AKSUM UNIVERSITY

AKSUM UNIVERSITY

APPLIED MATHEMATICS III

APPLIED(Math
MATHEMATICS
331) III

(Math
CREDIT 2062) 4
HOURS:

CREDIT HOURS: 4

June15, 2014

June15, 2014
 APPLIED MATHEMATICS III
(MATH 2062)

COLLEGE OF NATURAL AND COMPUTATIONAL SCIENCES

DEPARTEMENT OF COMPUTATIONAL SCIENCE
PRORGRAM OF MATHEMATICS

Written By:
Badri Ahmed (MSc.)
Moges Birhanu (MSc.)
Tekleberhan Berhe (MSc.)

Edited By:

June15, 2014
Course Description

Topics discussed in the course include methods of solving first-order differential equations,
second-order linear equations, Fourier series and integrals, Fourier and Laplace transformations,
Vector Calculus, divergence, Curl, line integral, Green’s and Stokes’ Theorem, functions of
complex variables, Cauchy integral theorem and formula.

Course Objective

At the successful Completion of this course student will be able to:

 Define the concept of ordinary differential equations.
 Distinguish separable, homogeneous and exact differential equations.
 Compute general and particular solutions of homogenous differential equations.
 Appreciate the applications of differential equation.
 Define the Fourier series and integrals, Fourier and Laplace Transformations
 Use the concept of vector and scalar fields to define arc length, curvature, gradient,
divergence and curl.
 Compute line and surface integrals.
 Appreciate the applications of vector differential and integral calculus.
 Use the concept of complex numbers,Complex Analytic function and Complex integrals.
 Compute limits, derivatives and integrals of complex variables.

i
MODULE INTRODUCTION

Applied Mathematics is an inherently interdisciplinary subject, which covers a wide spectrum of

scientific activities. It is the mathematics dealing with problems arising in the physical, life and
social sciences as well as in engineering, and provides a broad qualitative and quantitative
background for use in these fields.The methods of mathematical modeling and analysis provide a
unification and mutual enrichment of ideas from many different areas, and a deeper
understanding of the fields to which it is applied. Applied Mathematics draws upon concepts and
methods of mathematics from the fields of application and in turn, brings ideas, techniques and
scientific knowledge back to influence the development of mathematics.Owing to its nature,
Applied Mathematics appeals to people with a variety of different interests, ranging from those
with a desire to obtain a good quantitative background for use in some future career, to those
who wish to have a better understanding of the basic mathematical aspects of other fields, or to
those who are interested in the basic techniques and approaches in themselves.

The basic mathematical skills of Applied Mathematics come from a variety of sources, which
depend on the field of interest: the theory of ordinary and partial differential equations, statistical
sciences, probability and decision theory, operational analysis, optimization theory, the
mechanics of solid materials and of fluids flows, numerical analysis, scientific computation and
the science of modern computer-based modeling.
This module is prepared for those students who take as supportive especially for the students
study on engineering and other fields.
The courses in Applied Mathematics are designed for students with a wide range of goals and are
not limited to the needs of students following an applied mathematics concentration.
The module covers all the basic ideas which is used to study applied mathematics III.
This module must build upon the presumption that students studying the subject have already
gained some knowledge in applied mathematics I and II. In this module, you will learn the
concept of Ordinary differential equation with order one and two, Fourier series and integrals,
and vector calculus. You will also learn complex analytic functions and complex integrals. You
may have seen many solved examples, exercises and miscellaneous exercises at the end of each
chapter. We try to balance each examples and exercises to the level of students to understand it
easily.

ii
CONTENTS Page
Chapter One: ORDINARY DIFFERENTIAL EQUATIONS OF THE FIRST ORDER

1.1 Basic concepts and Ideas……………………………………………………………………………………………2

1.2 Separable Equations………………………………………………………………………………………………….6
1.3 Homogeneous equations…………………………………………………………………………………………..7
1.4 Exact Differential Equations………………………………………………………………….......................9
1.5 Integration Factors ………………………………………………………………………………………………….14
1.6 Linear First – Order Differential Equations……………………………………………………………….18

Chapter Two: ORDINARY LINEAR DIFFERENTIAL EQUATIONS OF THE SECOND ORDER

2.1 Homogeneous Linear Equations Of the second order……………………………………27

2.2 General Solution, Basis…………………………………………………………………………………..29

2.3 Homogeneous Second order Equations with constant coefficients…………………34

2.4 A Method for solving Non Homogeneous Linear Equations……………………………42

2.5 Systems of Differential Equations………………………………………………………………….48

Chapter Three: FOURIER SERIES AND INTEGRALS

3.1 Periodic Functions; Trigonometric Series………………………………………………………….61

3.2 Fourier Series and Fourier Integral…………………………………………………………………66

3.3 The Complex Fourier Series and Integrals…………………………………………………………80

Chapter Four: FOURIER AND LAPLACE TRANSFORMATIONS

4 .1 Fourier Cosine and Sine Transformation…………………………………………………92

4.2 Fourier Transformation…………………………………………………………………………..99

4.3 Laplace Transformation…………………………………………………………………………107

4.4 Differentiation and Integrations of Laplace Transformation…………………123

4.5 Convolution and Integral Equations…………………………………………………………………127

iii
Chapter Five: Vector Calculus

5.1 Scalar Fields and Vector Fields……………………………………………………………………140

5.2 Vector Calculus………………………………………………………………………………………………145

5.3 Curves, Arc Length and Tangent………………………………………………………………………149

5.4 Gradient of A Scalar Field; Divergence and Curl of a vector field………………………160

5.5 Line Integrals and Green’s Theorem………………………………………………………………..165

5.6 Surface Integral; Divergence Theorem of Gauss; Applications……………………………182

5.7 Stock’s Theorem; Applications…………………………………………………………………………… 195

5.8 Line Integral Independent of Path……………………………………………………………………… 200

Chapter Six: COMPLEX ANALYTIC FUNCTIONS

6.1 Complex Numbers,TheTriangle Inequality…………………………………………………………..209

6.2 Functions of Complex Variables, Limit, Derivative, Analytic Function…………………218

6.3 Cauchy- Riemann Equations, Laplace Equation………………………………………………….223

6.4 Elementary Functions, Exponential, Trigonometric…………………………………………..226

6.5 Hyperbolic and Logarithm Functions; General Power…………………………………….229

Chapter Seven: COMPLEX INTEGRALS

7.1 Line integral in the Complex Plane……………………………………………………………………..239

7.2 Cauchy Integral Theorem and Formula………………………………………………………… ….247

7.3 The Derivative of Analytic Function…………………………………………………………………..253

iv
 APPLIED MATHEMATICS III

Chapter One

Ordinary Differential Equations of the First Order

Introduction:

The study of physical phenomena becomes consistent and applicable by establishing

mathematical relationships between the involved physical quantities. Sometimes, these
relationships are algebraic. But in most cases, algebra is not enough to characterize the
phenomenon. The involved quantities may depend on other quantities, considered as independent
variables, and the relationships are no more algebraic, containing both the unknown function and
its derivatives. In the case of functions depending only on one variable, these are called ordinary
differential equations (ODEs). If the unknown function depends on several variables, the
equations will also contain its partial derivatives; such equations are called partial differential
equations (PDEs). In this introductory chapter we define ordinary differential equations, give
examples showing how they are used and show how to find solutions of some differential
equations of the first order.

Unit Objectives:

Upon successful completion of this chapter, the student will be able to:

 Define the very basic definitions and terminology of differential equations as well as a
discussion of central issues and objectives for the course.
 Classify Ordinary Differential Equations (ODEs) and distinguish ODEs from Partial
Differential Equations (PDEs).
 Explain what is meant by an integrating factor, find the integrating factor for a linear
first-order equation with constant coefficients and use it to solve the equation.
 Solve “Exact” and “Homogeneous” equations.
 Decide which (if any) of the above methods can be used to solve a given first-order
differential equation.
 Use a given change of variable to transform a first-order differential equation into one
that can more easily be solved.

Badri A, Moges B. and Teklebrhan B. 1 AKU

 APPLIED MATHEMATICS III

1.1Basic Concepts (Some Definitions and Terminology)

Section Objectives:

At the end of this section, students should able to:

 Define the term differential equation, ordinary differential equation and partial
differential equation.
 Define the order and degree of differential equation.

Definition 1.1 A Differential Equation (DE) is an equation involving a dependent variable and its
derivative with respect to one or more independent variable.

Definition 1.2 A differential equation involves functions of only a single variable is called an
Ordinary Differential Equation (ODE). If a differential equation contains only ordinary derivatives
of one or more unknown functions with respect to a single independent variable, it is said to be an
ordinary differential equation (ODE).

The PDE (Partial Differential Equation) is an equation which involves partial derivatives of
an unknown function of two or more variables. That is, the derivatives of functions of more than
one variable.
Definition 1.3 The order of the equation is the order of the highest derivatives that appears

Definition 1.4 The degree of a DE is the highest exponent of the highest order derivatives which
involves in the DE.

An ODE is said to be order n is the nth derivative of the unknown function Y is the highest
derivative of Y in the equation.

Badri A, Moges B. and Teklebrhan B. 2 AKU

 APPLIED MATHEMATICS III

Examples 1.1

Equation Type Order

1 𝑦 ′ = 𝑐𝑜𝑠𝑥 ODE first
2 𝜕 2𝑢 𝜕 2𝑢 PDE second
+ =0
𝜕𝑥 2 𝜕𝑦 2
3 𝑦 " + 9𝑦 = 0 ODE second
4 𝑑3 𝑑𝑦
2 𝑥
𝑑3 ODE third
𝑥 + 2𝑒 = (𝑥 2 + 2)2
𝑑𝑥 𝑑𝑥 𝑑𝑥

Example 1.2 In the differential equation

differential equation Degree Order

1 𝑦′ + 𝑥3𝑦 = 𝑥 1 1
2 𝑑3 𝑦 2 𝑑𝑦 2 3
( ) + = 𝑐𝑜𝑠𝑥
𝑑𝑥 3 𝑑𝑥
3 (𝑦 ′′′ )2 − (𝑦 ′ )𝑥 = 𝑥 − 1 2 3
4 √𝑦 ′ = 𝑦 ′′′ + 1 2 3
′′
5 𝑦 ′′ = 𝑒 𝑦 No 2

Notation: The first ordered differential equation can be written as

𝐹(𝑥, 𝑦, 𝑦 ′ ) = 0 (Implicit form) (1)
Or often in the form
𝑦’ = 𝑓(𝑥, 𝑦) (Explicit form)

Example 1. 3: The implicit ODE 𝑥 3 𝑦 ′ − 4𝑦 2 = 0 (Where x≠ 0)

can be written explicit as
𝑦′ =4𝑥 −3 𝑦 2
Generally, the ordinary differential equation (ODE) written as
𝐹(𝑥, 𝑦, 𝑦 (′) … … … … . 𝑦 (𝑛) ) = 0
The number n is called the order of ODE.

Definition 1.5 The most general nth order linear differential equation can be written
𝑎0 (𝑥)𝑦 𝑛 + 𝑎1 (𝑥)𝑦 𝑛−1 + ⋯ + 𝑎𝑛−1 (𝑥)𝑦 = 𝑓(𝑥) *
Badri A, Moges
where, 𝑎 (𝑥), 𝑎B.(𝑥)
and…Teklebrhan B. the coefficient of the
. 𝑎 (𝑥) called 3
equation AKU
0 1 𝑛
 APPLIED MATHEMATICS III

The known function 𝑓(𝑥) is called non homogeneous term, equation (*) is called homogeneous
if 𝑓(𝑥) = 0

If the coefficients in (*) are constant, so that (*) becomes

𝑎0 𝑦 𝑛 + 𝑎1 𝑦 𝑛−1 + ⋯ + 𝑎𝑛−1 𝑦 = 𝑓(𝑥)

which is called constant coefficient linear equation. 𝑎0 ≠ 0 , otherwise the equation would not
be the nth order. An ordinary differential equation that is not linear is said to be non-linear.

Example 1.4
ODE Property
𝑑𝑦
1. + 2𝑥𝑦 = 𝑠𝑖𝑛𝑥 linear variable coefficient,
𝑑𝑥

non homogeneous first order equation

𝑑3 𝑦 𝑑𝑦
2. 𝑥 3 𝑑𝑥 3 − 2𝑥 𝑑𝑥 + 6𝑦 = 𝑒 𝑥 linear variable coefficient, non-homogeneous, 3rd

order equation,
𝑑2 𝑦 𝑑𝑦
3. + 𝑎 𝑑𝑥 + 𝑏𝑦 = 𝑠𝑖𝑛𝜑𝑥 𝑤𝑖𝑡ℎ 𝜑 − 𝑐𝑜𝑛𝑠𝑡𝑎𝑛𝑡 linear variable coefficient
𝑑𝑥 2

non homogeneous 2nd order equation

𝑑2 𝜃
4. + 𝑘 𝑠𝑖𝑛𝜃 = 0 with 𝑘-constant non-linear 2nd order equation ,because 𝜃
𝑑𝑡 2

occurs non-linearly in the function sin𝜃

𝑑4 𝑦
5. + 𝑦2 = 0 non-linear, homogeneous, 4th order equation .
𝑑𝑥 4

Solutions of a Differential Equation

Definition 1.6 A function relation between two variables (dependent and independent variables)
which satisfy the given DE is called the solution of an ODE or integral curve. A solution of an nth
order equation that contains n arbitrary constants is called the general solution of the equation. If the
arbitrary constants in the general solution are assigned specific value, the result is called a particular
solution of the equation.

Note: It is possible to have more than one solution of a differential equation. For instance 𝑦 =
2𝑥 3 + 𝐴 and 𝑦 = 2𝑥 3 are solution of the differential equation 𝑦 ′ = 12𝑥.

Badri A, Moges B. and Teklebrhan B. 4 AKU

 APPLIED MATHEMATICS III

Definition 1.7 An ODE together with an initial condition is called an initial value problem.
Thus, if the ordinary differential equation is explicit
𝑦 ′ = 𝑓(𝑥, 𝑦)
the initial value problem is of the form
𝑦 ′ = 𝑓(𝑥, 𝑦) , 𝑦(𝑥0 ) = 𝑦0

Example 1.5 : Solve the initial value problem

𝑦 ′ = 3𝑦 , 𝑦(0) = 5.7
Existence and Uniqueness: Two fundamental questions arise in considering an initial –value
problem:

𝑑𝑦
 Existence : Does the differential equation 𝑑𝑥 = 𝑓(𝑥, 𝑦) possess solutions?
Do any of the solution curves pass through the point (𝑥0 , 𝑦0 )?
 Uniqueness : When can we be certain that there is precisely one solution curve passing
through the point (𝑥0 , 𝑦0 )?

Theorem: (Existence of a Unique Solution)

Let 𝑅 be a rectangular region in the 𝑥𝑦 plane defined by 𝑎 ≤ 𝑥 ≤ 𝑏, 𝑐 ≤ 𝑦 ≤ 𝑑 that contains

𝜕𝑓
the point (𝑥0 , 𝑦0 ) and 𝜕𝑦 are continuous on 𝑅, then there exists some interval 𝐼0 : (𝑥0 − ℎ, 𝑥0 +
ℎ), ℎ > 0, contained in [𝑎, 𝑏] and a unique function 𝑦(𝑥), defined on 𝐼0 , that is a solution of the
initial value problem 𝑦 ′ = 𝑓(𝑥, 𝑦) , 𝑦(𝑥0 ) = 𝑦0.

Badri A, Moges B. and Teklebrhan B. 5 AKU

 APPLIED MATHEMATICS III

1.2 Solutions of First Order Differential Equation

Section Objectives:

At the end of this section, students should able to:

 Distinguish linear, separable, homogeneous and exact differential equation.

 Solve differential equations, using different methods.
 Use a given change of variable to transform a first-order differential equation into one
that can more easily be solved.
 Define linear differential equations of first order.
 Solve linear differential equations of first order.

1.1.1 Elementary Method

If the differential equation has the form

𝑦 ′ = 𝑓(𝑥) ,
then

𝑑𝑦
= 𝑓(𝑥) ⟹ 𝑑𝑦 = 𝑓(𝑥)𝑑𝑥
𝑑𝑥

which up on integration gives

∫ 𝑑𝑦 = ∫ 𝑓(𝑥)𝑑𝑥
⟹ 𝑦 = ∫ 𝑓(𝑥)𝑑𝑥 + 𝑐 , which is a general solution.

Example 1.6 a) Solve 𝑦 ′ = cos 𝑥

Solution: We have
𝑑𝑦
= 𝑐𝑜𝑠𝑥
𝑑𝑥
This implies that
𝑑𝑦 = 𝑐𝑜𝑠𝑥𝑑𝑥 … (*)
Integrate equation (*) both sides, i.e.
∫ 𝑑𝑦 = ∫ 𝑐𝑜𝑠𝑥𝑑𝑥
Hence,
𝑦 = 𝑠𝑖𝑛𝑥 + 𝑐

b) Solve 𝑦 ′ = ln 𝑥 (exercise)

Badri A, Moges B. and Teklebrhan B. 6 AKU

 APPLIED MATHEMATICS III

1.3 Separable equations

We begin our study of how to solve differential equations with the simplest of all differential
equations: first-order equations with separable variables. Because the method in this section and
many techniques for solving differential equations involve integration you are urged to refresh
du
your memory on important formulas (such as∫ ) and techniques (such as integration by parts)
u
by consulting a calculus text.

Definition 1.8 An equation 𝑦 ′ = 𝑓(𝑥, 𝑦) is called separable if it can be written in the

form

𝑦 ′ = 𝐹(𝑥)𝐺(𝑦) (*)

For some function 𝐹(𝑥) a dependent only on 𝑥 and 𝐺(𝑦) dependent on 𝑦.

Equation (*) itself to be variable separable type.

Solution Method:
Equation (*) can be written as
𝑑𝑦 1
= 𝐹(𝑥)𝐺(𝑦) => 𝑑𝑦 = 𝐹(𝑥)𝑑𝑥
𝑑𝑥 𝐺(𝑦)
1
⟹ ∫ 𝐺(𝑦) 𝑑𝑦 = ∫ 𝐹(𝑥)𝑑𝑥

⟹ 𝑀1 (𝑦) = 𝑀2 (𝑥) + 𝑐,

Where,
1
∫ 𝐺(𝑦) 𝑑𝑦 = 𝑀1 (𝑦) and ∫ 𝐹(𝑥)𝑑𝑥 = 𝑀2 (𝑥) + 𝑐

Example 1.7: Solve (1 + 𝑥)𝑑𝑦 − 𝑦𝑑𝑥 = 0

Solution: Dividing by(1 + 𝑥)𝑦, we can write
𝑑𝑦 𝑑𝑥
=
𝑦 1+𝑥
From which it follows that
𝑑𝑦 𝑑𝑥
∫ = ∫
𝑦 1+𝑥
𝑙𝑛|y| = |1 + 𝑥| + 𝑐1
𝑦 = 𝑒 𝑙𝑛|1+x|+𝑐1 = 𝑒 𝑙𝑛|1+x| . 𝑒 𝑐1

Badri A, Moges B. and Teklebrhan B. 7 AKU

 APPLIED MATHEMATICS III

= |1 + 𝑥|. 𝑒 𝑐1
= ±. 𝑒 𝑐1 (1 + 𝑥)
Replacing, 𝑐 = ±. 𝑒 𝑐1 then gives
𝑦 = 𝑐(1 + 𝑥)
Example 1.8: (An initial value problem)
𝑑𝑦
Solve (𝑒 2𝑦 − 𝑦)𝑐𝑜𝑠𝑥 = 𝑒 𝑦 𝑠𝑖𝑛2𝑥 , 𝑦(0) = 0
𝑑𝑥

Solution: Dividing the equation by 𝑒 𝑦 𝑐𝑜𝑠𝑥 , gives

𝑒 2𝑦 −𝑦 𝑠𝑖𝑛2𝑥
𝑑𝑦 = dx
𝑒𝑦 𝑐𝑜𝑠𝑥

Before integrating we use term wise division on the left-hand side and the trigonometric identity
𝑠𝑖𝑛2𝑥 = 2𝑠𝑖𝑛𝑥𝑐𝑜𝑠𝑥
on the right-hand side. Then

∫(𝑒 𝑦 − 𝑦𝑒 −𝑦 )𝑑𝑦 = 2 ∫ 𝑠𝑖𝑛𝑥𝑑𝑥

Using Integration by parts yields:

𝑒 𝑦 + 𝑦𝑒 −𝑦 + 𝑒 −𝑦 = −2𝑐𝑜𝑠𝑥 + 𝑐
The initial condition 𝑦 = 0 when 𝑥 = 0 implies 𝑐 = 4.
Thus, a solution of the initial value problem is
𝑒 𝑦 + 𝑦𝑒 −𝑦 + 𝑒 −𝑦 = 4 − 2𝑐𝑜𝑠𝑥

Exercises

Solve the given differential equation by separation of variables

𝑑𝑦 𝑑𝑦
( a) 𝑦 ′ = 1 + 𝑦 2 ( c) = 𝑠𝑖𝑛5𝑥 ( e) 𝑥𝑦 ′ = 4𝑦 (g) = 𝑒 3𝑥+2𝑦
𝑑𝑥 𝑑𝑥

𝑦−1
( b) 𝑦 ′ = 𝑥 2 (1 + 𝑦) (d) 𝑑𝑥 + 𝑒 3𝑥 𝑑𝑦 = 0 (f) 𝑦 ′ + 1−𝑥 = 0

Badri A, Moges B. and Teklebrhan B. 8 AKU

 APPLIED MATHEMATICS III

1.4 Homogenous Equations

Sometimes, the best way of solving a DE is to use a change of variables that will put the DE into
a form whose solution method we know. We now consider a class of DEs that are not directly
solvable by separation of variables, but, through a change of variables, can be solved by that
method.
Definition 1.9 A function 𝑓(𝑥, 𝑦) is said to be algebraically homogenous of degree 𝑛, or simply
homogenous of degree 𝑛,
if
𝑓(𝑘𝑥, 𝑘𝑦) = 𝑘 𝑛 𝑓(𝑥, 𝑦),
for some real number 𝑛 and all 𝑘 > 0 for 𝑓(𝑥, 𝑦) ≠ (0,0).
Example 1.9.
a) 𝑓(𝑥, 𝑦) = 𝑥 2 + 3𝑥𝑦 + 4𝑦 2 is homogenous
b) Show that 𝑓(𝑥, 𝑦) = ln|𝑦| − ln|𝑥| for (𝑥, 𝑦) ≠ (0,0) is homogenous of degree 0.

Note: A homogenous function of degree 𝑛 in 𝑥 and 𝑦 can also be expressed as

𝑥 𝑦
𝑦 𝑛 𝐹(𝑦) or 𝑥 𝑛 𝐹(𝑥 ).

Example 1.10
𝑥 2 +𝑦 2
a) 𝑓(𝑥, 𝑦) = 𝑥 3 +𝑦 3 , homogenous

b) 𝑓(𝑥, 𝑦) = 𝑥 2 + 4 , not homogenous

c) 𝑓(𝑥, 𝑦) = 𝑥 2 + 3𝑥𝑦 + 4𝑦 2 , homogenous
d) 𝑓(𝑥, 𝑦) = tan(𝑥𝑦 + 1) , not homogenous.

Definition 1.10 The first order ODE in differential form

𝑃(𝑥, 𝑦) + 𝑄(𝑥, 𝑦) = 0
is called homogenous if 𝑃 and 𝑄 are homogeneous function of the same degree or equivalently,
if when written in the form
𝑑𝑦
= 𝑓(𝑥, 𝑦),
𝑑𝑥
𝑦
the function 𝑓(𝑥, 𝑦) can be written as 𝑓(𝑥, 𝑦) = 𝑔(𝑥 )

Badri A, Moges B. and Teklebrhan B. 9 AKU

 APPLIED MATHEMATICS III

General Solution of a Homogenous Differential Equation

The substitution 𝑦 = 𝑢𝑥 will reduce either form of the homogenous equation to an equation
involving the dependent variable 𝑥 and the new dependent variable 𝑢 in which the variable
separable. Let
𝑦 ′ = 𝑓(𝑥, 𝑦),
is a homogenous of degree 0.
𝑦 𝑦
𝑓(𝑥, 𝑦) = 𝐹 (𝑥 ) or 𝑦 ′ = 𝐹(𝑥 )
Now,
𝑑𝑦 𝑑𝑢
𝑦 = 𝑢𝑥 ⟹ =𝑢+𝑥
𝑑𝑥 𝑑𝑥
⟹ 𝑑𝑦 = 𝑢𝑑𝑥 + 𝑥𝑑𝑢
𝑦 𝑑𝑢
𝐹 ( ) = 𝐹(𝑢) = 𝑦 ′ = 𝑢 + 𝑥
𝑥 𝑑𝑥
It follows that,
𝑑𝑢
𝑥 = 𝐹(𝑢) − 𝑢
𝑑𝑥
𝑑𝑥 𝑑𝑢
⟹ = 𝐹(𝑢)−𝑢 (Separable variable)
𝑥

𝑑𝑥 𝑑𝑢
∫ =∫
𝑥 𝐹(𝑢) − 𝑢
Hence the solution of the given differential equation becomes
𝑑𝑢
𝑙𝑛|𝑥| = ∫ 𝐹(𝑢)−𝑢 + 𝑐

Example 1.11: Solve the differential equation

a) 2𝑥 2 𝑦 ′ = 𝑥 2 + 𝑦 2
Solution: If we divide both sides by 2𝑥 2 then we obtain
1 1 𝑥 2
𝑦′ = + ( )
2 2 𝑦
Which is homogeneous. Now, setting
𝑦
𝑢= ⟹ 𝑦 = 𝑢𝑥
𝑥
,yields the equation
1 2 1
𝑥𝑢′ = 𝑢 −𝑢+
2 2
After rearrange
2𝑢′ 1
2
=
(𝑢 − 1) 𝑥
Then integrating yields

Badri A, Moges B. and Teklebrhan B. 10 AKU

 APPLIED MATHEMATICS III

2𝑑𝑢 1
∫ 2
= ∫ 𝑑𝑥
(𝑢 − 1) 𝑥
−2
= 𝑙𝑛(𝑥) + 𝑐
𝑢−1
Solving for 𝑢 gives
2
𝑢 = 1−
ln(𝑥) + 𝑐
Hence,
2𝑥
𝑦 = 𝑥 − ln(𝑥)+𝑐

Example 1.12: Solve the differential equation

′
𝑥2 + 𝑦2
𝑦 =
𝑥𝑦

Solution: If we divide the numerator and denominator of the fraction by 𝑥 2 ,we obtain

𝑦
1 + (𝑥 )2
𝑦′ = 𝑦
(𝑥 )

𝑦
which is homogeneous. Now setting 𝑢 = 𝑥 or 𝑦 = 𝑢𝑥

1+𝑢2 1
Yields, 𝑥𝑢′ = −𝑢 =𝑢
𝑢

Separating and integrating yields

1
∫ 𝑢𝑑𝑢 = ∫ 𝑑𝑥
𝑥

𝑢2
⟹ = ln(𝑥) + 𝑐
2

Solving for 𝑢 gives 𝑢 = √2 ln(𝑥) + 𝑐

and then

𝑦 = √2 ln(𝑥) + 𝑐

Exercises

Badri A, Moges B. and Teklebrhan B. 11 AKU

 APPLIED MATHEMATICS III

Solve a) (𝑦 2 + 2𝑥𝑦)𝑑𝑥 − 𝑥 2 𝑑𝑦 = 0
𝑏) (𝑥 2 + 𝑦 2 )𝑑𝑥 = 2𝑥𝑦 𝑑𝑦
c) 𝑥 2 𝑦𝑑𝑥 − (𝑥 3 + 𝑦 3 )𝑑𝑦 = 0

1.4 Exact Differential Equation

Definition 1.11: The first ordered differential equation

𝑀(𝑥, 𝑦)𝑑𝑥 + 𝑁(𝑥, 𝑦)𝑑𝑦 = 0
is side to be exact if a function 𝐹(𝑥, 𝑦) exists such that the total differential
𝑑[𝐹(𝑥, 𝑦)] = 𝑀(𝑥, 𝑦)𝑑𝑥 + 𝑁(𝑥, 𝑦)𝑑𝑦
Test for exactness:
𝜕𝑀 𝜕𝑁
The DE 𝑀(𝑥, 𝑦)𝑑𝑥 + 𝑁(𝑥, 𝑦)𝑑𝑦 = 0 is exact if and only if =
𝜕𝑦 𝜕𝑥

Method of solution: Given an equation in the differential form

𝑀(𝑥, 𝑦)𝑑𝑥 + 𝑁(𝑥, 𝑦)𝑑𝑦 = 0
, which is exact. Then the solution can be obtained using the following systematic way.
𝜕𝑓 𝜕𝑓
Let (a) 𝜕𝑥 = 𝑀 and (b) =𝑁
𝜕𝑦

By integrating (a) with respect to x, we obtain

𝑓 = ∫ 𝑀𝑑𝑥 + 𝑘(𝑦), (*)

𝑘(𝑦), plays the role of a constant .

𝜕𝑓 𝑑𝑘
To determine𝑘(𝑦), we derive (differentiate with respect to y) from (*). Use (b) to get 𝑑𝑦 and
𝜕𝑦
𝑑𝑘
then integrating 𝑑𝑦 to get k.

Example 1.13: Solve 2𝑥𝑦𝑑𝑥 + (𝑥 2 − 1)𝑑𝑦 = 0

Solution: With 𝑀(𝑥, 𝑦) = 2𝑥𝑦 and 𝑁(𝑥, 𝑦) = 𝑥 2 − 1, we have
𝜕𝑀 𝜕𝑁
= 2𝑥 =
𝜕𝑦 𝜕𝑥
Thus, the equation is exact, and so there exist a function 𝑓(𝑥, 𝑦) such that
𝜕𝑓 𝜕𝑓
= 2𝑥𝑦 and = 𝑥2 − 1
𝜕𝑥 𝜕𝑦

From the first of these equations, we obtain after integrating

Badri A, Moges B. and Teklebrhan B. 12 AKU

 APPLIED MATHEMATICS III

𝑓(𝑥, 𝑦) = 𝑥 2 𝑦 + 𝑘(𝑦)
Taking the partial derivative of the last expression with respect to y and setting the result equal to
𝑁(𝑥, 𝑦), gives
𝜕𝑓
= 𝑥 2 + 𝑘 ′ (𝑦) = 𝑥 2 − 1
𝜕𝑦
It follows that 𝑘 ′ (𝑦) = −1 and 𝑘(𝑦) = −𝑦.
Hence 𝑓(𝑥, 𝑦) = 𝑥 2 𝑦 − 𝑦
Note: in the above example, the equation could be solved by separation of variables.
Example 1.14: Solve (𝑒 2𝑦 − 𝑦𝑐𝑜𝑠𝑥𝑦)𝑑𝑥 + (2𝑥𝑒 2𝑦 − 𝑥𝑐𝑜𝑠𝑥𝑦 + 2𝑦)𝑑𝑦 = 0
Solution: The equation is exact because
𝜕𝑀 𝜕𝑁
= 2𝑒 2𝑦 + 𝑥𝑦𝑠𝑖𝑛𝑥𝑦 − 𝑐𝑜𝑠𝑥𝑦 =
𝜕𝑦 𝜕𝑥
Hence a function 𝑓(𝑥, 𝑦) exists for which
𝜕𝑓 𝜕𝑓
𝑀(𝑥, 𝑦) = 𝜕𝑥 and 𝑁(𝑥, 𝑦) = 𝜕𝑦
𝜕𝑓
Now for variety we shall start with the assumption that = 𝑁(𝑥, 𝑦) ;that is
𝜕𝑦
𝜕𝑓
=2𝑥𝑒 2𝑦 − 𝑥𝑐𝑜𝑠𝑥𝑦 + 2𝑦
𝜕𝑦

𝑓(𝑥, 𝑦) = 2𝑥 ∫ 𝑒 2𝑦 𝑑𝑦 − 𝑥 ∫ 𝑐𝑜𝑠𝑥𝑦𝑑𝑥 + 2 ∫ 𝑦𝑑𝑦.

It follows that
𝑓(𝑥, 𝑦) = 𝑥𝑒 2𝑦 − 𝑠𝑖𝑛𝑥𝑦 + 𝑦 2 + ℎ(𝑥)
𝜕𝑓
= 𝑒 2𝑦 − 𝑦𝑐𝑜𝑠𝑥𝑦 − ℎ′ (𝑥) = 𝑒 2𝑦 − 𝑦𝑐𝑜𝑠𝑥𝑦
𝜕𝑥
and so
ℎ′ (𝑥) = 0 or ℎ(𝑥) = 𝑐.
Hence a family of solution is
𝑥𝑒 2𝑦 − 𝑠𝑖𝑛𝑥𝑦 + 𝑦 2 + 𝑐 = 0.

Exercises

Solve the following

𝑦−2𝑥−1
a) 𝑦 ′ = 2𝑦−𝑥−1 b) (𝑐𝑜𝑠 𝑦 𝑠𝑖𝑛 ℎ𝑥 + 1)𝑑𝑥 − 𝑠𝑖𝑛 𝑦 𝑐𝑜𝑠 ℎ𝑥 𝑑𝑦 = 0

Badri A, Moges B. and Teklebrhan B. 13 AKU

 APPLIED MATHEMATICS III

c) 𝑥 2 𝑦𝑑𝑥 − (𝑥 3 + 𝑦 3 )𝑑𝑦 = 0 d) 𝑥 2 𝑦 3 + 2𝑦 )𝑑𝑥 + (2𝑥 − 2𝑥 3 𝑦 2 )𝑑𝑦 = 0

1.5 Reduction to Exact form, integrating factors

If the partial derivatives are not equal i.e. for a non-exact differential equation 𝑀(𝑥, 𝑦)𝑑𝑥 +
𝑁(𝑥, 𝑦)𝑑𝑦 = 0. It is sometimes possible to find an integrating factor 𝑰(𝑥 , 𝑦) so that after
multiplying, the left-hand side of
𝐼(𝑥, 𝑦)𝑀(𝑥, 𝑦)𝑑𝑥 + 𝐼(𝑥, 𝑦)𝑁(𝑥, 𝑦)𝑑𝑦 = 0 (a)
is an exact differential. Equation (a) is exact if and only if (𝐼𝑀)𝑦 = (𝐼𝑁 )𝑥, where the
subscripts denote partial derivatives. By the Product Rule of differentiation the last equation is
the same as
𝐼𝑀𝑦 + 𝐼𝑦𝑀 = 𝐼𝑁𝑥 + 𝐼𝑥 𝑁 or
𝐼𝑥 𝑁 + 𝐼𝑦𝑀 = (𝑀𝑦 _ 𝑁𝑥)𝐼. (b)

Definition1. 12: An integrating factor is a factor which we multiply the given non exact equation
to make it exact.

How to Find Integrating Factor

Finding 𝐼(𝑥, 𝑦) in general, requires solving the partial differential equation

𝐼𝑦 𝑀 − 𝐼𝑥 𝑁 + 𝐼(𝑀𝑦 − 𝑁𝑥 ) = 0
This is just as tricky to solve as the original equation. Only in a few special cases are there
methods for computing the integrating factor 𝐼(𝑥, 𝑦).
Case1: Suppose we want to see if there exists an integrating factor that depends only on x (and
𝜕𝐼
not on y). Then, 𝜕𝑦 would be zero, since 𝐼 does not depend on y, and so 𝐼(𝑥) we need to satisfy
𝐼′ 𝑀𝑦 −𝑁𝑥
= .
𝐼 𝑁
𝑀𝑦 −𝑁𝑥
This can only happen if the ratio is a function of 𝑃(𝑥) only of 𝑥 (and not𝑦), then
𝑁

𝐼(𝑥) = 𝑒 ∫ 𝑃(𝑥)𝑑𝑥

Badri A, Moges B. and Teklebrhan B. 14 AKU

 APPLIED MATHEMATICS III

Case2: We see also look to see if there is an integrating factor that depends only on 𝑦 and not
𝜕𝐼
on 𝑥 .We can do the same calculation ,this time using 𝜕𝑦 would be zero, to see that such an
𝑁𝑥 −𝑀𝑦
integrating factor exists if the ratio is a function of 𝑄(𝑦) only of 𝑦 (and not x); then
𝑀

𝐼(𝑦) = 𝑒 ∫ 𝑄(𝑦)𝑑𝑦
Example 1.15: Solve for 𝑦(𝑥), if (2𝑥𝑦 2 − 4𝑦) + (3𝑥 2 𝑦 − 8𝑥)𝑦 ′ = 0
Solution: In differential form the equation is
(2𝑥𝑦 2 − 4𝑦)𝑑𝑥 + (3𝑥 2 𝑦 − 8𝑥)𝑑𝑦 = 0
Therefore,
𝑀 = 2𝑥𝑦 2 − 4𝑦 and 𝑁 = 3𝑥 2 𝑦 − 8𝑥
𝑀𝑦 = 4𝑥𝑦 − 4 and 𝑁𝑥 = 6𝑥𝑦 − 8
These are not equal (i.e. 𝑀𝑦 ≠ 𝑁𝑥 ), so the equation is not exact. We look for integrating factors
using the two criteria we know.
𝑀𝑦 −𝑁𝑥 −2𝑥𝑦+4
 First we have = is not a function of 𝑥 only.
𝑁 3𝑥 2 𝑦−8𝑥

𝑁𝑥 −𝑀𝑦 2𝑥𝑦—4 1
 Second we have = = 𝑦 is a function of 𝑦 only.
𝑀 2𝑥𝑦 2 −4𝑦

Therefore, we need to multiply by the integrating factor

1
∫(𝑦)𝑑𝑦
𝐼(𝑦) = 𝑒 =𝑦
Now, our new equation is
(2𝑥𝑦 3 − 4𝑦 2 ) + (3𝑥 2 𝑦 2 − 8𝑥𝑦)𝑦 ′ = 0
i.e.
(2𝑥𝑦 3 − 4𝑦 2 )𝑑𝑥 + (3𝑥 2 𝑦 2 − 8𝑥𝑦)𝑑𝑦 = 0
𝑀(𝑥, 𝑦) = 2𝑥𝑦 3 − 4𝑦 2 and 𝑁(𝑥, 𝑦) = 3𝑥 2 𝑦 2 − 8𝑥𝑦
To find 𝑓 with
𝑓𝑥 = 2𝑥𝑦 3 − 4𝑦 2 and 𝑓𝑦 = 3𝑥 2 𝑦 2 − 8𝑥𝑦
Taking the anti-partial derivative of the first equation gives:
𝑓(𝑥, 𝑦) = 𝑥 2 𝑦 3 − 4𝑥𝑦 2 + 𝑘(𝑦) … (*)

To find 𝑘(𝑦) differentiate equation (*) with respect to 𝑦,we obtain

𝑓𝑦 (𝑥, 𝑦) = 3𝑥 2 𝑦 2 − 8𝑥𝑦 + 𝑘 ′ (𝑦).

Badri A, Moges B. and Teklebrhan B. 15 AKU

 APPLIED MATHEMATICS III

But , we have
𝑓𝑦 (𝑥, 𝑦) = 3𝑥 2 𝑦 2 − 8𝑥𝑦
Then
3𝑥 2 𝑦 2 − 8𝑥𝑦 = 3𝑥 2 𝑦 2 − 8𝑥𝑦 + 𝑘 ′ (𝑦)
It follows that,
𝑘 ′ (𝑦) = 0
Integrating both sides ,
𝑘(𝑦) = 𝑐.
Therefore our solutions are given implicitly by 𝑥 2 𝑦 3 − 4𝑥𝑦 2 = 𝑐
Example 1.16: Solve 𝑥𝑦 𝑑𝑥 + (2𝑥 2 + 3𝑦 2 − 20)𝑑𝑦 = 0
Solution: We have 𝑀 = 𝑥𝑦 and 𝑁 = 2𝑥 2 + 3𝑦 2 − 20.
We find the partial derivatives 𝑀𝑦 = 𝑥 and 𝑁𝑥 = 4𝑥. Since 𝑀𝑦 ≠ 𝑁𝑥 ,the nonlinear first-
order differential equation is not exact. We have
𝑀𝑦 −𝑁𝑥 𝑥 − 4𝑥 −3𝑥
= 2 = 2
𝑁 2𝑥 + 3𝑦 − 20 2𝑥 + 3𝑦 2 − 20
2

Depends on 𝑥 and 𝑦. But

𝑁𝑥 −𝑀𝑦 4𝑥 − 𝑥 3𝑥 3
= = =
𝑀 𝑥𝑦 𝑥𝑦 𝑦
The integrating factor is then
3
∫𝑦𝑑𝑦 3
𝑒 = 𝑒 3𝑙𝑛𝑦 = 𝑒 𝑙𝑛𝑦 = 𝑦 3
After we multiply the given differential equation by 𝐼(𝑦) = 𝑦 3 the resulting equation is
𝑥𝑦 4 𝑑𝑥 + (2𝑥 2 𝑦 3 + 3𝑦 5 − 20𝑦 3 )𝑑𝑦 = 0.
You should verify that the last equation is now exact as well as show using exact method, the
solution is
1 2 4 1 6
𝑥 𝑦 + 𝑦 − 5𝑦 4 = 𝑐
2 2

Exercises

1. Determine whether the given differential equation is exact. If it is exact, solve it.

Badri A, Moges B. and Teklebrhan B. 16 AKU

 APPLIED MATHEMATICS III

a) (2𝑥 − 1)𝑑𝑥 + (3𝑦 + 7)𝑑𝑦 = 0

b) (2𝑥 + 𝑦)𝑑𝑥 − (𝑥 + 6𝑦)𝑑𝑦 = 0
c) (𝑠𝑖𝑛𝑦 − 𝑦𝑠𝑖𝑛𝑥)𝑑𝑥 = (𝑐𝑜𝑠𝑥 + 𝑥𝑐𝑜𝑠𝑦 − 𝑦)𝑑𝑦 = 0
d) (2𝑥𝑦 2 − 3)𝑑𝑥 + (2𝑥 2 𝑦 + 4)𝑑𝑦 = 0
𝑦
e) (1 + 𝑙𝑛𝑥 + 𝑥 ) 𝑑𝑥 = (1 − 𝑙𝑛𝑥)𝑑𝑦

f) (5𝑦 − 2𝑥)𝑦 ′ − 2𝑦 = 0
g) (𝑡𝑎𝑛𝑥 − 𝑠𝑖𝑛𝑥𝑠𝑖𝑛𝑦)𝑑𝑥 + 𝑐𝑜𝑠𝑥𝑐𝑜𝑠𝑦𝑑𝑦 = 0
𝑑𝑦
h) 𝑥 𝑑𝑥 = 2𝑥𝑒 𝑥 − 𝑦 + 6𝑥 2

2. Solve the given differential equation by finding, an appropriate integrating factor.

3 2
a) 𝑦𝑑𝑥 − 𝑥𝑑𝑦 + 3𝑥 2 𝑦 2 𝑒 𝑥 = 0 d) 𝑐𝑜𝑠𝑥𝑑𝑥 + (1 + 𝑦 𝑠𝑖𝑛𝑥) 𝑑𝑦 = 0
−2 3𝑦
b) 𝑦′ = − 2𝑥 e) 𝑦(𝑥 + 𝑦 + 1)𝑑𝑥 + (𝑥 + 2𝑦)𝑑𝑦 = 0
𝑦

c) 6𝑥𝑦𝑑𝑥 + (4𝑦 + 9𝑥 2 )𝑑𝑦 = 0 f) (−𝑥𝑦𝑠𝑖𝑛𝑥 + 2𝑦𝑐𝑜𝑠𝑥)𝑑𝑥 + 2𝑥𝑐𝑜𝑠𝑥𝑑𝑦 = 0

Badri A, Moges B. and Teklebrhan B. 17 AKU

 APPLIED MATHEMATICS III

1.6 Linear Differential Equations of First Order

Linear ODEs that can be transformed to linear form are models of various phenomena, for
instance, in physics, biology, population dynamics, and ecology.

Definition13: A first order ODE is said to be linear if it can be written as:

𝑦 ′ + 𝑃(𝑥)𝑦 = 𝑟(𝑥) (Standard form ) (*)
Where 𝑝 and 𝑟 are a function of 𝑥.

If the first term is 𝑓(𝑥)𝑦’ (instead of 𝑦’), divided the equation by 𝑓(𝑥) to get the “ standard form
“ (*) with 𝑦’ as the first term.
For instance,
𝑦’𝑐𝑜𝑠𝑥 + 𝑦𝑠𝑖𝑛 𝑥 = 𝑥
is a linear ODE and its standard form is
𝑦 ′ + 𝑦 tan 𝑥 = 𝑥 sec 𝑥

To find the general solution of (a) we use an “ integrating factor”; we multiply the equation by a
function 𝐼(𝑥), to obtain
𝑑𝑦
𝐼(𝑥) 𝑑𝑥 + 𝐼(𝑥)𝑃(𝑥)𝑦 = 𝐼(𝑥)𝑟(𝑥) (b)

What we would like to happen is for

𝑑𝑦
𝐼(𝑥) + 𝐼(𝑥)𝑃(𝑥)𝑦
𝑑𝑥
to be the derivative of something nice. When written thus way, this sum looks sort of like the
output of the product rule. If we can find 𝐼(𝑥) so that the derivative of 𝐼(𝑥) is 𝐼(𝑥)𝑝(𝑥) then
this sum will be
𝑑
the derivative 𝑑𝑥 [𝐼(𝑥). 𝑦].What we want is

𝐼 ′ (𝑥) = 𝐼(𝑥)𝑝(𝑥)

This is now a (very easy) separable equation for the function 𝐼(𝑥), and the solution is

𝐼(𝑥) = 𝑒 ∫ 𝑝(𝑥)𝑑𝑥 .
If 𝑟(𝑥) = 0, then the ODE (a) becomes

Badri A, Moges B. and Teklebrhan B. 18 AKU

 APPLIED MATHEMATICS III

𝑦 ′ + 𝑃(𝑥)𝑦 = 0 (c)
and is called homogenous. By separating variables and integrating we then obtain
𝑑𝑦
= −𝑃(𝑥)𝑑𝑥
𝑦

Thus ,
ln|𝑦| = − ∫ 𝑃(𝑥)𝑑𝑥 + 𝑐 ∗
Taking exponents on both sides, we obtain the general solution of the homogenous ODE (c)
∗
𝑦(𝑥) = 𝑐𝑒 − ∫ 𝑃(𝑥)𝑑𝑥 where, 𝑐 = ±𝑒 𝑐

Rule for Solving Linear First Order Equation

Step1. If the equation is not in standard form then write in the standard form. i.e.
𝑦 ′ + 𝑃(𝑥)𝑦 = 𝑟(𝑥)
Step 2. Find the integrating factor,
𝐼(𝑥) = 𝑒 ∫ 𝑃(𝑥)𝑑𝑥
Step 3. Multiply the both sides of the standard form equation by the integrating factor. The left-
hand side of the resulting equation is automatically the derivative of the product of the
integrating factor 𝑒 ∫ 𝑝(𝑥)𝑑𝑥 and y:
𝑑 ∫ 𝑝(𝑥)𝑑𝑥
[𝑒 . 𝑦] = 𝑒 ∫ 𝑝(𝑥)𝑑𝑥 𝑟(𝑥)
𝑑𝑥
Step 4. Integrate both sides of the last equation and solve for y.
Step 5. If an initial condition 𝑦(𝑥0 ) = 𝑦0 is given the required solution of the initial value
problem is obtained by choosing the arbitrary constant 𝑐 in the general solution found in step 4
so that
𝑦 = 𝑦0 when, 𝑥 = 𝑥0
𝑑𝑦
Example 1.17: Solve − 3𝑦 = 0
𝑑𝑥

Solution: This linear equation can be solved by separation of variables. Alternatively, since the
differential equation is already in standard form , we identify 𝑃(𝑥) = −3, and so the integrating
factor is 𝑒 ∫(−3)𝑑𝑥 = 𝑒 −3𝑥 . We then multiply the given equation by this factor and recognize that
𝑑𝑦
𝑒 −3𝑥 − 3𝑒 −3𝑥 𝑦 = 𝑒 −3𝑥 . 0
𝑑𝑥
is the same as

Badri A, Moges B. and Teklebrhan B. 19 AKU

 APPLIED MATHEMATICS III

𝑑 −3𝑥
[𝑒 𝑦] = 0
𝑑𝑥
Integration on the last equation
𝑑
∫ 𝑑𝑥 [𝑒 −3𝑥 𝑦]𝑑𝑥 = ∫ 0 𝑑𝑥
then yields,
𝑒 −3𝑥 𝑦 = 𝑐 or 𝑦 = 𝑐𝑒 3𝑥 , ∞ < 𝑥 < −∞.
Example 1.18: Solve 𝑥𝑦 ′ = 𝑥 4 − 4𝑦
4
Solution: We have 𝑦 ′ + 𝑥 𝑦 = 𝑥 3

so
4
𝑝(𝑥) = 𝑥 and 𝑟(𝑥) = 𝑥 3
Multiplying both sides by
𝑒 ∫ 𝑝(𝑥)𝑑𝑥 = 𝑒 4ln(𝑥) = 𝑥 4
to get
𝑥 4 𝑦 ′ + 4𝑥 3 𝑦 = 𝑥 7

Taking the anti-derivative on both sides, yields

1 8
𝑖𝑦 = 𝑥 +𝑐
8
Solving for 𝑦 gives
1 4
𝑦= 𝑥 + 𝑐. 𝑥 −4
8
1
Example 1.19: Find 𝑦 given that y ′ . cot(x) = y + 2cos(x) and y(0) = 2

Solution: We have the standard form

y ′ − tan(x) y = 2sin(x)
With 𝑝(𝑥) = − tan(𝑥) and 𝑟(𝑥) = 2sin(𝑥)
Multiply both sides by 𝑒 ∫ 𝑝(𝑥)𝑑𝑥 = 𝑒 ln(cos(𝑥)) = cos(𝑥). To get
𝑦 ′ . cos(𝑥) − 𝑦. 𝑠𝑖𝑛𝑥 = 2sin(𝑥)cos(𝑥)
−1
Taking the anti-derivative on both sides yields [𝑦. cos(𝑥)] = cos(2𝑥) + 𝑐
2
−1 −1
Plugging in yields = .1 +𝑐
2 2

Hence, 𝑐 = 0.
−cos(2𝑥)
Solving for 𝑦 gives 𝑦=
2cos(𝑥)

Example 1.20 Solve the IVP cos 𝑥 𝑦 ′ + 𝑦 = sin 𝑥 , 𝑦(0) = 2 ( Similar, do it)

Badri A, Moges B. and Teklebrhan B. 20 AKU

 APPLIED MATHEMATICS III

Exercises

Find the solution of the following:

a. 𝑦 ′ = −4𝑥
b. 𝑦 ′ − 2𝑥 = 𝑥
𝑑𝑦 2
c. − 𝑥 2 𝑦 = 𝑥𝑐𝑜𝑠 𝑥 , for 𝑥 > 0
𝑑𝑥
𝑑𝑦
d. (𝑥 + 2𝑦 3 ) 𝑑𝑥 = 𝑦

Badri A, Moges B. and Teklebrhan B. 21 AKU

 APPLIED MATHEMATICS III

Unit Summary:

 A Differential Equation (DE) is an equation involving a dependent variable and its

derivative with respect to one or more independent variable.
 A differential equation involves functions of only a single variable is called an Ordinary
Differential Equation (ODE). If a differential equation contains only ordinary derivatives of
one or more unknown functions with respect to a single independent variable, it is said to be
an ordinary differential equation (ODE).
The PDE (Partial Differential Equation) is an equation which involves partial derivatives of an
unknown function of two or more variables. That is, the derivatives of functions of more than
one variable.

 The order of the equation is the order of the highest derivatives that appears.
 The degree of a DE is the highest exponent of the highest order derivatives which involves in
the DE.
 The nth order linear differential equation is written as,
𝑎0 (𝑥)𝑦 𝑛 + 𝑎1 (𝑥)𝑦 𝑛−1 + ⋯ + 𝑎𝑛−1 (𝑥)𝑦 = 𝑓(𝑥) … (*)

where, 𝑎0 (𝑥), 𝑎1 (𝑥) … . 𝑎𝑛 (𝑥) called the coefficient of the equation .The known function
𝑓(𝑥) is called non homogeneous term , equation (*) is called homogeneous if 𝑓(𝑥) = 0

If the coefficients in (*) are constant, so that (*) becomes 𝑎0 𝑦 𝑛 + 𝑎1 𝑦 𝑛−1 + ⋯ + 𝑎𝑛−1 𝑦 = 𝑓(𝑥)

which is called constant coefficient linear equation. 𝑎0 ≠ 0 , otherwise the equation would
not be the nth order. An ordinary differential equation that is not linear is said to be non-linear.

 A function relation between two variables (dependent and independent variables) which
satisfy the given DE is called the solution of an ODE or integral curve. A solution of an nth
order equation that contains n arbitrary constants is called the general solution of the
equation. If the arbitrary constants in the general solution are assigned specific value, the
result is called a particular solution of the equation.
 If the differential equation has the form 𝑦 ′ = 𝑓(𝑥) , then 𝑦 = ∫ 𝑓(𝑥)𝑑𝑥 + 𝑐 which is a
general solution.
 An equation 𝑦 ′ = 𝑓(𝑥, 𝑦) is called separable if it can be written in the form

Badri A, Moges B. and Teklebrhan B. 22 AKU

 APPLIED MATHEMATICS III

𝑦 ′ = 𝐹(𝑥)𝐺(𝑦) (*)

For some function 𝐹(𝑥) a dependent only on 𝑥 and 𝐺(𝑦) dependent on 𝑦. Equation (*) itself to be
variable separable type.
 A function 𝑓(𝑥, 𝑦) is said to be algebraically homogenous of degree 𝑛, or simply
homogenous of degree 𝑛,
if 𝑓(𝑘𝑥, 𝑘𝑦) = 𝑘 𝑛 𝑓(𝑥, 𝑦),
for some real number 𝑛 and all 𝑘 > 0 for 𝑓(𝑥, 𝑦) ≠ (0,0).
 The first ordered differential equation 𝑀(𝑥, 𝑦)𝑑𝑥 + 𝑁(𝑥, 𝑦)𝑑𝑦 = 0 is side to be exact if a
function 𝐹(𝑥, 𝑦) exists such that the total differential 𝑑[𝐹(𝑥, 𝑦)] = 𝑀(𝑥, 𝑦)𝑑𝑥 + 𝑁(𝑥, 𝑦)𝑑𝑦
𝜕𝑀 𝜕𝑁
 The differential equation 𝑀(𝑥, 𝑦)𝑑𝑥 + 𝑁(𝑥, 𝑦)𝑑𝑦 = 0 is exact if and only if =
𝜕𝑦 𝜕𝑥

 An integrating factor is a factor which we multiply the given non exact equation to make it
exact. To find an integrating factor 𝐼(𝑥, 𝑦) , solve the partial differential equation
𝐼𝑦 𝑀 − 𝐼𝑥 𝑁 + 𝐼(𝑀𝑦 − 𝑁𝑥 ) = 0
This is just as tricky to solve as the original equation. Only in a few special cases are there
methods for computing the integrating factor 𝐼(𝑥, 𝑦).
Case1: Suppose we want to see if there exists an integrating factor that depends only in x (and
𝜕𝐼
not on y). Then, 𝜕𝑦 would be zero, since 𝐼 does not depend on y, and so 𝐼(𝑥) we need to satisfy
𝐼′ 𝑀𝑦 −𝑁𝑥 𝑀𝑦 −𝑁𝑥
= .This can only happen if the ratio is a function of 𝑃(𝑥) only of 𝑥 (and not𝑦),
𝐼 𝑁 𝑁

then 𝐼(𝑥) = 𝑒 ∫ 𝑃(𝑥)𝑑𝑥 .

Case2: We see also look to see if there is an integrating factor that depends only on 𝑦 and not
𝜕𝐼
on 𝑥 .We can do the same calculation ,this time using would be zero, to see that such an
𝜕𝑦
𝑁𝑥 −𝑀𝑦
integrating factor exists if the ratio is a function of 𝑄(𝑦) only of 𝑦 (and not x); then
𝑀

𝐼(𝑦) = 𝑒 ∫ 𝑄(𝑦)𝑑𝑦
 A first order ODE is said to be linear if it can be written as:𝑦 ′ + 𝑃(𝑥)𝑦 = 𝑟(𝑥)
(Standard form ) (*)

where 𝑝 and 𝑟 are a function of 𝑥. If the first term is 𝑓(𝑥)𝑦’ (instead of 𝑦’), divided the equation
by 𝑓(𝑥) to get the “ standard form “ (*) with 𝑦’ as the first term.

Badri A, Moges B. and Teklebrhan B. 23 AKU

 APPLIED MATHEMATICS III

Miscellaneous Exercises

1. Explain the terms ordinary differential equation (ODE), partial differential

equation(PDE),order, general solution and particular solution. Give examples, why are the
concepts of importance?
2. What is an initial condition? How is this condition used in an initial value problem?
3. What is a homogeneous linear ODE? A nonhomogeneous linear ODE? Why are these
equations simpler than nonlinear ODEs?
4. Make a list of the solution methods considered. Explain each method with a few short
sentences and illustrate it by a typical example.
5. Can certain ODEs be solved by more than one method? Give three examples.
6. Does every first order ODE have a solution? What do you know about uniqueness of
solutions?
7. Without solving explain why the initial value problem
dy
= √y , y(x0 ) = y0 has no solution for y0 = 0.
dx

I. Find the general solution; indicate which method in this chapter you are using. Show the
details of your work.
dy −4y2 +6xy
a. (y 2 + 1)dx = ysec 2 xdy i. =
dx 3y2 +2x
dQ
b. y(lnx − lny)dx = (xlnx − xlny − y)dy j. t dt + Q = t 4 lnt
dy
c. (6x + 1)y 2 dx + 3x 2 + 2y 3 = 0 k. yy ′ + xy 2 =
y
d. xy ′ = (x)3 + y l. (x 2 + 4)dy = (2x − 8xy)dx

e. 3sin2ydx + 2xcos2ydy = 0 m. (2x + y + 1)y ′ = 1

y
f. xy ′ = xtan (x) + y n. y ′ = x(y − x 2 + 1)

g. xy ′ = (y − 2x)2 + y , (Hint , set y − 2x = z). o. yy ′ + xy 2 = x

h. −πsinπxcosh3ydx + 3cosπxsinh3ydy = 0

Badri A, Moges B. and Teklebrhan B. 24 AKU

 APPLIED MATHEMATICS III

II. Solve the following initial value problem. Indicate the method used. Show the details of
your work.
a. yy ′ + x = 0 , y(3) = 4
dy 7π
b. sinx dx + (cosx)y = 0, y ( 6 ) = −2
dy −1
c. + 2(x + 1)y 2 = 0, y(0) =
dx 8

d. (2xy 2 − sinx)dx + (2 + 2x 2 y)dy = 0 , y(0) = 1

y2 1
e. [2y + + ex (1 + x)] dx + (x + 2y)dy = 0 , y(1) = 1
x

f. y ′ + πy = 2bcosπx , y(0) = 0
g. y ′ − 3y = −12y 2 , y(0) = 0

References

1. Erwin Kreyszing,Advanced engineering mathematics, 2006,John Wiley and Sons,Inc.

2. Dennis Gzill, A first course in differential equations with modelin equations,
2011Brooks/Cole Cengage Learning

3.PeterPhili,Methods for Solving Ordinary Differential Equations,2011, Spring

Badri A, Moges B. and Teklebrhan B. 25 AKU

 APPLIED MATHEMATICS III

4. HeneryRecardo, A Modern to Introduction to Differential equation,2009,Elsevier Inc.

5. A.C. King, J. Billingnam, Differential Equations ,linear, nonlinear, Ordinary, Partial,2003,

CAMBRIGE .
6. Matthew R and Etla, differential equation with linear algebra,2009, Oxford University press
7. K. Soetaert, T P, Solving Initial Value Diferential Equations in R,2010,R package deS

8.Mircea V. Soare , Ordinary differential equations with applications to Mechanics, 2007

Springer.

9. Larson Edwards, Calculus, 2010, Brooks/Cole CengageLearning

10. A. Ganesh and Etla, Engineering Mathematics II, 2009 New age International press

11. Wilfred Kaplan,Advanced Calculus, 5th edition, publishing house of electronics industry

12. Salas Hille Etgen, Calculus – One and Several variables,10th edition, WILLEY PLUS
13. Boyce. Diprima, Elementary differential equations and boundary value problem, 2001 ,John
Wiley and Sons.Inc
14. Ravi P. Agarwal. An introduction to differential equation, 2000,Spring
15. Rudolph E. Longer,Ordinary Differential equations,1954,John Wiley and Sons.Inc

Chapter Two
Ordinary Linear Differential Equation of The 2nd order
Introduction

In chapter one we saw that we could solve a few first-order differential equations by recognizing
them as separable, linear, exact or homogeneous equations. We turn now to the solution of
ordinary differential equation of order two or higher. These equations have important

Badri A, Moges B. and Teklebrhan B. 26 AKU

 APPLIED MATHEMATICS III

engineering application, especially with connection mechanical and electrical engineering

vibrations.
Linear second order differential equations with constant coefficients are the simplest of the
higher order differential equation and they have many applications.
The most general linear second order differential equation is in the form:
𝑝(𝑥)𝑦 ′′ + 𝑞(𝑥)𝑦 ′ + 𝑟(𝑥)𝑦 = 𝑔(𝑥)
The equation is called non-homogenous when 𝑔(𝑥)zero is not identically; otherwise it is called
homogenous.
Example 2.1: 𝑦 ′′ + 𝑥𝑦 ′ + 𝑦 = 0 is a homogenous 2nd order linear differential equation
𝑦 ′′ = sin 𝑥 is a non-homogenous 2nd order linear differential equation.

2.1. Homogenous Linear Equation of The Second Order

A second order ODE is called linear if it can be written as
𝑦 ′′ + 𝑝(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 𝑟(𝑥) (1)
where, p and q are called the coefficient of the ODEs and is a function of x and nonlinear if it
cannot be written in this form.
Example 2.2. 𝑦 ′′ + 25𝑦 = cos 𝑥 is a linear 2nd order ODE
𝑦 ′′ + 𝑦𝑦 ′ + 2𝑦 = 𝑒 𝑥 is nonlinear 2nd order ODE
In equation (1) if 𝑟(𝑥) = 0, the equation reduce to 𝑦 ′′ + 𝑝(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0
and is called homogenous. If 𝑟(𝑥) ≠ 0 then equation (1) is called non homogenous.
Example 2.3. 𝑦 ′′ + 25𝑦 = 𝑒 −𝑥 cos 𝑥 Non homogenous
𝑥𝑦 ′′ + 𝑦 ′ + 𝑥𝑦 = 0 Homogenous

Definition 2.1. A function 𝑦 = ℎ(𝑥) is called a solution of a (linear or non linear) second order ODE on
some open interval 𝐼 if ℎ is defined and twice differentiable throughout the interval and is such that the
ODE becomes an identify if we replace the unknown 𝑦 by ℎ, and its successive derivatives.

Theorem: (Superposition or Linear principle)

If 𝑦1 and 𝑦2 are solution of the differential equation
𝑦 ′′ + 𝑝(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0

Badri A, Moges B. and Teklebrhan B. 27 AKU

 APPLIED MATHEMATICS III

then 𝑐1 𝑦1, 𝑐2 𝑦2 and 𝑐1 𝑦1 + 𝑐2 𝑦2 are solutions of the given differential equation.

Proof: Let 𝑦1 and 𝑦2 be solution of the differential equation on I. Then by substituting
𝑦 = 𝑐1 𝑦1 + 𝑐2 𝑦2
and its derivatives into the differential equation, and using the familiar rule
( 𝑐1 𝑦1 + 𝑐2 𝑦2 )′ = 𝑐1 𝑦1 ′ + 𝑐2 𝑦2 ′ .
We get
𝑦 ′′ + 𝑝𝑦 ′ + 𝑞𝑦 = (𝑐1 𝑦1 + 𝑐2 𝑦2 )′′ + (𝑐1 𝑦1 + 𝑐2 𝑦2 )′ + 𝑞(𝑐1 𝑦1 + 𝑐2 𝑦2 )
= 𝑐1 𝑦1 ′′ + 𝑐2 𝑦2 ′′ + 𝑝(𝑐1 𝑦1 ′ + 𝑐2 𝑦2 ′ ) + 𝑞(𝑐1 𝑦1 + 𝑐2 𝑦2 )
= 𝑐1 (𝑦1 ′′ + 𝑝𝑦1 ′ + 𝑞𝑦1 ) + 𝑐2 (𝑦2 ′′ + 𝑝𝑦2 ′ + 𝑞𝑦2 ) = 0
Since in the last line (… ) = 0, because 𝑦1 and 𝑦2 are solution of differential equation on I.
Note: Superposition principle holds for homogenous linear ODEs only but doesn’t hold for non-
homogenous linear or non linear.
Example 2.4. Show that 𝑦 = cos 𝑥 𝑎𝑛𝑑 𝑦 = 𝑠𝑖𝑛 𝑥 are solution of
𝑦 ′′ + 𝑦 = 0
for all 𝑥 and show that any linear combination of the two functions is a solution of the given
ODE.
Example 2.5. Verify that

a. 𝑦1 = 1 and 𝑦2 = 𝑥 2 are solution of 𝑦𝑥 3 − 𝑦𝑦 ′ = 0

b. 𝑦 = 4𝑒 3𝑥 − 2 and 𝑦 = 𝑒 3𝑥 − 2 are solution of 𝑦 ′′ − 9𝑦 = 18 ,but their linear
combination are not.
This will be seen later in the next topics.

2.2 Initial Value Problem (IVP), Basis, General solution.

The initial condition is used to determine the arbitrary constant in the general solution of the
ODE.
This results is a unique solution and is called a particular solution.
For a second order homogenous linear ODE
𝑦 ′′ + 𝑝𝑦 + 𝑞𝑦 = 0 (1)

Badri A, Moges B. and Teklebrhan B. 28 AKU

 APPLIED MATHEMATICS III

an initial value problem(IVP) consists of (1) and two initial conditions

𝑦(𝑥0 ) = 𝑦0 , 𝑦(𝑥0 ) = 𝑦1

these condition helps to determine the constants 𝑐1 and 𝑐2 in the general solution.
Generally: we defined an initial-value problem for a general nth-order differential equation. For
a linear differential equation an
Example 2.6: Solve the following IVP
𝑦 ′′ − 9𝑦 = 0, 𝑦(0) = 2, 𝑦 ′ (0) = −1
Solution: The two function 𝑦(𝑡) = 𝑒 3𝑡 and 𝑦(𝑡) = 𝑒 −3𝑡 are enough to form the general
solution to the differential equation. The general solution to our differential equation is then
𝑦(𝑡) = 𝑐1 𝑒 −3𝑡 + 𝑐2 𝑒 3𝑡
Now all we need to do is apply the initial conditions
𝑦 ′ (𝑡) = −3𝑐1 𝑒 −3𝑡 + 3𝑐2 𝑒 3𝑡
Plugging in the initial conditions
𝑦(0) = 2 = 𝑐1 + 𝑐2
𝑦 ′ (0) = −1 = −3𝑐1 + 3𝑐2
This gives us a system of two equations and two unknowns that can be solved. Doing this yields
7 5
𝑐1 = , 𝑐2 =
6 6
The solution to the IVP is then
7 −3𝑡 5 3𝑡
𝑦(𝑡) = 𝑒 + 𝑒
6 6

We will see it more examples in detail next

Exercises

Find a general solution of the IVP

𝜋 𝜋
a. 𝑦 ′′ + 4𝑦 = 0 𝑦 ( 4 ) = 1 , 𝑦 ( 4 ) = 1

b. 𝑦 ′′ + 9𝑦 = 0 𝑦(0) = 2 , 𝑦(0) = −1

Badri A, Moges B. and Teklebrhan B. 29 AKU

 APPLIED MATHEMATICS III

Linear Dependent and Independent

Two function 𝑦1 (𝑥) and 𝑦2 (𝑥) are said to be linearly independent (LI) over an interval I if the
equation
𝑐1 𝑦1 (𝑥) + 𝑐2 𝑦2 (𝑥) = 0 (*)
is only true for all x in the interval if 𝑐1 = 𝑐2 = 0. The functions are said to be linearly dependent
(LD) if (*) is true for some non vanishing constants 𝑐1 and 𝑐2 .
When the functions are linearly dependent, provided 𝑐1 ≠ 0 equation (*) can be written:
𝑐2
𝑦1 = − 𝑦2 (𝑥)
𝑐1
with a corresponding result
𝑐1
𝑦2 = − 𝑦 (𝑥)
𝑐2 1
Showing that in each case the linear dependent of the function means they are proportional.
Definition: Two functions 𝑦1 (𝑥) and 𝑦2 (𝑥) will be linearly independent (LI) over an interval I
if they are not proportional over the interval: otherwise will be (linear dependent) LD.
Example 2.7:
𝒚𝟏 (𝒙) and 𝒚𝟐 (𝒙) Type Reason
1. ln 𝑥 and ln 𝑥 2 LD ln 𝑥 2 = 2 ln 𝑥
𝑒 2𝑥
2. 𝑒 𝑥 and 𝑒 2𝑥 LI = 𝑒 𝑥 is not constant
𝑒𝑥

3. sin ℎ𝑥 and sin ℎ𝑥 cos ℎ𝑥 LD sin ℎ𝑥 = 2 sin ℎ𝑥 cos ℎ𝑥

𝑥
4. 𝑥 and 𝑒 𝑥 LI = 𝑥𝑒 𝑥 is not constant
𝑒𝑥

Note: The number of linearly independent solution is equal to the order of the differential
equation , so the 2nd order differential equation has two linearly independent solution.

Definition: The Wronskian of 𝑛 function 𝑦1 (𝑥) ,𝑦2 (𝑥), 𝑦3 (𝑥),… 𝑦𝑛 (𝑥) each (n-1) time
differentiable is the form
𝑦1 𝑦2 ⋯⋯ 𝑦𝑛
𝑦1 ′ 𝑦2 ′ ⋯⋯ 𝑦𝑛 ′
|
𝑊(𝑦) = | 𝑦1 ′′ 𝑦2 ′′ ⋯⋯ 𝑦𝑛 ′′ | , where | | denotes the determinant.
⋯⋯ |
⋮ ⋮ ⋮
𝑦1 (𝑛−1)
𝑦2 (𝑛−1) ⋯⋯ 𝑦𝑛 (𝑛−1)

Badri A, Moges B. and Teklebrhan B. 30 AKU

 APPLIED MATHEMATICS III

Example 7. Consider the functions 𝑦1 (𝑥) = 𝑒 𝑥 , 𝑦2 (𝑥) = 𝑒 2𝑥 . Then

𝑦1 𝑦2
𝑊(𝑒 𝑥 , 𝑒 2𝑥 ) = |𝑦 ′ 𝑦 ′ |
1 2
𝑥
= |𝑒 𝑥 𝑒 2𝑥 |
𝑒 2𝑒 2𝑥
= 𝑒 𝑥 2𝑒 2𝑥 − 𝑒 𝑥 𝑒 2𝑥 = 𝑒 3𝑥
Theorem: If 𝑊(𝑦1 , 𝑦2 , … 𝑦𝑛 )(𝑥0 ) ≠ 0 for some 𝑥0 ∈ 𝐼 then the set {𝑦1 , 𝑦2 , … , 𝑦𝑛 } is LI on 𝐼.
Theorem: Given
𝑦 ′′ + 𝑝𝑦 ′ + 𝑞𝑦 = 0 (*)
If 𝑦1 and 𝑦2 are any two solution of (*) on I, then 𝑊(𝑦) ≠ 0(𝑖, 𝑒 𝑊(𝑦1 , 𝑦2 ) ≠ 0)
Corollary: Any two solutions 𝑦1 and 𝑦2 of the differential equation 𝑦 ′′ + 𝑝𝑦 ′ + 𝑞𝑦 = 0 are LI.
(Proofs of the above Theorems and corollary are left as exercise for you)

Definition: a general solution of an ODE

𝑦 ′′ + 𝑝(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0 (*)
On an open interval I is a solution of
𝑦 = 𝑐1 𝑦1 + 𝑐2 𝑦2 (**)
In which 𝑦1 and 𝑦2 are solution of (*) on 𝐼 that are not proportional (LI) and 𝑐1 and 𝑐2 are
arbitrary constants. These 𝑦1 and 𝑦2 are called a basis (or a fundamental system of the solution
(*) on 𝐼)

A particular solution of (*) on I is obtained if we assign specific values to 𝑐1 and 𝑐2 in (**).

′′
Example7.Acos
Definition: 𝑥 and
basis sin 𝑥 in example
of solution (*) on an3open
forminterval
a basis 𝐼ofissolution
a pair ofofLIthe ODE 𝑦of
solution + 𝑦on
(*) = 𝐼.0 for all

Example 2.8. Show that 𝑦1 = 𝑒 𝑥 and 𝑦2 = 𝑒 −𝑥 are solution of ODE 𝑦 ′′ + 𝑦 = 0 then solve the
initial value problem .
Find a Basis If One Solution is Known, Reduction of Order
Finding solutions to non- constant coefficients, second order differential equations can be much
more differential equations can be much more difficult than finding solutions to constant
coefficient differential equations. However if we already know one solution to the differential
equation we can use the method that we used in the first order differential equation. This method
is called reduction of order.

Badri A, Moges B. and Teklebrhan B. 31 AKU

 APPLIED MATHEMATICS III

Given the homogenous linear ODE (in the standard form)

𝑦 ′′ + 𝑝(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0
If 𝑦1 is its solution. Then the second linearly independent solution is given by
𝑒 − ∫ 𝑝𝑑𝑥
𝑦2 = 𝑦1 ∫ 𝑑𝑥
𝑦1 2
Method of Finding 𝒚𝟐
We substitute
𝑦 = 𝑦2 = 𝑢𝑦1 , 𝑦 ′ = 𝑦2 ′ = 𝑢′ 𝑦1 + 𝑢𝑦1 ′ , 𝑦 ′′ = 𝑢′′ 𝑦1 + 2𝑢′ 𝑦1 ′ + 𝑢𝑦1 ′′
into a homogenous linear equation
𝑦 ′′ + 𝑝(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0
This gives
𝑢′′ 𝑦1 + 2𝑢′ 𝑦1 ′ + 𝑢𝑦1 ′′ + 𝑝(𝑢′ 𝑦1 + 𝑢𝑦1 ′ ) + 𝑞𝑢𝑦1 = 0
Collecting terms in 𝑢′′ , 𝑢′ and 𝑢, we have
𝑢′′ 𝑦1 + 𝑢′ (2𝑦1 ′ + 𝑝𝑦1 ) + 𝑢(𝑦1 ′′ + 𝑝𝑦1 ′ + 𝑞𝑦1 ) = 0
Now comes the point. Since 𝑦1 is a solution of
𝑦 ′′ + 𝑝(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0
i.e. the expression
𝑦1 ′′ + 𝑝𝑦 ′ + 𝑞𝑦1 = 0.
Hence , 𝑢 is gone, and we are left with an ODE by 𝑦1 and set
𝑢′ = 𝑈, 𝑢′′ = 𝑈
2𝑦1 ′ +𝑝𝑦1
𝑢′′ + 𝑢′ = 0.
𝑦1

Thus
2𝑦1 ′
𝑈′ + ( + 𝑝)𝑈 = 0
𝑦1

This is the desired first order ODE, the reduced ordinary differential equation. Separation of
variables and integrations gives
𝑑𝑈 2𝑦1 ′
= −( + 𝑝)dx ,and
𝑈 𝑦1

𝑙𝑛|𝑈| == 2𝑙𝑛|𝑦1 | − ∫ 𝑝𝑑𝑥 .

By taking exponents we finally obtain

Badri A, Moges B. and Teklebrhan B. 32 AKU

 APPLIED MATHEMATICS III

1 − ∫ 𝑝𝑑𝑥
𝑈= 𝑒
𝑦1 2
Here, 𝑈 = 𝑢′ so that 𝑢 = ∫ 𝑈𝑑𝑥
Hence, the desired solution is
1 − ∫ 𝑝𝑑𝑥
𝑦2 = 𝑦1 𝑢 = 𝑦1 ∫ 𝑈𝑑𝑥 = 𝑦1 ∫ 𝑒 𝑑𝑥
𝑦1 2
Example 2.9: Find a basis of solutions of the ODE
(𝑥 2 − 𝑥)𝑦 ′′ − 𝑥𝑦 ′ + 𝑦 = 0
Given that 𝑦1 (𝑥) = 𝑥 is a solution.
Solution: Given 𝑦1 = 𝑥 is a solution, (inspection).To find the second solution ,substitute
𝑦2 = 𝑦 = 𝑢𝑦1 = 𝑢𝑥
𝑦 ′ = 𝑢′ 𝑥 + 𝑢 , 𝑦 ′′ = 𝑢′′ 𝑥 + 𝑢′ 𝑥 ′ + 𝑢′ = 𝑢′′ 𝑥 + 2𝑢′
into the ODE. This gives
(𝑥 2 − 𝑥)(𝑢′′ 𝑥 + 2𝑢′ ) − 𝑥(𝑢′ 𝑥 + 𝑢) + 𝑢𝑥 = 0
𝑢𝑥 and −𝑢𝑥 cancel and we are left with the following ODE, which we divide by x, order and
simplify.

(𝑥 2 − 𝑥)(𝑢′′ 𝑥 + 2𝑢′ ) − 𝑥 2 𝑢′ = 0
After simplification, we get
(𝑥 2 − 𝑥)𝑢′′ + (𝑥 − 2)𝑢′ = 0
This ODE is of first order in 𝑈 = 𝑢′ , namely (𝑥 2 − 𝑥)𝑈 ′ + (𝑥 − 2)𝑈 = 0
Separation of variables and integration gives
dU −(x−2) 1 2
= dx= (x−1 − x)dx
U x2 −x

ln|U| = ln|x − 1| + 2ln|x|

|𝑥 − 1|
= 𝑙𝑛
𝑥2
Taking exponents and integrating again, we obtain
𝑥−1 1 1
𝑈= = − 2
𝑥2 𝑥 𝑥
1
𝑢 = ∫ 𝑈𝑑𝑥 = 𝑙𝑛|𝑥| +
𝑥
Hence, 𝑦2 = 𝑢𝑥 = 𝑥𝑙𝑛|𝑥| + 1.

Badri A, Moges B. and Teklebrhan B. 33 AKU

 APPLIED MATHEMATICS III

Exercises

I. Solve the given differential equations, whose one solution is known

a. 𝑦 ′′ − 5𝑥𝑦 ′ + 9𝑦 = 0 , 𝑦1 = 𝑥 3
b. 𝑦 ′′ = 𝑦 ′ , 𝑦1 = 𝑥 3
c. 𝑥𝑦 ′′ + 𝑦 ′ = 0 , 𝑦1 = 𝑙𝑛 𝑥
d. 2𝑥 2 𝑦 ′′ + 𝑥𝑦 ′ − 3𝑦 = 0 , 𝑦1 = 𝑥 2
e. 𝑥 2 𝑦 ′′ − 𝑥𝑦 ′ + 𝑦 = 0 , 𝑦1 = 𝑥
II. Verify by substitution that the given function a basis , solve the given initial value
problem
a. 𝑦 ′′ + 25 𝑦 = 0 , 𝑐𝑜𝑠 5𝑥, 𝑠𝑖𝑛 5𝑥 𝑦(0) = 0.8 𝑦 ′ (0) = −6.5
b. 𝑥 2 𝑦 ′′ − 7𝑥 𝑦 ′ + 15 𝑦 = 0 , 𝑥3 , 𝑥5 𝑦(1) = 0.4 𝑦 ′ (1) = 1

2.3. Homogenous linear ODE with constant coefficients

Let us consider second – order homogenous linear ODEs whose coefficients 𝑎 and 𝑏
are constant.
General form

𝑦 ′′ + 𝑎𝑦 ′ + 𝑏𝑦 = 0 (1)

where a and b are constant.

Definition 2.3 The equation

𝜆2 + 𝑎𝜆 + 𝑏 = 0 (2)

is called the characteristic equation of the given differential equation.

The roots of the characteristic equation are :

1
𝜆1 = (−𝑎 + √𝑎2 − 4𝑏)
2

1
𝜆2 = (−𝑎 − √𝑎2 − 4𝑏)
2

Badri A, Moges B. and Teklebrhan B. 34 AKU

 APPLIED MATHEMATICS III

Depending on the sign of the discriminant 𝑎2 − 4𝑏 the quadratic equation (2) may have three
kinds of roots.

Case I: Two real roots, if 𝑎2 − 4𝑏 > 0

Case II: A real double root , if 𝑎2 − 4𝑏 = 0

Case III: Complex conjugate roots, if 𝑎2 − 4𝑏 < 0

Let us consider the conditions:

Case I: Two distinct real roots 𝜆1 and 𝜆2 .

In this case, a basis of solutions of (1) on any interval is

𝑦1 = 𝑒 𝜆1 𝑥 and 𝑦2 = 𝑒 𝜆2 𝑥

Because, 𝑦1 and 𝑦2 are defined (and real) for all x and their quotient is not constant. The
corresponding general solution is

𝑦 = 𝑐1 𝑒 𝜆1 𝑥 + 𝑐2 𝑒 𝝀𝟐 𝑥

Example 2.10: Solve the following IVP

a. 𝑦 ′′ + 11𝑦 ′ +24𝑦 = 0 , 𝑦(0) = 0 , 𝑦 ′ (0) = −7

Solution: The characteristic equation is

𝜆2 + 11𝜆 + 24 = 0

(𝜆 + 8)(𝜆 + 3) = 0.

Its roots are 𝜆1 = −8 and 𝜆2 = −3 , and the general solution and its derivative is

𝑦(𝑥) = 𝑐1 𝑒 −8𝑥 + 𝑐2 𝑒 −𝟑𝑥

𝑦 ′ (𝑥) = −8𝑐1 𝑒 −8𝑥 − 3𝑐2 𝑒 −𝟑𝑥

Now, plug in the initial conditions to get the following system of equations

𝑦(0) = 0 = 𝑐1 + 𝑐2
Badri A, Moges B. and Teklebrhan B. 35 AKU
 APPLIED MATHEMATICS III

𝑦 ′ (0) = −7 = −8𝑐1 − 3𝑐2

Solving this system gives

7 −7
𝑐1 = 5 and 𝑐2 = 5

The actual solution to the differential equation is , then

7 −8𝑥 7 −𝟑𝑥
𝑦(𝑥) = 𝑒 − 𝑒
5 5

b. y ′′ + 3y ′ − 10y = 0, y(0) = 4, y ′ (0) = −2

Solution: The characteristic equation is

𝜆2 + 3𝜆 − 10 = 0

(𝜆 + 5)(𝜆 − 2) = 0

The roots are, 𝜆1 = −5 and 𝜆2 = 2

, and so the general solution and its derivative is

𝑦(𝑥) = 𝑐1 𝑒 −5𝑥 + 𝑐2 𝑒 𝟐𝑥

𝑦 ′ (𝑥) = −5𝑐1 𝑒 −5𝑥 + 2𝑐2 𝑒 𝟐𝑥

Now, plug in the initial conditions to get the following system of equations.

𝑦(0) = 4 = 𝑐1 + 𝑐2

𝑦 ′ (0) = −2 = −5𝑐1 + 2𝑐2

Solving this system gives

10 18
𝑐1 = and 𝑐2 =
7 7

The actual (particular) solution to the differential equation is then

10 −5𝑥 18 𝟐𝑥
𝑦(𝑥) = 𝑒 + 𝑒
7 7

Badri A, Moges B. and Teklebrhan B. 36 AKU

 APPLIED MATHEMATICS III

𝒂
Case II: Real double root 𝝀 = − 𝟐

If the discriminant 𝑎2 − 4𝑏 = 0, the characteristic equation has only one root 𝜆 = 𝜆1 = 𝜆2 =

𝑎
𝒂
− 𝟐. Hence one solution is 𝑦1 = 𝑒 (−2)𝑥 . To obtain a second independent solution 𝑦2 (needed for a

basis) we use the method of reduction of order is

𝑒 − ∫ 𝑝(𝑥)𝑑𝑥
𝑦2 = 𝑦1 ∫ , but 𝑝(𝑥) is a constant .
𝑦1 2

Then it become,

𝑒 − ∫ 𝑎𝑑𝑥 𝑎
(− )𝑥 𝑒 −𝑎𝑥 𝑎
(− )𝑥
𝑦2 = 𝑦1 ∫ 𝑑𝑥 = 𝑒 2 ∫ 𝑑𝑥 = 𝑥𝑒 2
𝑦1 2 𝑒 −𝑎𝑥

𝑎
Hence in the case of a double root of (2), a basis of solution of (1) on any interval is: 𝑒 (−2)𝑥 ,
𝑎
𝑒 (−2)𝑥 .

The corresponding general solution is:

𝑎
𝑦 = 𝑒 (−2)𝑥 (𝑐1 + 𝑐2 𝑥)

i.e. if the roots of the characteristic equation are 𝜆 = 𝜆1 = 𝜆2 , then the general solution is then

𝑦(𝑥) = 𝑐1 𝑒 𝜆𝑥 + 𝑐2 𝑥𝑒 𝜆𝑥

Example 2.11: Solve the following IVP.

′
a. 𝑦 ′ − 4𝑦 ′ + 4𝑦 = 0 , 𝑦(0) = 12, 𝑦 ′ (0) = −3
Solution: The characteristic equation and its roots are

𝜆2 − 4𝜆 + 4 = (𝜆 − 2)2 = 0

i.e.

𝜆1 = 𝜆2 = 2

The general solution and its derivative are

Badri A, Moges B. and Teklebrhan B. 37 AKU

 APPLIED MATHEMATICS III

𝑦(𝑥) = 𝑐1 𝑒 2𝑥 + 𝑐2 𝑥𝑒 𝟐𝑥

𝑦 ′ (𝑥) = 2𝑐1 𝑒 2𝑥 + 𝑐2 𝑒 2𝑥 + 2𝑐2 𝑥𝑒 𝟐𝑥

Plugging in the initial conditions gives the following system

12 = 𝑦(0) = 𝑐1

−3 = 𝑦 ′ (0) = 2𝑐1 + 𝑐2

This system is easily solve to get 𝑐1 = 12 and 𝑐2 = −27.

The actual solution (particular solution ) to the IVP is , then

𝑦(𝑥) = 12𝑒 2𝑥 − 27𝑥𝑒 𝟐𝑥

′ −9
b. 16𝑦 ′ − 40𝑦 ′ + 25𝑦 = 0 , 𝑦(0) = 3, 𝑦 ′ (0) = 4

Solution: The characteristic equation and its roots are:

5
16𝜆2 − 40𝜆 + 25 = (4𝜆 − 5)2 = 0 , 𝜆 1,2 =
4

The general solution and its derivative are

5𝑥 𝟓𝒙
𝑦(𝑥) = 𝑐1 𝑒 4 + 𝑐2 𝑥𝑒 𝟒

5 5𝑥 𝟓𝒙 5 5𝑥
𝑦 ′ (𝑥) = 𝑐1 𝑒 4 + 𝑐2 𝑒 𝟒 + 𝑐2 𝑥𝑒 4
4 4

Plugging in the initial conditions gives the following system:

3 = 𝑦(0) = 𝑐1

−9 5
= 𝑦 ′ (0) = 𝑐1 + 𝑐2
4 4

Solving this system gives, 𝑐1 = 3 and 𝑐2 = −6

The particular solution to the IVP is then

Badri A, Moges B. and Teklebrhan B. 38 AKU

 APPLIED MATHEMATICS III

5𝑥 𝟓𝒙
𝑦(𝑥) = 3𝑒 4 − 6𝑥𝑒 𝟒

𝟏 𝟏
Case III: Complex roots: − 𝟐 𝒂 + 𝒊𝝎 and − 𝟐 𝒂 − 𝒊𝝎.

This case occurs if the discriminant 𝑎2 − 4𝑏 < 0 of the characteristic equation is negative,
then the characteristic equation has no real root but it has a complex root

That is,

1
𝜆1 , 𝜆2 = (−𝑎 ± √𝑎2 − 4𝑏)
2

1
= (−𝑎 ± 𝑖 √4𝑏 − 𝑎2 )
2

𝑎 1
Put, 𝛼 = − 2 and 𝛽 = 2 √4𝑏 − 𝑎2

We get the complex roots of the characteristic equation as

𝜆1 , 𝜆2 = 𝛼 ± 𝑖𝛽 with 𝛼, 𝛽 ∈ 𝑅

Thus ,

𝑦1 = 𝑒 𝜆1 𝑥 = 𝑒 (𝛼+𝑖𝛽)𝑥

and

𝑦2 = 𝑒 𝜆2 𝑥 = 𝑒 (𝛼−𝑖𝛽)𝑥

are the basis of complex solution of the ODE and the general solution is

𝑦 = 𝑐1 𝑒 (𝛼+𝑖𝛽)𝑥 + 𝑐2 𝑒 (𝛼−𝑖𝛽)𝑥 ,with 𝑐1 , 𝑐2 constants.

Remark: (Euler’s formula)

𝑒 𝑖𝑥 = cos 𝑥 + 𝑖 sin 𝑥 ⇒ 𝑒 𝑖𝛽 = cos 𝛽𝑥 + 𝑖 sin 𝛽𝑥

𝑒 (𝛼±𝑖𝛽)𝑥 = 𝑒 𝛼𝑥±𝑖𝛽𝑥 = 𝑒 𝛼𝑥 (𝑒 ±𝑖𝛽𝑥 ) = 𝑒 𝛼𝑥 (cos 𝛽𝑥 ± 𝑖 sin 𝛽𝑥)

So the complex solution of (1) can be written as

Badri A, Moges B. and Teklebrhan B. 39 AKU

 APPLIED MATHEMATICS III

𝑦 = 𝑐1 𝑒 𝛼𝑥 (cos 𝛽𝑥 + 𝑖 sin 𝛽𝑥) + 𝑐2 𝑒 𝛼𝑥 (cos 𝛽𝑥 − 𝑖 sin 𝛽𝑥)

However, we expect to have a real solution. That is :

𝑦1 = 𝑒 𝛼𝑥 cos 𝛽𝑥 , 𝑦2 = 𝑒 𝛼𝑥 sin 𝛽𝑥

and the general solution:

𝑦 = 𝑒 𝛼𝑥 (cos 𝛽𝑥 + sin 𝛽𝑥)

Example 2.12: Solve the following IVP

′
𝑦 ′ − 8𝑦 ′ + 17𝑦 = 0 , 𝑦(0) = −4, 𝑦 ′ (0) = −1
Solution: The characteristic equation is :
𝜆2 − 8𝜆 + 17 = 0
The roots are, 𝜆 1,2 = 4 ± 𝑖
The general solution as well as its derivative is

𝑦(𝑥) = 𝑐1 𝑒 4𝑥 cos(𝑥) + 𝑐2 𝑒 4𝑥 sin(𝑥)

𝑦 ′ (𝑥) = 4𝑐1 𝑒 4𝑥 cos(𝑥) − 𝑐1 𝑒 4𝑥 sin(𝑥) + 4𝑐2 𝑒 4𝑥 sin(𝑥) + 𝑐2 𝑒 4𝑥 cos(𝑥)

Applying the initial conditions gives the following system,

−4 = 𝑦(0) = 𝑐1

−1 = 𝑦 ′ (0) = 4𝑐1 + 𝑐2

Solving this system gives 𝑐1 = −4 , and 𝑐2 = 15

The particular solution to the IVP is then

𝑦(𝑥) = −4𝑒 4𝑥 cos(𝑥) + 15𝑒 4𝑥 sin(𝑥)

Summary of case I-III

Case Roots Basis General Solution

I Distinct real 𝑒 𝜆1 𝑥 , 𝑒 𝜆2 𝑥 𝑦 = 𝑐1 𝑒 𝜆1 𝑥 + 𝑐2 𝑒 𝜆2 𝑥

Badri A, Moges B. and Teklebrhan B. 40 AKU

 APPLIED MATHEMATICS III

𝜆1 , 𝜆2

𝑎 𝑎 𝑎
II Real double root 𝑒 −2𝑥 , 𝑥𝑒 −2𝑥 𝑦 = 𝑒 −2𝑥 (𝑐1 + 𝑐2 𝑥)

𝑎
𝜆 = −2

III Complex Conjugate

𝜆1 = 𝛼 + 𝑖𝛽 𝑒 𝛼𝑥 cos 𝛽𝑥 𝑦 = 𝑒 𝛼𝑥 (𝑐1 cos 𝛽𝑥 + 𝑐2 sin 𝛽𝑥)

𝜆2 = 𝛼 − 𝑖𝛽 𝑒 𝛼𝑥 sin 𝛽𝑥

Exercises

Solve the following

a. 𝑦 ′′ + 2𝑘𝑦 ′ + 𝑘 2 𝑦 = 0 , 𝑘 ≠ 0, 𝑦(0) = 2, 𝑦 ′ (0) = 4 e) 9𝑦 ′′ − 30𝑦 ′ + 25𝑦 = 0

b. 𝑦 ′′ + 𝑦 ′ + 𝑦 = 0, 𝑦(0) = 1, 𝑦 ′ (0) = 3 f) 𝑦 ′′ + 𝑦 ′ − 6𝑦 = 0
c. 𝑦 ′′ + 2𝑦 ′ + 3𝑦 = 0 𝑦(0) = 1, 𝑦 ′ (0) = 3 g) 4𝑦 ′′ − 4𝑦 ′ − 3𝑦 = 0
d. 𝑦 ′′ + 𝑦 = 0 𝑦(𝜋) = 2 , 𝑦 ′ (𝜋) = 1 h) 𝑦 ′′ + 9𝑦 ′ + 20𝑦 = 0

2.4 A Method for Solving Non Homogenous Linear Equations

A form 𝑦 ′′ + 𝑝(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 𝑟(𝑥) (1)

where 𝑟(𝑥) ≠ 0 is the non homogenous linear equation.
The equation
𝑦 ′′ + 𝑝(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0 (2)
is called related homogenous equation of (1).

Badri A, Moges(General
Definition2.3 B. and Teklebrhan B.
solution, Particular solution) 41 AKU

A general solution of the non Homogenous ODE (1) on an open interval I is a solution of the
 APPLIED MATHEMATICS III

A particular solution of (1) on I is a solution obtained from (3) by assigning specific value to the
arbitrary constant 𝑐1 and 𝑐2 in 𝑦ℎ .

Procedures to find the general solution of (1)

i. Find a general solution of (2)

ii. Find a particular solution of (1)
iii. Sum of (i) and (ii) is the general solution of (1)

Question: how we can find a solution 𝑦𝑝 (i,e a particular solution of (1) )

The method of undetermined coefficients is suitable for linear ODEs with constant coefficients a
and b

𝑦 ′′ + 𝑎𝑦 ′ + 𝑏𝑦 = 𝑟(𝑥) (4)

when 𝑟(𝑥) is an exponential function, a power of x, a cosine or sine or sum or product of such
function.

Choice rule for the method of undetermined coefficients

a. Basic rule: if r(x) in (4) is one of the function in the first column in table 2.1 , choose 𝑦𝑝
in the same line and determine its undetermined coefficients by substitute 𝑦𝑝 and its
derivates into (4).
b. Modification rule: if a term in your choice for 𝑦𝑝 happens to be a solution of the
homogenous ODE corresponding to (4), multiply this term by 𝑥 (or 𝑥 2 if this solution
corresponding to a double root of the characteristic equation of the homogenous ODE)

Badri A, Moges B. and Teklebrhan B. 42 AKU

 APPLIED MATHEMATICS III

c. Sum rule: if 𝑟(𝑥) is a sum of functions in the first column, choose 𝑦𝑝 the sum of
functions in the corresponding lines of the 2nd column.

Table 2.1. Method of undetermined coefficients

Terms in 𝒓(𝒙) Choice for 𝑦𝑝 (𝑥)

𝑘𝑒 𝛾𝑥 𝑐𝑒 𝛾𝑥
𝑘𝑥 𝑛 𝑛 = 0,1,2, … 𝑘𝑛 𝑥 𝑛 + 𝑘𝑛−1 𝑥 𝑛−1 + ⋯ + 𝑘1 𝑥 + 𝑘0
𝑘𝑐𝑜𝑠 𝛽𝑥 𝑘𝑐𝑜𝑠 𝛽𝑥 + 𝑀𝑠𝑖𝑛 𝛽𝑥
𝑘 sin 𝛽𝑥
𝑘𝑒 𝛼𝑥 cos 𝛽𝑥 𝑒 𝛼𝑥 (𝑘 cos 𝛽𝑥 + 𝑀𝑠𝑖𝑛 𝛽𝑥
𝑘𝑒 𝛼𝑥 sin 𝛽𝑥

Example 2.13: a. Solve the initial value problem

𝑦 ′′ + 𝑦 ′ = 0.001𝑥 2 , 𝑦(0) = 0 , 𝑦 ′ (0) = 1.5

Step 1: General solution of the homogeneous ODE

The characteristic equation for this differential equation and its roots are :

𝜆2 + 1 = 0

⟹ 𝜆=𝑖

The complementary (associated) or homogeneous solution is, then

𝑦ℎ = 𝐴𝑐𝑜𝑠𝑥 + 𝐵𝑠𝑖𝑛𝑥

Step 2: Solution 𝒚𝒑 of the nonhomogeneous ODE .

Let us guess for 𝑦𝑝 (𝑥) , i.e.

𝑦𝑝 = 𝐾2 𝑥 2 + 𝐾1 𝑥 + 𝐾0

(Plugging this into the differential equation)

Badri A, Moges B. and Teklebrhan B. 43 AKU

 APPLIED MATHEMATICS III

Then,

𝑦𝑝 ′′ + 𝑦𝑝 = 2𝐾2 + 2𝑥 2 + 𝐾1 𝑥 + 𝐾0 = 0.001𝑥 2

Collecting like terms gives,

𝐾2 = 0.001, 𝐾1 = 0 , 2𝐾2 + 𝐾 0 = 0

Hence,

𝐾0 = −2𝐾2 = 0.002

This gives

𝑦𝑝 = 0.001𝑥 2 − 0.002 , and

𝑦 = 𝑦ℎ + 𝑦𝑝 = 𝐴𝑐𝑜𝑠𝑥 + 𝐵𝑠𝑖𝑛𝑥 + 0.001𝑥 2 − 0.002

Step 3: Solution of the initial value problem

Setting = 0 , and using the first initial condition gives ,

𝑦 = 𝐴𝑐𝑜𝑠(0) + 𝐵𝑠𝑖𝑛(0) + 0.001(0)2 − 0.002

= 𝐴 − 0.002 = 0 or 𝐴 = 0.002

By differentiation and from the second initial condition

𝑦 ′ = 𝑦ℎ ′ + 𝑦𝑝 ′ = −𝐴𝑠𝑖𝑛𝑥 + 𝐵𝑐𝑜𝑠𝑥 + 0.002𝑥 ,and

𝑦 ′ (0) = 𝐵 = 1.5

Thus a particular solution to the differential equation is then

𝑦 = 0.002𝑐𝑜𝑠𝑥 + 1.5𝑠𝑖𝑛𝑥 + 0.001𝑥 2 − 0.002

b. Determine a particular solution to

𝑦 ′′ − 4𝑦 ′ − 12𝑦 = 3𝑒 5𝑥

Solution: To find a particular solution, let us guess 𝑦𝑝 (𝑥) as

Badri A, Moges B. and Teklebrhan B. 44 AKU

 APPLIED MATHEMATICS III

𝑦𝑝 (𝑥) = 𝐴𝑒 5𝑥
Plugging into the differential equation gives,
25𝐴𝑒 5𝑥 − 20𝐴𝑒 5𝑥 − 12(𝐴𝑒 5𝑥 ) = 3𝑒 5𝑥
−7𝐴𝑒 5𝑥 = 3𝑒 5𝑥
−3
𝐴=
7
A particular solution to the differential equation is
−3 5𝑥
𝑦𝑝 (𝑥) = 𝑒
7

C. Solve the initial value problem

𝑦 ′′ + 3𝑦 ′ − 2.25𝑦 = −10𝑒 −1.5𝑥 , 𝑦(0) = 1 , 𝑦 ′ (0) = 0

Answer: (𝟏 + 𝟏. 𝟓𝒙 − 𝟓𝒙𝟐 )𝒆−𝟏.𝟓𝒙

Exercises

I. Find a general solution

a. 𝑦 ′′ + 3𝑦 ′ + 2𝑦 = 30𝑒 2𝑥
1
b. 𝑦 ′′ + 9𝑦 = 𝑐𝑜𝑠 𝑥 + 3 𝑐𝑜𝑠 3𝑥

c. 𝑦 ′′ + 4𝑦 ′ + 6.25𝑦 = 3.125(𝑥 + 1)2

d. 𝑦 ′′ + 9𝑦 = 8𝑥 + 36 𝑠𝑖𝑛 3𝑥
e. 𝑦 ′′ − 3𝑦 ′ + 2𝑦 = 4𝑥 + 𝑒 3𝑥
II. Solve the following initial value problem
a. 𝑦 ′′ + 2𝑦 ′ + 0.75𝑦 = 2 𝑐𝑜𝑠(𝑥) − 0.25 𝑠𝑖𝑛(𝑥) + +0.09𝑥,
, 𝑦(0) = 2.78, 𝑦 ′ (0) = −0.43

b. 𝑦 ′′ − 3𝑦 ′ + 2.25𝑦 = 27(𝑥 2 + 𝑥) 𝑦(0) = 20, 𝑦 ′ (0) = 30

Badri A, Moges B. and Teklebrhan B. 45 AKU

 APPLIED MATHEMATICS III

c. 𝑦 ′′ − 2𝑦 ′ = 12𝑒 2𝑥 − 8𝑒 −2𝑥 𝑦(0) = −2, 𝑦 ′ (0) = 12

d. 𝑦 ′′ + 2𝑦′ + 10𝑦 = 17 𝑠𝑖𝑛 𝑥 − 37 𝑠𝑖𝑛 3𝑥 𝑦(0) = 6.6, 𝑦 ′ (0) = 2.2

e. 𝑦 ′′ + 3𝑦 ′ + 2.25𝑦 = −10𝑒 −1.5 𝑦(0) = 1 , 𝑦 ′ (0) = 0

Solution by Variation of parameters

We continue our discussion of nonhomogeneous linear ODEs

𝑦 ′′ + 𝑝(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 𝑟(𝑥) … (1)

In the previous discussion, we have seen that a general solution of (1) is the sum of a general
solution 𝑦ℎ of the corresponding homogeneous ODE and any particular solution 𝑦𝑝 of (1). To
obtain 𝑦𝑝 when 𝑟(𝑥) is not too complicated we can often use the method of undetermined
coefficients . However, since this method restricted to function , 𝑟(𝑥) whose derivatives are of a
form similar to 𝑟(𝑥) itself (powers, exponential functions etc.), it is desirable to have a method
valid for more general ODEs (1), we shall now develop. It is called the method of variation of
parameters and is credited to Lagrange. Here 𝑝, 𝑞, 𝑟 in (1) may be variable (given functions of
𝑥), but we assume that they are continuous on some open I.

Aim: To find 𝒚𝒑 for (1)

Lagrange’s method gives a particular solution 𝑦𝑝 of (1)

𝑦2 𝑟 𝑦1 𝑟
𝑦𝑝 = −𝑦1 ∫ 𝑑𝑥 + 𝑦2 ∫ 𝑑𝑥 … (2)
𝑊 𝑊

, where W is the Wronskian of 𝑦1 and 𝑦2 form a basis of solutions of the corresponding

homogeneous ODE

𝑦 ′′ + 𝑝(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0

on I and W is the Wronskian of 𝑦1 , 𝑦2

𝑊 = 𝑦1 𝑦 ′ 2 − 𝑦2 𝑦 ′1

Badri A, Moges B. and Teklebrhan B. 46 AKU

 APPLIED MATHEMATICS III

CAUTION! The solution formula (2) is obtained under the assumption that the ODE is written
in standard form, with 𝑦 ′′ as the first term as shown in (1). If it starts with 𝑓(𝑥)𝑦 ′′ , divide first
by 𝑓(𝑥).

Example 2.14: Solve the nonhomogeneous ODE

1
𝑦 ′′ + 𝑦 = 𝑠𝑒𝑐𝑥 =
𝑐𝑜𝑠𝑥

Solution: A basis of solution of the homogeneous ODE an any interval is

𝑦1 = 𝑐𝑜𝑠𝑥 , 𝑦2 = 𝑠𝑖𝑛𝑥

This gives the Wronskian

𝑊(𝑦1 , 𝑦2 ) = 𝑐𝑜𝑠𝑥 𝑐𝑜𝑠𝑥 − 𝑠𝑖𝑛𝑥(−𝑠𝑖𝑛𝑥) = 1

From (2) choosing zero constants of integration, we get the particular solution of the given
ODE 𝑦𝑝 = −𝑐𝑜𝑠𝑥 ∫ 𝑠𝑖𝑛𝑥𝑠𝑒𝑐𝑥 𝑑𝑥 + 𝑠𝑖𝑛𝑥 ∫ 𝑐𝑜𝑠𝑥𝑠𝑒𝑐𝑥 𝑑𝑥

= 𝑐𝑜𝑠𝑥𝑙𝑛|𝑐𝑜𝑠𝑥| + 𝑥𝑠𝑖𝑛𝑥

Now using 𝑦𝑝 and the general solution

𝑦ℎ = 𝑐1 𝑦1 + 𝑐2 𝑦2

,of the homogeneous ODE we obtain the answer

𝑦 = 𝑦ℎ + 𝑦𝑝

= (𝑐1 + 𝑙𝑛|𝑐𝑜𝑠𝑥| )𝑐𝑜𝑠𝑥 + (𝑐2 + 𝑥)𝑠𝑖𝑛𝑥

Exercises

Solve the given nonhomogeneous ODE by variation of parameters or undetermined coefficients.

Give a general solution.

a. 𝑦 ′′ + 2𝑦 ′ + 𝑦 = 𝑥𝑒 −𝑥 e. 𝑦 ′′ − 2𝑦 ′ + 𝑦 = 𝑒 𝑥 𝑠𝑖𝑛𝑥
b. 𝑦 ′′ + 𝑦 = 𝑠𝑒𝑐 𝑥 f. 𝑦 ′′ + 𝑦 = 𝑡𝑎𝑛 𝑥

Badri A, Moges B. and Teklebrhan B. 47 AKU

 APPLIED MATHEMATICS III

c. 𝑥 2 𝑦 ′′ − 2𝑥 2 𝑦 ′ + 𝑥 2 𝑦 = 𝑒 𝑥 g. 𝑦 ′′ + 𝑦 = 𝑐𝑜𝑠𝑥 + 𝑠𝑒𝑐𝑥
d. 𝑥 2 𝑦 ′′ + 𝑥𝑦 ′ − 𝑦 = 𝑥 2 𝑙𝑛 𝑥 f. 𝑥𝑦 ′′ − 2𝑥𝑦 ′ + 2𝑦 = 𝑥 3 𝑐𝑜𝑠𝑥

2.5 System of Differential Equation

A general system of n first order linear variable coefficient DE involving the n dependent
variables 𝑥1 (𝑡), 𝑥2 (𝑡), … 𝑥𝑛 (𝑡) that are functions of the independent variable t (in application t is
often the time), the variable coefficients 𝑎𝑖𝑗 (𝑥) and the non homogenous terms
𝑓1 (𝑡), 𝑓2 (𝑡), … , 𝑓𝑛 (𝑡) has the form

𝑥′1 (𝑡) = 𝑎11 (𝑡)𝑥1 (𝑡) + 𝑎12 (𝑡)𝑥2 (𝑡) + 𝑎13 (𝑡)𝑥3 (𝑡) + ⋯ + 𝑎1𝑛 (𝑡)𝑥𝑛 (𝑡) + 𝑓1 (𝑡)
(1) 𝑥′2 (𝑡) = 𝑎21 (𝑡)𝑥1 (𝑡) + 𝑎22 (𝑡)𝑥2 (𝑡) + 𝑎23 (𝑡)𝑥3 (𝑡) + ⋯ + 𝑎2𝑛 (𝑡)𝑥𝑛 (𝑡) + 𝑓2 (𝑡)
𝑥′3 (𝑡) = 𝑎31 (𝑡)𝑥1 (𝑡) + 𝑎32 (𝑡)𝑥2 (𝑡) + 𝑎33 (𝑡)𝑥3 (𝑡) + ⋯ + 𝑎3𝑛 (𝑡)𝑥𝑛 (𝑡) + 𝑓3 (𝑡)
⋮
𝑥′𝑛 (𝑡) = 𝑎𝑛1 (𝑡)𝑥1 (𝑡) + 𝑎𝑛2 (𝑡)𝑥2 (𝑡) + 𝑎𝑛3 (𝑡)𝑥3 (𝑡) + ⋯ + 𝑎𝑛𝑛 (𝑡)𝑥𝑛 (𝑡) + 𝑓𝑛 (𝑡)
System (1) is said to be:
 Homogenous: when all the functions 𝑓𝑖 (𝑡) are zero.
 Non Homogenous: when at least one of them is non zero.
 Variable Coefficient system: whenever at least one of the Coefficients 𝑎𝑖𝑗 (𝑡) is a function
of t: otherwise, it is a constant coefficient system.
 The system (1) is linear system b/c it is linear in the function 𝑥1 (𝑡), 𝑥2 (𝑡), … 𝑥𝑛 (𝑡) and
their derivatives.
 An initial value problem for system (1) involves seeking a solution of (1) such that at 𝑡 =
𝑡0 the variables 𝑥1 (𝑡), 𝑥2 (𝑡), … 𝑥𝑛 (𝑡) satisfies the initial conditions.
𝑥1 (𝑡0 ) = 𝑘1 , 𝑥2 (𝑡0 ) = 𝑘2 , … , 𝑥𝑛 (𝑡0 ) = 𝑘𝑛 (2)
where ∀𝑘𝑖 , 𝑖 = 1,2, . . , 𝑛 are given constant.
The matrix notation of (1) is
𝑥 ′ (𝑡) = 𝐴(𝑡)𝑥(𝑡) + 𝑏(𝑡) (3)
or more simplify as 𝑥 ′ = 𝐴𝑥 + 𝑏 where

Badri A, Moges B. and Teklebrhan B. 48 AKU

 APPLIED MATHEMATICS III

𝑥1 (𝑡) 𝑥′1 (𝑡)

𝑥 (𝑡)
𝑥(𝑡) = [ 2 ] 𝑥 ′ (𝑡) = [𝑥′2 (𝑡) ] (4)
⋮ ⋮
𝑥𝑛 (𝑡) 𝑥′𝑛 (𝑡)
𝑎11 (𝑡) 𝑎12 (𝑡) ⋯ 𝑎1𝑛 (𝑡) 𝑓1 (𝑡)
𝑎 (𝑡) 𝑎22 (𝑡) ⋯ 𝑎2𝑛 (𝑡) 𝑓 (𝑡)
𝐴(𝑡) = [ 21 ] 𝑏(𝑡) = [ 2 ]
⋮ ⋮ ⋮ ⋮ ⋮
𝑎𝑛1 (𝑡) 𝑎𝑛2 (𝑡) ⋯ 𝑎𝑛𝑛 (𝑡) 𝑓𝑛 (𝑡)

The 𝑛 × 1 vector 𝑥(𝑡) is called the solution vector. The 𝑛 × 𝑛 matrix 𝐴(𝑡) is
called the coefficient matrix and the 𝑛 × 1 vector 𝑏(𝑡) is called the non
homogenous form of the system. ,The system (3) becomes an initial value
problem for the solution 𝑥(𝑡) when at 𝑡 = 𝑡0 the vector 𝑥(𝑡) is required to satisfy
the initial condition

𝑘1
𝑘
𝑥(𝑡0 ) = [ 2 ] (5)
⋮
𝑘𝑛

where, 𝑥(𝑡0 ) is the initial vector and 𝑘1 , 𝑘2 , … 𝑘𝑛 are given constants.

Example 2.15. Express in matrix form the IVP

𝑥′1 = 2𝑥1 − 𝑥2 + 4 − 𝑡 2
𝑥′2 = −𝑥1 + 2𝑥2 + 1 with 𝑥1 (0) = 1, 𝑥2 (0)=0
Solution: In matrix form this is 𝑥 ′ (𝑡) = 𝐴(𝑡)𝑥(𝑡) + 𝑏(𝑡) ,

2 −1 𝑥1 2 1
where, 𝐴 = [ ] 𝑥(𝑡) = [𝑥 ] and, 𝑏(𝑡) = [4 − 𝑡 ],with initial vector 𝑥(0) = [ ].
−1 2 2 1 0

Eigen value and Eigen vectors of a matrix

We now restrict our discussion to homogeneous first-order systems with constant coefficients:
those of the form:
𝑥′1 (𝑡) = 𝑎11 (𝑡)𝑥1 (𝑡) + 𝑎12 (𝑡)𝑥2 (𝑡) + 𝑎13 (𝑡)𝑥3 (𝑡) + ⋯ + 𝑎1𝑛 (𝑡)𝑥𝑛 (𝑡)

Badri A, Moges B. and Teklebrhan B. 49 AKU

 APPLIED MATHEMATICS III

𝑥′2 (𝑡) = 𝑎21 (𝑡)𝑥1 (𝑡) + 𝑎22 (𝑡)𝑥2 (𝑡) + 𝑎23 (𝑡)𝑥3 (𝑡) + ⋯ + 𝑎2𝑛 (𝑡)𝑥𝑛 (𝑡)
𝑥′3 (𝑡) = 𝑎31 (𝑡)𝑥1 (𝑡) + 𝑎32 (𝑡)𝑥2 (𝑡) + 𝑎33 (𝑡)𝑥3 (𝑡) + ⋯ + 𝑎3𝑛 (𝑡)𝑥𝑛 (𝑡)
⋮
𝑥′𝑛 (𝑡) = 𝑎𝑛1 (𝑡)𝑥1 (𝑡) + 𝑎𝑛2 (𝑡)𝑥2 (𝑡) + 𝑎𝑛3 (𝑡)𝑥3 (𝑡) + ⋯ + 𝑎𝑛𝑛 (𝑡)𝑥𝑛 (𝑡)

Let 𝐴 = [𝑎𝑗𝑘 ] be an 𝑛 × 𝑛 matrix and consider the equation 𝐴 𝑥= 𝜆𝑥 where 𝜆 is a scalar (real or
complex). A scalar 𝜆 which satisfies
𝐴𝑥 = 𝜆x … (*)
for some x ≠ 0 is called an eigenvalue of A and this non zero vector is called an
eigenvector of A corresponding to this 𝜆 .(*) can be written as
𝐴𝑥 − 𝜆𝑥=0 … (**)
(𝐴 − 𝜆)x=0 … (***)
For these equations to have a solution x≠ 0 the determinant of the coefficient matrix
𝐴 − 𝜆𝐼 must be zero. Then (***) be comes
𝑎11 − 𝜆 𝑎12
[ ] (𝑥1 ) = (00)
𝑎21 𝑎22 − 𝜆 𝑥2
(𝑎11 − 𝜆)𝑥1 + 𝑎12 𝑥2 = 0
𝑎21 𝑥1 + (𝑎22 − 𝜆)𝑥2 = 0 … (𝑖)
𝑎11 − 𝜆 𝑎12
Solving [ ]=0
𝑎21 𝑎22 − 𝜆
for 𝜆 and substituting in (i) we get 𝑋1and 𝑋2 for 𝜆1 and 𝑋3 , 𝑋4 for 𝜆2 and thus, The Eigen vector
corresponds to 𝜆1 will be
𝑋1
𝑥⃗1 = ( )
𝑋2
and the Eigen vector corresponding to 𝜆2 will be
𝑋3
𝑋⃗2 = ( )
𝑋4
Theorem; Given 𝑋⃗ = 𝐴𝑥⃗ if 𝜆1 and 𝜆2 are the eigen values of the matrix A and 𝑦⃗1 and 𝑦⃗2 are
corresponding Eigen vectors then the general solution of the system is
𝑦 = 𝐶1 𝑒 𝜆1 t 𝑦⃗1 + 𝐶2 𝑒 𝜆2 t 𝑦⃗2
1 12
Example 2.16: Solve 𝑥⃗ = [ ] 𝑥⃗
3 1

Badri A, Moges B. and Teklebrhan B. 50 AKU

 APPLIED MATHEMATICS III

1 12
Solution: we have 𝐴 = [ ]
3 1
And (A − λI)=[1 − λ 12
].
3 1−λ
1−λ 12
det(A − λI)=| |=0
3 1−λ
= (1 − λ)2 − 36 ⟹ 1 − λ = ±6
This implies, 𝜆1 = 7 𝜆2 = −5
Now
1−𝜆 12 𝑥1
[ ] [𝑥 ] = 0
3 1−𝜆 2

For 𝜆1 = 7 , we have
−6 12 𝑥1
[ ] [ ]=0
3 −6 𝑥2

-6𝑥1 + 12𝑥2 = 0, i.e. 𝑥1 = 2𝑥2

So the eigenvectors are of the form
𝑥1
[2𝑥 ]
2

and the Eigen space is 1-dimensional with a basis given

1
[ ]
2
Similarly for 𝜆2 = −5, we have
6 12 𝑥1
[ ] [ ]=0
3 6 𝑥2
6𝑥1 + 12𝑥2 = 0, i.e. 𝑥1 = −2𝑥2
𝑥1
So the eigenvectors are of the form [−2𝑥 ]
2
1
and the Eigen space is 1-dimensional with a basis given [ ]
−2
Hence, the general solution of the system is
𝑦 = 𝐶1 𝑒 𝜆1 t 𝑦⃗1 + 𝐶2 𝑒 𝜆2 t 𝑦⃗2
1 −1
Therefore, y = C1 e7t [ ] + c2 e−5t [ ]
2 2
Example 2.17: Find all functions 𝑥1 and 𝑥2 such that
𝑥′1 = 𝑥1 − 3𝑥2

Badri A, Moges B. and Teklebrhan B. 51 AKU

 APPLIED MATHEMATICS III

𝑥′2 = 𝑥1 + 5𝑥2
Answer: The general solution is
x1 −3 −1
[x ] = C1 [ ] ∙ e2t + C2 [ ] . e4t
2 1 1

Exercises

Find the general solution to the following system

a. 𝑋⃗ = 𝐴𝑥⃗ given by
1 −1 4
⃗
𝑋 = [3 2 −1] 𝑥⃗ , where 𝜆 = 1, −2,3
2 1 −1

b. 𝑥′1 = 4𝑥1 − 𝑥2 − 2𝑥3

𝑥′2 = 2𝑥1 + 𝑥2 − 2𝑥3
𝑥′3 = 5𝑥1 − 3𝑥3

c. 𝑥′1 = 𝑥1 + 𝑥2
𝑥′2 = 4𝑥1 + 𝑥2

d. Given 𝑋⃗ = 𝐴𝑥⃗ , where

4 0 −1
𝐴 = [2 2 − 1]
3 1 0

Matrix Exponentials (for general coefficient matrices)

If the coefficient matrix is not diagonalizable, life is more difficult, as we cannot generate a basis
for the solution space using eigenvectors alone. We can still solve the system using chains of
generalized eigenvectors. However, there are some slightly cumbersome technical problems that
can occur when any defective eigenspace has more than one independent eigenvector: in order to
generate enough solutions, in general one must construct a chain above each element of a basis
for a defective eigenspace.

Badri A, Moges B. and Teklebrhan B. 52 AKU

 APPLIED MATHEMATICS III

But some of the generalized eigenvectors obtained from these chains may yield linearly
dependent solution functions.
If we consider the 1 × 1 system 𝑥′ = 𝑘𝑥 with the initial condition x(0) = 𝐶 we know that the
general solution is
𝑥(𝑡) = 𝑒 𝑘𝑡 𝐶.
 We would like to find some way to extend this result to 𝑛 × 𝑛 systems.
 This leads us to define the exponential of a matrix 𝑒 𝐴 .

Definition: If 𝐴 is an 𝑛 × 𝑛 matrix then we define the exponential of 𝐴, denoted by 𝑒 𝐴 , to

𝐴𝑛
be the infinite sum 𝑒 𝐴 = ∑∞
𝑛=0 𝑛! .

Note: The definition is motivated by the Tyalor series for the exponential of a real or complex
𝑧𝑛
number 𝑧; namely 𝑒 𝑧 = ∑∞
𝑛=0 𝑛! .

𝐴𝑛
Theorem: The infinite sum 𝑒 𝐴 = ∑∞
𝑛=0 Converges for every matrix 𝐴.
𝑛!

Theorem: If 𝐴 is an 𝑛 × 𝑛 matrix, then the unique solution to the initial value problem
⃗⃗⃗⃗(0) = 𝑦⃗0 is given by 𝑦⃗(𝑥) = 𝑒 𝐴𝑥 ∙ 𝑦⃗0 .
𝑦⃗ = 𝐴 ∙ 𝑦⃗ with 𝑦

−1 𝐴𝑃
Proposition: For any invertiable matrix 𝑃, 𝑒 𝑝 = 𝑃−1 [𝑒 𝐴 ]𝑃.

Proof:
−1 𝐴𝑃 (𝑝−1 𝐴𝑃)𝑛 𝐴𝑛
𝑒𝑝 = ∑∞
𝑛=0 = 𝑃−1 (∑∞ −1 𝐴
𝑛=0 𝑛! ) 𝑃 = 𝑃 [𝑒 ]𝑃,
𝑛!

,where the middle step uses the fact that (𝑝−1 𝐴𝑃)𝑛 = 𝑃(𝐴𝑛 )𝑃.

Proposition: If 𝐷 is a diagonal matrix with diagonal entries 𝜆1 , 𝜆2 , 𝜆3 , . . . 𝜆𝑛 then 𝑒 𝐷 is the

diagonal matrix with diagonal entries 𝑒 𝜆1 , 𝑒 𝜆2 , . .. , 𝑒 𝜆𝑛 .

0 −2
Example 2.18: Find 𝑒 𝐴𝑥 , if 𝐴 = [ ]
3 5
Solution: First we attempt to diagonalize the matrix 𝐴. We calculate
det(𝑡𝐼 − 𝐴) = 𝑡(𝑡 − 5) + 6 = (𝑡 − 2)(𝑡 − 3)

Badri A, Moges B. and Teklebrhan B. 53 AKU

 APPLIED MATHEMATICS III

So, the eigenvalues are 𝜆 = 2,3.

For 𝜆 = 2, we need to solve

0 −2 𝑎 𝑎
[ ] ∙ [ ] = 2[ ]
3 5 𝑏 𝑏
So,
−2𝑏 2𝑎
[ ]=[ ]
3𝑎 + 5𝑏 2𝑏
and thus 𝑎 = −𝑏.
−𝑏 −1
The eigenvectors are of the form [ ], so a basis for 𝜆 = 2 eigenspace is [ ].
𝑏 1

For 𝜆 = 3, we need to solve

0 −2 𝑎 𝑎
[ ] ∙ [ ] = 3[ ]
3 5 𝑏 𝑏
So,
−2𝑏 3𝑎
[ ]=[ ]
3𝑎 + 5𝑏 3𝑏
−2
and thus 𝑎= 𝑏.
3
−2
𝑏 −2
The eigenvectors are of the form [ 3 ], so a basis for the 𝜆 = 3 eigenspace is [ ].
𝑏 3
Since the eigenvalues are distinct ,clearly 𝐴 is diagonalizable: we can write
2 0 −1 2
𝐴 = 𝑃−1 𝐷𝑃 for 𝐷 = [ ] and 𝑃 = [ ].
0 3 1 3
−3 −2
We also compute 𝑃−1 = [ ].
1 1
2𝑥
Now we compute 𝑒 𝐷𝑥 = [ 𝑒 0]
0 𝑒 3𝑥
From the formula for exponentiating diagonal matrices.

Finally we have

𝑒 𝐴𝑥 = 𝑃𝑒 𝐷𝑥 𝑃−1 = [
−1 − 2 𝑒 2𝑥 0 −3 − 2
][ ] [ ] = [ 3𝑒 2𝑥 − 2𝑒 3𝑥 2𝑒 2𝑥 − 2𝑒 3𝑥 ]
1 3 0 𝑒 3𝑥 1 1 −3𝑒 2𝑥 + 3𝑒 3𝑥 − 2𝑒 3𝑥 + 3𝑒 3𝑥

Unit Summary:

Badri A, Moges B. and Teklebrhan B. 54 AKU

 APPLIED MATHEMATICS III

 Linear second order differential equations with constant coefficients are the simplest of the
higher order differential equation and they have many applications.
 The most general linear second order differential equation is in the form:

𝑝(𝑥)𝑦 ′′ + 𝑞(𝑥)𝑦 ′ + 𝑟(𝑥)𝑦 = 𝑔(𝑥)

The equation is called non-homogenous when 𝑔(𝑥)zero is not identically; otherwise it is
called homogenous.

 A second order ODE is called linear if it can be written as

𝑦 ′′ + 𝑝(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 𝑟(𝑥) (1)

where, p and q are called the coefficient of the ODEs and is a function of x and nonlinear
if it cannot be written in this form.

 A function 𝑦 = ℎ(𝑥) is called a solution of a (linear or non linear) second order ODE on
some open interval 𝐼 if ℎ is defined and twice differentiable throughout the interval and is
such that the ODE becomes an identify if we replace the unknown 𝑦 by ℎ, and its successive
derivatives.
 Two function 𝑦1 (𝑥) and 𝑦2 (𝑥) are said to be linearly independent (LI) over an interval I if
the equation
𝑐1 𝑦1 (𝑥) + 𝑐2 𝑦2 (𝑥) = 0
 The Wronskian of 𝑛 function 𝑦1 (𝑥) ,𝑦2 (𝑥), 𝑦3 (𝑥),… 𝑦𝑛 (𝑥) each (n-1) time differentiable is
the form
𝑦1 𝑦2 ⋯⋯ 𝑦𝑛
𝑦1 ′ 𝑦2 ′ ⋯⋯ 𝑦𝑛 ′
𝑊(𝑦) = || 𝑦1 ′′ 𝑦2 ′′ ⋯⋯ 𝑦𝑛 ′′ | , where |
| | denotes the determinant.
⋮ ⋮ ⋯⋯ ⋮
𝑦1 (𝑛−1) 𝑦2 (𝑛−1) ⋯⋯ 𝑦𝑛 (𝑛−1)
 A general solution of an ODE
𝑦 ′′ + 𝑝(𝑥)𝑦 ′ + 𝑞(𝑥)𝑦 = 0
On an open interval I is a solution of
𝑦 = 𝑐1 𝑦1 + 𝑐2 𝑦2

Badri A, Moges B. and Teklebrhan B. 55 AKU

 APPLIED MATHEMATICS III

In which 𝑦1 and 𝑦2 are solution of (*) on 𝐼 that are not proportional (LI) and 𝑐1 and 𝑐2 are
arbitrary constants. These 𝑦1 and 𝑦2 are called a basis (or a fundamental system of the solution
(*) on 𝐼)
 A basis of solution (*) on an open interval 𝐼 is a pair of LI solution of (*) on 𝐼.
 The equation
𝜆2 + 𝑎𝜆 + 𝑏 = 0 (2)
is called the characteristic equation of the given differential equation.
 Depending on the sign of the discriminant 𝑎2 − 4𝑏 the quadratic equation (2) may have
three kinds of roots.

Case I: Two real roots, if 𝑎2 − 4𝑏 > 0

Case II: A real double root , if 𝑎2 − 4𝑏 = 0

Case III: Complex conjugate roots, if 𝑎2 − 4𝑏 < 0

 A general solution of the non Homogenous ODE (1) on an open interval I is a solution of the
form 𝑦(𝑥) = 𝑦ℎ (𝑥) + 𝑦𝑝 (𝑥) (3)

Where , 𝑦ℎ = 𝑐1 𝑦1 + 𝑐2 𝑦2 is a general solution of the homogeneous ODE (2) on I and 𝑦𝑝 is

any solution of (1) on I containing no arbitrary constant.

 If 𝐴 is an 𝑛 × 𝑛 matrix then we define the exponential of 𝐴, denoted by 𝑒 𝐴 , to be the infinite

𝐴𝑛
sum 𝑒 𝐴 = ∑∞
𝑛=0 𝑛! .

 A general system of n first order linear variable coefficient DE involving the n dependent
variables 𝑥1 (𝑡), 𝑥2 (𝑡), … 𝑥𝑛 (𝑡) that are functions of the independent variable t (in application
t is often the time), the variable coefficients 𝑎𝑖𝑗 (𝑥) and the non homogenous terms
𝑓1 (𝑡), 𝑓2 (𝑡), … , 𝑓𝑛 (𝑡) has the form
𝑥′1 (𝑡) = 𝑎11 (𝑡)𝑥1 (𝑡) + 𝑎12 (𝑡)𝑥2 (𝑡) + 𝑎13 (𝑡)𝑥3 (𝑡) + ⋯ + 𝑎1𝑛 (𝑡)𝑥𝑛 (𝑡) + 𝑓1 (𝑡)
(1) 𝑥′2 (𝑡) = 𝑎21 (𝑡)𝑥1 (𝑡) + 𝑎22 (𝑡)𝑥2 (𝑡) + 𝑎23 (𝑡)𝑥3 (𝑡) + ⋯ + 𝑎2𝑛 (𝑡)𝑥𝑛 (𝑡) + 𝑓2 (𝑡)
𝑥′3 (𝑡) = 𝑎31 (𝑡)𝑥1 (𝑡) + 𝑎32 (𝑡)𝑥2 (𝑡) + 𝑎33 (𝑡)𝑥3 (𝑡) + ⋯ + 𝑎3𝑛 (𝑡)𝑥𝑛 (𝑡) + 𝑓3 (𝑡)
⋮
𝑥′𝑛 (𝑡) = 𝑎𝑛1 (𝑡)𝑥1 (𝑡) + 𝑎𝑛2 (𝑡)𝑥2 (𝑡) + 𝑎𝑛3 (𝑡)𝑥3 (𝑡) + ⋯ + 𝑎𝑛𝑛 (𝑡)𝑥𝑛 (𝑡) + 𝑓𝑛 (𝑡)

Badri A, Moges B. and Teklebrhan B. 56 AKU

 APPLIED MATHEMATICS III

Miscellaneous Exercises

1. What is a general solution of linear ODE ? A basis of solution?

2. How would you obtain a general solution of a nonhomogeneous linear ODE if you know a general
solution of the corresponding homogeneous ODE?
3. What does an initial value problem for a second order ODE look like?
4. What is a particular solution and why is it more common than a general solution as the answer to
practical problems?
5. Why are second –order ODEs more important in modeling than ODEs of higher order ?
6. Test the given pairs of functions for linear independence or dependence over the stated intervals

a. 𝑠𝑖𝑛ℎ2 𝑥, 𝑐𝑜𝑠ℎ2 , 𝑓𝑜𝑟 all 𝑥

b. 𝑥 + 𝑙𝑛|𝑥|, 𝑥 + 2𝑙𝑛|𝑥|, for|𝑥| ≥ 0.
c. 1 + 𝑥, 𝑥 + 𝑥 2 ,for all x.
d. 𝑠𝑖𝑛𝑥𝑐𝑜𝑠𝑥, 𝑠𝑖𝑛2𝑥, for all 𝑥,
e. 𝑒 2𝑥 , 𝑥𝑒 2𝑥 , for all x.
7. Find a differential equation whose general solution is
a. 𝑦 = 𝑐1 𝑒 2𝑡 + 𝑐2𝑒 −3𝑡
−𝑡
b. 𝑦 = 𝑐1 𝑒 2 + 𝑐2𝑒 −2𝑡
8. Find a general solution. Indicate the method you are using.
a. 𝑦 ′′ − 2𝑦 ′ − 8𝑦 = 52𝑐𝑜𝑠6𝑥 g. 𝑦 ′′ + 𝑦 = 𝑐𝑠𝑐𝑥
′′ ′ −3𝑥 2
b. 𝑦 + 6𝑦 + 9𝑦 = 𝑒 − 27𝑥 h. 𝑦 ′′ − 4𝑦 ′ + 4𝑦 = 𝑥 2 𝑒 𝑥
c. 𝑦 ′′ + 8𝑦 ′ + 25𝑦 = 26𝑠𝑖𝑛3𝑥 i. 𝑥 2 𝑦 ′′ = 𝑥𝑦 ′ + 𝑦 = 𝑥𝑙𝑛|𝑥|
d. 𝑦𝑦 ′′ = 2𝑦 ′2
e. 𝑦 ′′ − 2𝑦 ′ − 3𝑦 = 𝑒 4𝑡
f. 𝑦 ′′ − 2𝑦 ′ + 2𝑦 = 𝑒 𝑡 + 𝑡𝑐𝑜𝑠𝑡

9. Use the general solution of

𝑦 ′′ + 𝑦 = 0
to find if a solution exists and is unique, exists but is non unique, or does not exist for each set of
boundary conditions.
a. 𝑦(0) = 0, 𝑦(𝜋) = 0,
b. 𝑦(0) = 1, 𝑦(2𝜋) = 2
c. 𝑦′(0) = 0, 𝑦 ′ (0) = 0, 𝑦 ′ (𝜋) = 0
d. 𝑦(0) = 0, 𝑦 ′ (𝜋) = 0.
10. Solve the following initial value problems.
a.𝑦 ′′ + 5𝑦 ′ − 14𝑦 = 0, 𝑦(0) = 6, 𝑦 ′ (0) = −6
b.𝑦 ′′ + 6𝑦 ′ + 18𝑦 = 0, 𝑦(0) = 5, 𝑦 ′ (0) = −21
c. 𝑥 2 𝑦 ′′ − 𝑥𝑦 ′ − 24𝑦 = 0, 𝑦(1) = 15, 𝑦 ′ (1) = 0
d. 𝑦 ′′ + 5𝑦 ′ + 6𝑦 = 108𝑥 2 , 𝑦(0) = 18, 𝑦 ′ (0) = −26
e.𝑦 ′′ + 𝑦 ′ + 2.5𝑦 = 13𝑐𝑜𝑠𝑥, 𝑦(0) = 8, 𝑦 ′ (0) = 4.5

Badri A, Moges B. and Teklebrhan B. 57 AKU

 APPLIED MATHEMATICS III

References

1. Erwin Kreyszing,Advanced engineering mathematics, 2006,John Wiley and Sons,Inc.

2. Dennis Gzill, A first course in differential equations with modelin equations,
2011Brooks/Cole Cengage Learning

3.PeterPhili,Methods for Solving Ordinary Differential Equations,2011, Spring

4. HeneryRecardo, A Modern to Introduction to Differential equation,2009,Elsevier Inc.

5. A.C. King, J. Billingnam, Differential Equations ,linear, nonlinear, Ordinary, Partial,2003,

CAMBRIGE .

6.Matthew R and Etla, differential equation with linear algebra,2009, Oxford University press

7.K. Soetaert, T P, Solving Initial Value Diferential Equations in R,2010,R package deS

8. Mircea V. Soare , Ordinary differential equations with applications to Mechanics, 2007

Springer.

9. Larson Edwards, Calculus, 2010, Brooks/Cole CengageLearning

10. A. Ganesh and Etla, Engineering Mathematics II, 2009 New age International press

11. Wilfred Kaplan,Advanced Calculus, 5th edition, publishing house of electronics industry

12.Salas Hille Etgen, Calculus – One and Several variables,10th edition, WILLEY PLUS

13 .Boyce. Diprima, Elementary differential equations and boundary value problem, 2001 ,John
Wiley and Sons.Inc

14. Ravi P. Agarwal. An introduction to differential equation, 2000,Spring

15. Rudolph E. Longer,Ordinary Differential equations,1954,John Wiley and Sons.Inc

Badri A, Moges B. and Teklebrhan B. 58 AKU

 APPLIED MATHEMATICS III

Chapter-Three
Fourier series And Integrals
Introduction

The central starting point of Fourier analysis is Fourier series. They arise naturally while
analyzing many physical phenomenon like electrical oscillations, vibrating mechanical systems,
longitudinal oscillations crystals etc. They are infinite series designed to represent general
periodic functions in terms of simple ones, namely, 𝑐𝑜𝑠𝑖𝑛𝑒𝑠 and 𝑠𝑖𝑛𝑒𝑠.

Fourier series are very important to the engineer and physicist because they allow the solution of
ODEs in connection with forced oscillations and the Approximation of periodic functions.
Where as Fourier integrals extend the concept of Fourier series to non -periodic functions
defined for all x as in many practical problems we come across functions defined on −∞ < 𝑥 <
∞.More over periodic functions can be represented in complex Fourier series and non-periodic
functions can be represented in complex Fourier integral form.

In this chapter we will explore representations of periodic functions in Fourier series, in complex
Fourier series and representations of non-periodic functions in Fourier integral, in complex
Fourier integral. Before introducing Fourier series we will see some preliminary concepts, like
Periodic functions, even and odd functions orthogonal functions and trigonometric series.

Unit Objectives:

On the completion of this unit, students should be able to:

 understand the definition of periodic ,even, odd functions and orthogonal functions;
 understand and find Fourier series representation of periodic functions;
 find Fourier integral representation of non- periodic functions
 Understand and find the Complex Fourier series representation of periodic functions;
 Identify and understand the idea of complex Fourier integral representation.

Badri A, Moges B. and Teklebrhan B. 59 AKU

 APPLIED MATHEMATICS III

3.1 Periodic Functions; Trigonometric Series

Overview:

In this section, we are going to consider the definition of periodic, even, odd, orthogonal
functions and trigonometric series with examples.

Section Objectives:

At the end of this subtopic, students will be able to:

 Define periodic functions, even and odd functions

 Define and understand about orthogonal functions and trigonometric series.

Definition(periodic functions) A function f (x) is called a periodic function if f ( x) is defined

for all real x, except possibly at some points, and if there is some positive number p, called a
period of f ( x) ,
Such that
𝑓 (𝑥 + 𝑝) = 𝑓 (𝑥) 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑥

The smallest positive period is often called the fundamental period. The graph of a periodic
function has the characteristic that it can be obtained by periodic repetition of its graph in any
interval of length p as shown in the figure 1 bellow

Fig.3.1. Periodic function of period p

Familiar periodic functions are the cosine, sine, tangent, and cotangent with fundamental periods
p= 2𝜋, 2𝜋, 𝜋 𝑎𝑛𝑑 𝜋 𝑟𝑒𝑠𝑝𝑒𝑐𝑡𝑖𝑣𝑒𝑙𝑦. Examples of functions that are not periodic are 𝑥, 𝑥 2 , 𝑥 3 , 𝑒 𝑥 ,
𝑐𝑜𝑠ℎ𝑥 𝑎𝑛𝑑 𝑙𝑛𝑥 .

Badri A, Moges B. and Teklebrhan B. 60 AKU

 APPLIED MATHEMATICS III

If f (x) has period p, it also has the period2𝑝 because by the above definition of a periodic
function

𝑓(𝑥 + 2𝑝) = 𝑓((𝑥 + 𝑝) + 𝑝) = 𝑓(𝑥 + 𝑝) = 𝑓(𝑥) 𝑒𝑡𝑐

Thus, for any integer 𝑛 = 1,2,3, . , . ,. 𝑓(𝑥 + 𝑛𝑝) = 𝑓(𝑥) 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑥

Example 3.1:
State the period for each of the following functions

𝑎) 𝑓(𝑥) = 𝑐𝑜𝑠4𝑥 𝑏) 𝑓(𝑥) = 𝑠𝑖𝑛3𝑥

Solution a) Since cosine is a periodic function of period 𝑝 = 2𝜋,we have

𝜋 𝜋
𝑓(𝑥) = 𝑐𝑜𝑠4𝑥 = cos(4𝑥 + 2𝜋) = cos(4(𝑥 + 2 ) ) = 𝑓(𝑥 + 2 )

𝜋 𝜋
⇒ 𝑓(𝑥 + 2 ) = 𝑓(𝑥) ⇒ P = is the period of 𝑓(𝑥)
2

b) Since the sine is a periodic function of period 𝑝 = 2𝜋,

2𝜋 2𝜋
𝑓(𝑥) = 𝑠𝑖𝑛3𝑥 = sin(3𝑥 + 2𝜋) = sin(3(𝑥 + ) ) = 𝑓(𝑥 + )
3 3

2𝜋 2𝜋
⇒ 𝑓(𝑥 + ) = 𝑓(𝑥) ⇒ P = is the period of 𝑓(𝑥)
3 3

Definition (Even and odd functions) a function 𝑓(𝑥) is called an Even function if
𝑓(−𝑥) = 𝑓(𝑥) 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑥 𝑖𝑛 𝑡ℎ𝑒 𝑑𝑜𝑚𝑎𝑖𝑛 𝑜𝑓 𝑓 ; 𝑖𝑡 𝑖𝑠 𝑐𝑎𝑙𝑙𝑒𝑑 𝑎𝑛 Odd function if
𝑓(−𝑥) = −𝑓(𝑥) for all x in the domain of f .

If 𝑓 and 𝑔 are both odd functions, and both even functions; then the product 𝑓𝑔 is even; and if
one is even and the other is odd then the product is odd. More over even functions are symmetric
about the y- axis where as odd functions are symmetric about the origin.

𝑙 𝑙
Note: 1.for any even function 𝑓(𝑥), ∫−𝑙 𝑓(𝑥) 𝑑𝑥 = 2∫0 𝑓(𝑥)dx
𝑙
2. for any odd function 𝑓(𝑥), ∫−𝑙 𝑓(𝑥) 𝑑𝑥 = 0

Example 3.2: Show that 𝑓(𝑥) = 𝑠𝑖𝑛ℎ(𝑥) is an odd function

Solution To show that a function 𝑓 is odd, we need to evaluate it at – 𝑥 and show that this is the
same as the value of – 𝑓 at 𝑥; that is

Badri A, Moges B. and Teklebrhan B. 61 AKU

 APPLIED MATHEMATICS III

𝑒 (−𝑥) −𝑒 −(−𝑥) 𝑒 −𝑥 −𝑒 𝑥 𝑒 𝑥 −𝑒 −𝑥
𝑓(−𝑥) = 𝑆𝑖𝑛ℎ(−𝑥) = = = −( ) = − sinh(𝑥) = −𝑓(𝑥)
2 2 2
⟹ 𝑓(−𝑥) = −𝑓(𝑥)
Hence 𝑓 is odd function

Definition (Orthogonal functions) Two functions 𝑓(𝑥) 𝑎𝑛𝑑 𝑔(𝑥) are said to be Orthogonal on
the interval [-l,-l] if
𝑙
∫−𝑙 𝑓(𝑥) 𝑔(𝑥)𝑑𝑥 = 0

Example 3.3: show that 𝑓(𝑥)= 𝑠𝑖𝑛𝑥 and 𝑔(𝑥) = 𝑐𝑜𝑠2𝑥 are orthogonal on [−𝜋, 𝜋]

𝜋 𝜋 1
Solution: ∫−𝜋 𝑠𝑖𝑛𝑥 𝑐𝑜𝑠2𝑥𝑑𝑥 = ∫−𝜋 2 (sin(1 − 2) 𝑥 + cos(1 + 2)𝑥)𝑑𝑥
𝜋
1
=∫ (sin(−1) 𝑥 + cos(3)𝑥)𝑑𝑥
−𝜋 2
𝜋
1
= −∫ (sin𝑥 + cos 3𝑥)𝑑𝑥
−𝜋 2
1 𝜋
= − ∫ (sin 𝑥 + cos 3𝑥)𝑑𝑥
2 −𝜋
1 𝜋 1
= − ∫ sin𝑥𝑑𝑥 + − ∫ cos3x 𝑑𝑥
2 −𝜋 2

1 𝑠𝑖𝑛3𝑥 𝜋 1 𝑠𝑖𝑛3𝜋 𝑠𝑖𝑛3(−𝜋)

= 2 [𝑐𝑜𝑠𝑥 − ] = [(𝑐𝑜𝑠𝜋 − ) – (𝑐𝑜𝑠(−𝜋) − ]
6 −𝜋 2 6 6

1 𝑠𝑖𝑛3𝜋 𝑠𝑖𝑛3(−𝜋) 1 1
[(𝑐𝑜𝑠𝜋 − ) – (𝑐𝑜𝑠(−𝜋) − ] = 2 [(−1 − 0)– (−1 − 0] = [(−1 − 0)– (−1 − 0)]
2 6 6 2

1
= (−1 + 1) = 0
2
Hence 𝑓 and 𝑔 are orthogonal functions.

Definition (trigonometric system)

The system 1,cos𝑥, 𝑠𝑖𝑛𝑥, 𝑐𝑜𝑠2𝑥, 𝑠𝑖𝑛2𝑥, 𝑐𝑜𝑠3𝑥, 𝑠𝑖𝑛3𝑥, . , . , . , 𝑐𝑜𝑠𝑛𝑥 , 𝑠𝑖𝑛𝑛𝑥 is called trigonometric
system

Theorem (Orthogonality of the Trigonometric System)

The trigonometric system is orthogonal on the interval−𝜋 ≤ 𝑥 ≤ 𝜋 (hence also on 0 ≤ 𝑥 ≤ 2𝜋
or any other interval of length because of periodicity); that is, the integral of the product of any
two functions in the Trigonometric System over that interval is 0, so that for any integers n and
m,

Badri A, Moges B. and Teklebrhan B. 62 AKU

 APPLIED MATHEMATICS III

𝜋 𝜋, 𝑛 = 𝑚 𝜋
𝑎) ∫−𝜋𝑐𝑜𝑠𝑛𝑥 𝑐𝑜𝑐𝑠𝑚𝑥𝑑𝑥 = { d) ∫−𝜋𝑐𝑜𝑠𝑛𝑥 𝑑𝑥 = 0
0, 𝑛 ≠ 𝑚

𝜋 𝜋, 𝑛 = 𝑚 𝜋
b). ∫−𝜋 𝑠𝑖𝑛𝑛𝑥 𝑠𝑖𝑛𝑚𝑥𝑑𝑥 = { e) ∫−𝜋𝑐𝑜𝑠𝑛𝑥 𝑑𝑥 = 0
0, 𝑛 ≠ 𝑚

𝜋 0, 𝑛 ≠ 𝑚
c) ∫−𝜋 𝑠𝑖𝑛𝑛𝑥 𝑐𝑜𝑠𝑚𝑥𝑑𝑥 = {
0, 𝑛 = 𝑚
Proof
a) we use the trigonometric identities to prove this Orthogonality that is
For 𝑛 ≠ 𝑚
𝜋 𝜋
1
∫ 𝑐𝑜𝑠𝑛𝑥 𝑐𝑜𝑐𝑠𝑚𝑥𝑑𝑥 = ∫ [cos(𝑛 − 𝑚)𝑥 + 𝑐𝑜𝑠(𝑛 + 𝑚)𝑥]𝑑𝑥
−𝜋 −𝜋 2
1 𝜋 1 𝜋
= ∫ cos(𝑛 − 𝑚) 𝑥 + ∫ 𝑐𝑜𝑠(𝑛 + 𝑚)𝑥𝑑𝑥
2 −𝜋 2 –𝜋
1 𝜋 1 𝜋
= 2(𝑛−𝑚) [sin(n − m)𝑥 ] + [sin(𝑛 + 𝑚) 𝑥 ]
−𝜋 2(𝑛+𝑚) −𝜋
1 1
= 2(𝑛−𝑚) [sin(n − m) 𝜋 − sin(n − m)(− 𝜋)] + 2(𝑛+𝑚) [sin(n + m) 𝜋 − sin(n + m)(− 𝜋)]
1 1
= (0 − 0) + 2(𝑛+𝑚) (0 − 0)
2(𝑛−𝑚)
1 1
= (0) + 2(𝑛+𝑚) (0) = 0 + 0 = 0
2(𝑛−𝑚)

𝐻𝑒𝑛𝑐𝑒 𝑐𝑜𝑠𝑛𝑥 𝑎𝑛𝑑 𝑐𝑜𝑠𝑚𝑥 𝑎𝑟𝑒 𝑜𝑟𝑡ℎ𝑜𝑔𝑜𝑛𝑎𝑙 𝑤ℎ𝑒𝑛 𝑛 ≠ 𝑚

For n=m,
𝜋 𝜋 𝜋 𝜋 1
∫−𝜋𝑐𝑜𝑠𝑛𝑥 𝑐𝑜𝑐𝑠𝑚𝑥𝑑𝑥 = ∫−𝜋𝑐𝑜𝑠𝑛𝑥 𝑐𝑜𝑐𝑠𝑛𝑥𝑑𝑥 = ∫−𝜋 𝑐𝑜𝑠 2 𝑛𝑥 𝑑𝑥 = ∫−𝜋 2 (1 + 𝑐𝑜𝑠2𝑛𝑥)𝑑𝑥

1 𝑠𝑖𝑛2𝑛𝑥 𝜋 1 𝑠𝑖𝑛2𝑛𝜋 𝑠𝑖𝑛2𝑛(−𝜋) 1 𝑠𝑖𝑛2𝑛𝜋 𝑠𝑖𝑛2𝑛𝜋

= 2[𝑥+ ] = 2[(𝜋+ 2𝑛 ) − (−𝜋+ 2𝑛 )] = 2[(𝜋+ 2𝑛 ) − (−𝜋 − 2𝑛 )]
2𝑛 −𝜋

1 1 2𝜋
= 2[(𝜋+0) − (−𝜋 − 0)] = 2 (𝜋+𝜋) = =𝜋
2
Similarly b and c can be proved by the same fashion

Definition (trigonometric series) any series that can be expressed in the form of

𝑎𝑜 + 𝑎1 𝑐𝑜𝑠𝑥 + 𝑏1 𝑠𝑖𝑛𝑥 + 𝑎2 𝑐𝑜𝑠2𝑥 + 𝑏2 𝑠𝑖𝑛2𝑥 + ⋯ + = ∑∞

𝑛=1(𝑎𝑛 𝑐𝑜𝑠𝑛𝑥 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥)

Where𝑎𝑜 , 𝑎1 , 𝑏1 , 𝑎2 , 𝑏2 … are constants, called the coefficients of the series and each term has
the period 2𝜋 𝑖𝑠 𝑐𝑎𝑙𝑙𝑒𝑑 𝑎 𝒕𝒓𝒊𝒈𝒐𝒏𝒐𝒎𝒆𝒕𝒓𝒊𝒄 𝒔𝒆𝒓𝒊𝒆𝒔.

Badri A, Moges B. and Teklebrhan B. 63 AKU

 APPLIED MATHEMATICS III

Note: If the coefficients𝑎𝑡𝑟𝑖𝑔𝑜𝑛𝑜𝑚𝑒𝑡𝑟𝑖𝑐 𝑠𝑒𝑟𝑖𝑒𝑠are such that the series converges, its sum will
be a function of period 2𝜋.

Exercises 3.1

1. In each of the following find the fundamental period 𝑝

2𝜋𝑛𝑥 2𝜋𝑥
(𝑎) 𝑓(𝑥) = 𝑐𝑜𝑠𝜋𝑥 (𝑏) 𝑓(𝑥) = 𝑠𝑖𝑛𝑛𝑥 (𝑐). 𝑔(𝑥) = 𝑐𝑜𝑠 𝑘
(𝑑) ℎ(𝑥) = 𝑠𝑖𝑛 𝑘

2. If 𝑓(𝑥) 𝑎𝑛𝑑 𝑔(𝑥) have period 𝑝, show that ℎ(𝑥) = 𝑎𝑓(𝑥) + 𝑏𝑔(𝑥) (𝑎, 𝑏, constants) has period 𝑝.

3. Show that 𝑓 = constant is periodic with any period but has no fundamental period.

4. Determine whether the following functions are even or odd functions;

(a)𝑓(𝑥) = 𝑐𝑜𝑠ℎ(𝑥) (𝑏) 𝑔(𝑥) = 𝑡𝑎𝑛(𝑥) (𝑐) 𝑘(𝑥) = 𝑥 2 𝑐𝑜𝑠(𝑥) (𝑑) ℎ(𝑥) = 𝑥|𝑥|

5) Show that if 𝑓 is an even function and 𝑔 is odd, then the product 𝑓. 𝑔 is an odd function

5. Show that the following functions are orthogonal on[−𝜋, 𝜋];

(𝑎)𝑓(𝑥) = 𝑠𝑖𝑛4𝑥, 𝑔(𝑥) = 𝑐𝑜𝑠𝑥 (𝑏) 𝑓(𝑥) = −6, 𝑔(𝑥) = 𝑠𝑖𝑛3𝑥

(𝑐) = 𝑠𝑖𝑛2𝑥, 𝑔(𝑥) = 𝑠𝑖𝑛3𝑥

−𝑎
6. Show that ∫𝑎 𝑥 5 𝑐𝑜𝑠𝑥𝑑𝑥 = 0, what would you conclude from this?

Badri A, Moges B. and Teklebrhan B. 64 AKU

 APPLIED MATHEMATICS III

3.2 Fourier series and Fourier Integrals

Overview:

In this section, we are going to consider the definition of Fourier series together with some
examples

Section Objectives:

At the end of this subtopic, students will be able to:

 Learn and understand about Fourier series of periodic functions;

 Develop Fourier series representation of periodic functions of arbitrary period.
 Define and understand Fourier integral representations of non-periodic functions

Definition (Fourier series): If 𝑓 is a periodic function of period 2𝜋 and integrable on the interval
(−𝜋, 𝜋), then the Fourier series of 𝑓 is defined as:
𝑓(𝑥) = 𝑎𝑜 + ∑∞
𝑛=1(𝑎𝑛 𝑐𝑜𝑠𝑛𝑥 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥) (1)
where 𝑎𝑜 ,𝑎𝑛 , 𝑎𝑛𝑑 𝑏𝑛 𝑎𝑟𝑒 𝑐𝑎𝑙𝑙𝑒𝑑 𝑡ℎ𝑒 Fourier coefficients of 𝑓 and are given by the Euler’s Formula
1 𝜋
a) 𝑎𝑜 = 2𝜋 ∫−𝜋 𝑓(𝑥) 𝑑𝑥

1 𝜋
b) 𝑎𝑛 = 𝜋 ∫−𝜋 𝑓(𝑥)𝑐𝑜𝑠𝑛𝑥 𝑑𝑥 𝑛 = 1,2, …

1 𝜋
c) 𝑏𝑛 = 𝜋 ∫−𝜋 𝑓(𝑥)𝑠𝑖𝑛𝑛𝑥 𝑑𝑥 𝑛 = 1,2, …
Proof:

Verification

a) For 𝑎𝑜 integrating on both sides of (1) from – 𝜋 𝑡𝑜 𝜋 we get

𝜋 𝜋
∫−𝜋 𝑓(𝑥) 𝑑𝑥 = ∫–𝜋 [𝑎𝑜 + ∑∞ 𝑛=1(𝑎𝑛 𝑐𝑜𝑠𝑛𝑥 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥)]𝑑𝑥
𝜋 𝜋
= ∫−𝜋 𝑎𝑜 𝑑𝑥 + ∫−𝜋 ∑∞
𝑛=1(𝑎𝑛 𝑐𝑜𝑠𝑛𝑥 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥)𝑑𝑥
𝜋 𝜋
= 𝑎𝑜 ∫−𝜋 1𝑑𝑥 + ∑∞𝑛=1 ∫–𝜋(𝑎𝑛 𝑐𝑜𝑠𝑛𝑥 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥)d𝑥
𝜋 ∞ 𝜋 𝜋
]
= [𝑎𝑜 𝑥 – 𝜋 + ∑ 𝑛=1[ ∫–𝜋
𝑎 𝑛 𝑐𝑜𝑠𝑛𝑥𝑑𝑥 + ∫−𝜋 𝑛
𝑏 𝑠𝑖𝑛𝑛𝑥𝑑𝑥]
𝜋 𝜋 𝜋
= 𝑎𝑜 [𝑥]– 𝜋 + ∑∞
𝑛=1[𝑎𝑛 ∫–𝜋 𝑐𝑜𝑠𝑛𝑥𝑑𝑥 + 𝑏𝑛 ∫−𝜋 𝑠𝑖𝑛𝑛𝑥𝑑𝑥]

Badri A, Moges B. and Teklebrhan B. 65 AKU

 APPLIED MATHEMATICS III

𝑎𝑛 𝜋 𝑏𝑛 𝜋
= 𝑎𝑜 [𝜋 − (−𝜋)] + ∑∞
𝑛=1[ 𝑛 [𝑠𝑖𝑛𝑛𝑥]– 𝜋 − 𝑛 [ 𝑐𝑜𝑠𝑛𝑥]– 𝜋 ]

𝑎 𝑏𝑛 𝜋
= 𝑎 𝑜 2 𝜋 + ∑∞ 𝑛
𝑛=1[ 𝑛 [𝑠𝑖𝑛𝑛𝜋 + 𝑠𝑖𝑛𝜋] − [ 𝑐𝑜𝑠𝑛𝜋 − 𝑐𝑜𝑠𝑛𝜋]– 𝜋 ]
𝑛
𝑎𝑛 𝑏𝑛
= 𝑎 𝑜 2 𝜋 + ∑∞
𝑛=1[ 𝑛 (0) − (0)] ]
𝑛
= 𝑎𝑜 2 𝜋 + 0
= 𝑎𝑜 2 𝜋
𝜋
⇒ ∫−𝜋 𝑓(𝑥) 𝑑𝑥 = 𝑎𝑜 2 𝜋 𝟏 𝝅
𝒂𝒐 = ∫ 𝒇(𝒙) 𝒅𝒙
1 𝜋 𝟐𝝅 −𝝅
⇒ 𝑎𝑜 = 2𝜋 ∫−𝜋 𝑓(𝑥) 𝑑𝑥
1 𝜋
Hence 𝑎𝑜 = 2𝜋 ∫−𝜋 𝑓(𝑥) 𝑑𝑥
Proof: for 𝑎𝑛
Multiplying on both sides of (1) by 𝑐𝑜𝑠𝑚𝑥 for any fixed positive integer 𝑚 and integrating on
both sides from – 𝜋 𝑡𝑜 𝜋 we get
𝜋 𝜋
∫−𝜋 𝑓(𝑥)𝑐𝑜𝑠𝑚𝑥 𝑑𝑥 = ∫–𝜋 [𝑎𝑜 + ∑∞ 𝑛=1(𝑎𝑛 𝑐𝑜𝑠𝑛𝑥 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥)]𝑐𝑜𝑠𝑚𝑥𝑑𝑥
𝜋
= ∫–𝜋 [𝑎𝑜 𝑐𝑜𝑠𝑚𝑥 + ∑∞
𝑛=1(𝑎𝑛 𝑐𝑜𝑠𝑛𝑥𝑐𝑜𝑠𝑚𝑥 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥𝑐𝑜𝑠𝑚𝑥)]𝑑𝑥
𝜋 𝜋 𝜋
= ∫–𝜋 𝑎𝑜 𝑐𝑜𝑠𝑚𝑥𝑑𝑥 + ∑∞
𝑛=1 (∫–𝜋 𝑎𝑛 𝑐𝑜𝑠𝑛𝑥𝑐𝑜𝑠𝑚𝑥𝑑𝑥 + ∫–𝜋 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥𝑐𝑜𝑠𝑚𝑥𝑑𝑥 )
𝜋 𝜋 𝜋
= 𝑎𝑜 ∫–𝜋 𝑐𝑜𝑠𝑚𝑥𝑑𝑥 + ∑∞𝑛=1 (𝑎𝑛 ∫–𝜋 𝑐𝑜𝑠𝑛𝑥𝑐𝑜𝑠𝑚𝑥𝑑𝑥 + 𝑏𝑛 ∫–𝜋 𝑠𝑖𝑛𝑛𝑥𝑐𝑜𝑠𝑚𝑥𝑑𝑥 )
𝑎 𝜋
= 𝑚𝑜 [𝑠𝑖𝑛𝑚𝑥]– 𝜋 + ∑∞𝑛=1(𝑎𝑛 𝜋 + 𝑏𝑛 (0) ) (by the Orthogonality theorem above)
𝑎
= 𝑚𝑜 [𝑠𝑖𝑛𝑚𝜋 + 𝑠𝑖𝑛𝑚𝜋 ] + 𝑎𝑛 𝜋 (𝑚 = 𝑛 𝑖𝑠 𝑎 𝑓𝑖𝑥𝑒𝑑 𝑖𝑛𝑡𝑒𝑔𝑒𝑟)
𝑎𝑜
= [0 + 0] + 𝑎𝑛 𝜋 = 𝑎𝑛 𝜋
𝑚
𝜋
⇒ ∫−𝜋 𝑓(𝑥)𝑐𝑜𝑠𝑚𝑥 𝑑𝑥 = 𝑎𝑛 𝜋 𝟏 𝝅
𝒂𝒏 = ∫−𝝅 𝒇(𝒙)𝒄𝒐𝒔𝒏𝒙 𝒅𝒙
1 𝜋 𝝅
⇒ 𝑎𝑛 = 𝜋 ∫−𝜋 𝑓(𝑥)𝑐𝑜𝑠𝑚𝑥 𝑑𝑥
1 𝜋
⇒ 𝑎𝑛 = ∫−𝜋 𝑓(𝑥)𝑐𝑜𝑠𝑛𝑥 𝑑𝑥 as required
𝜋

1 𝜋
Hence 𝑎𝑛 = 𝜋 ∫−𝜋 𝑓(𝑥)𝑐𝑜𝑠𝑛𝑥 𝑑𝑥
Proof: for 𝑏𝑛
Multiplying on both sides of (1) by 𝑠𝑖𝑛𝑚𝑥 for any fixed positive integer 𝑚 and integrating on
both sides from – 𝜋 𝑡𝑜 𝜋 we get
𝜋 𝜋
∫−𝜋 𝑓(𝑥)𝑠𝑖𝑛𝑚𝑥 𝑑𝑥 = ∫–𝜋 [𝑎𝑜 + ∑∞ 𝑛=1(𝑎𝑛 𝑐𝑜𝑠𝑛𝑥 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥)]𝑠𝑖𝑛𝑚𝑥𝑑𝑥
𝜋
= ∫–𝜋 [𝑎𝑜 𝑠𝑖𝑛𝑚𝑥 + ∑∞
𝑛=1(𝑎𝑛 𝑐𝑜𝑠𝑛𝑥𝑠𝑖𝑛𝑚𝑥 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥𝑠𝑖𝑛𝑚𝑥)]𝑑𝑥
𝜋 𝜋 𝜋
= ∫–𝜋 𝑎𝑜 𝑠𝑖𝑛𝑚𝑥𝑑𝑥 + ∑∞
𝑛=1 (∫–𝜋 𝑎𝑛 𝑐𝑜𝑠𝑛𝑥𝑠𝑖𝑛𝑚𝑥𝑑𝑥 + ∫–𝜋 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥𝑠𝑖𝑛𝑚𝑥𝑑𝑥 )
𝜋 𝜋 𝜋
= 𝑎𝑜 ∫–𝜋 𝑠𝑖𝑛𝑚𝑥𝑑𝑥 + ∑∞
𝑛=1 (𝑎𝑛 ∫–𝜋 𝑐𝑜𝑠𝑛𝑥𝑠𝑖𝑛𝑚𝑥𝑑𝑥 + 𝑏𝑛 ∫–𝜋 𝑠𝑖𝑛𝑛𝑥𝑠𝑖𝑛𝑚𝑥𝑑𝑥 )

= 𝑎𝑜 (0) + ∑∞
𝑛=1(𝑎𝑛 (0) + 𝑏𝑛 𝜋) (By the Orthogonality theorem above)

Badri A, Moges B. and Teklebrhan B. 66 AKU

 APPLIED MATHEMATICS III

= 0 + 𝑏𝑛 𝜋 (𝑚 = 𝑛 𝑖𝑠 𝑎 𝑝𝑜𝑠𝑖𝑡𝑖𝑣𝑒 𝑓𝑖𝑥𝑒𝑑 𝑖𝑛𝑡𝑒𝑔𝑒𝑟)

= 𝑏𝑛 𝜋
𝜋
⇒ ∫−𝜋 𝑓(𝑥)𝑠𝑖𝑛𝑚𝑥 𝑑𝑥 = 𝑏𝑛 𝜋
1 𝜋
⇒ 𝑏𝑛 = 𝜋 ∫−𝜋 𝑓(𝑥)𝑠𝑖𝑛𝑚𝑥 𝑑𝑥 As required
𝜋 𝟏 𝝅
1
Hence 𝑏𝑛 = 𝜋 ∫−𝜋 𝑓(𝑥)𝑠𝑖𝑛𝑚𝑥 𝑑𝑥 𝒃𝒏 = ∫−𝝅 𝒇(𝒙)𝒔𝒊𝒏𝒎𝒙 𝒅𝒙
𝝅

Example 3.4: find the Fourier series representation of the periodic function
−𝑘 𝑖𝑓 − 𝜋 ≤ 𝑥 ≤ 𝜋
𝑓(𝑥) = { 𝑎𝑛𝑑 𝑓(𝑥 + 2𝜋) = 𝑓(𝑥)
𝑘 𝑖𝑓 0 < 𝑥 < 𝜋
Solution: to find the Fourier series representation of 𝑓 we first find the Fourier coefficients
𝑎𝑜 , 𝑎𝑛 𝑎𝑛𝑑 𝑏𝑛 of 𝑓 𝑢𝑠𝑖𝑛𝑔 𝐸𝑢𝑙𝑒𝑟 ′ 𝑠 𝑓𝑜𝑟𝑚𝑢𝑙𝑎𝑠.

1 𝜋 1 0 𝜋
𝑎𝑜 = 2𝜋 ∫−𝜋 𝑓(𝑥) 𝑑𝑥 = [∫ 𝑓(𝑥) 𝑑𝑥 + ∫0 𝑓(𝑥) 𝑑𝑥]
2𝜋 −𝜋
1 0 𝜋
= [∫ −𝑘 𝑑𝑥 + ∫0 𝑘 𝑑𝑥]
2𝜋 −𝜋
−𝑘 0 𝑘 𝜋
= [0 − (−𝜋)] + [𝜋 − 0]
2𝜋 −𝜋 2𝜋 0
−𝑘𝜋 𝑘𝜋
= + = 0
2𝜋 2𝜋
Hence, 𝑎𝑜 = 0
1 𝜋
Similarly, 𝑎𝑛 = 𝑎𝑛 = 𝜋 ∫–𝜋 𝑓(𝑥)𝑐𝑜𝑠𝑛𝑥 𝑑𝑥

1 0 𝜋
= [∫−𝜋 𝑓(𝑥) 𝑐𝑜𝑠𝑛𝑥𝑑𝑥 + ∫0 𝑓(𝑥) 𝑐𝑜𝑠𝑛𝑥𝑑𝑥]
𝜋

1 0 1 𝜋
= ∫ −𝑘 𝑐𝑜𝑠𝑛𝑥𝑑𝑥 + 𝜋 ∫0 𝑘 𝑐𝑜𝑠𝑛𝑥𝑑𝑥
𝜋 −𝜋

𝜋
[𝑠𝑖𝑛𝑥] 0 + [𝑠𝑖𝑛𝑥]
−𝑘 𝑘
= 𝑛𝜋 −𝜋 𝑛𝜋 0
−𝑘 𝑘
= [𝑠𝑖𝑛0 − sin(−𝜋)] + (𝑠𝑖𝑛𝜋 − sin 0)
𝑛𝜋 𝑛𝜋

−𝑘 𝑘
= [𝑠𝑖𝑛0 − sin(−𝜋)] + (𝑠𝑖𝑛𝜋 − sin 0)
𝑛𝜋 𝑛𝜋

−𝑘 𝑘
= (0) + 𝑛𝜋 (0) = 0
𝑛𝜋

Hence , 𝑎𝑛 = 0
1 𝜋
More over, 𝑏𝑛 = ∫−𝜋 𝑓(𝑥)𝑠𝑖𝑛𝑚𝑥 𝑑𝑥
𝜋
1 0 𝜋 1 0 1 𝜋
= [∫−𝜋 −𝑘 𝑠𝑖𝑛𝑛𝑥𝑑𝑥 + ∫0 𝑘𝑠𝑖𝑛𝑛𝑥 𝑑𝑥] = ∫ −𝑘 𝑠𝑖𝑛𝑛𝑥𝑑𝑥 + 𝜋 ∫0 𝑘𝑠𝑖𝑛𝑛𝑥 𝑑𝑥]
𝜋 𝜋 −𝜋

Badri A, Moges B. and Teklebrhan B. 67 AKU

 APPLIED MATHEMATICS III

𝑘 0 −𝑘 𝜋 𝑘 −𝑘
= 𝑛𝜋 [𝑐𝑜𝑠𝑛𝑥] + [𝑐𝑜𝑠𝑛𝑥] = 𝑛𝜋 [𝑐𝑜𝑠0 − cos(−𝑛𝜋)] + 𝑛𝜋 (𝑐𝑜𝑠𝑛𝜋 − 𝑐𝑜𝑠0)
−𝜋 2𝑛𝜋 0
𝑘 −𝑘 𝑘 𝑘
= 𝑛𝜋 [1 − 𝑐𝑜𝑠𝑛𝜋)] + 𝑛𝜋 (𝑐𝑜𝑠𝑛𝜋 − 1) = 𝑛𝜋 [1 − 𝑐𝑜𝑠𝑛𝜋)] + 2𝑛𝜋 (1 − 𝑐𝑜𝑠𝑛𝜋)

2𝑘 2𝑘
= 𝑛𝜋 (1 − 𝑐𝑜𝑠𝑛𝜋) = 𝑛𝜋 (1 − (−1)𝑛 )……. Because 𝑐𝑜𝑠𝑛𝜋 = (−1)𝑛
4𝑘
−1, 𝑓𝑜𝑟 𝑜𝑑𝑑 𝑛
𝑛 2𝑘 , 𝑓𝑜𝑟 𝑜𝑑𝑑 𝑛
Here, 𝑐𝑜𝑠𝑛𝜋 = (−1) = { ⇒ 𝑛𝜋 (1 − (−1)𝑛 ) = {𝑛𝜋
1, 𝑓𝑜𝑟 𝑒𝑣𝑒𝑛 𝑛 0, 𝑓𝑜𝑓 𝑛 𝑒𝑣𝑒𝑛
4𝑘
⇒ 𝑏𝑛 = 𝑛𝜋
4𝑘
Hence, 𝑏𝑛 = 𝑛𝜋
From these, the Fourier series of 𝑓 is

4𝑘
𝑓(𝑥) = 𝑎𝑜 + ∑∞
𝑛=1(𝑎𝑛 𝑐𝑜𝑠𝑛𝑥 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥) = 0 + ∑∞
𝑛=1 (0. 𝑐𝑜𝑠𝑛𝑥 + 𝑛𝜋 𝑠𝑖𝑛𝑛𝑥)

4𝑘 4𝑘 4𝑘 4𝑘
= ∑∞
𝑛=1 𝑛𝜋 𝑠𝑖𝑛𝑛𝑥 = 𝑠𝑖𝑛𝑥 + 0 + 3𝜋 𝑠𝑖𝑛3𝑥 + 0 + 𝑠𝑖𝑛5𝑥 + ⋯
𝜋 5𝜋

4𝑘 4𝑘 4𝑘 4𝑘 4𝑘 4𝑘
= 𝑠𝑖𝑛𝑥 + 𝑠𝑖𝑛3𝑥 + 𝑠𝑖𝑛5𝑥 + ⋯ = 𝑠𝑖𝑛𝑥 + 𝑠𝑖𝑛3𝑥 + 𝑠𝑖𝑛5𝑥 + ⋯
𝜋 3𝜋 5𝜋 𝜋 3𝜋 5𝜋

4𝑘 1 1
= (𝑠𝑖𝑛𝑥 + 3 𝑠𝑖𝑛3𝑥 + 𝑠𝑖𝑛5𝑥 + ⋯ )
𝜋 5

4𝑘 1 1
Hence, 𝑓(𝑥) = (𝑠𝑖𝑛𝑥 + 3 𝑠𝑖𝑛3𝑥 + 𝑠𝑖𝑛5𝑥 + ⋯ )
𝜋 5

Fig 3.2 Periodic rectangular wave

0, 𝐼𝑓 − 𝜋 ≤ 𝑥 ≤ 0
Example 3.5: Find the Fourier series representation of 𝑓(𝑥) = { so that 𝑓 is
1, 𝐼𝑓0 ≤ 𝑥 ≤ 𝜋
periodic with period 2𝜋 and 𝑓(𝑥 + 2𝜋) = 𝑓(𝑥)

Solution: to find a Fourier series representation of this periodic function we first find the Fourier
coefficients 𝑎𝑜 , 𝑎𝑛 𝑎𝑛𝑑 𝑏𝑛 .
1 𝜋
That is by Euler’s formulas, 𝑎𝑜 = 2𝜋 ∫−𝜋 𝑓(𝑥) 𝑑𝑥

Badri A, Moges B. and Teklebrhan B. 68 AKU

 APPLIED MATHEMATICS III

1 0 𝜋 1 0 𝜋 1 𝜋
= 2𝜋 [∫−𝜋 𝑓(𝑥) 𝑑𝑥 + ∫0 𝑓(𝑥) 𝑑𝑥] = 2𝜋 [∫−𝜋 0 𝑑𝑥 + ∫0 1 𝑑𝑥] = ∫ 1 𝑑𝑥]
2𝜋 0

1 𝜋 1 𝜋 1
[𝑥] = (𝜋 − 0) = =2
2𝜋 0 2𝜋 2𝜋

1
Hence, 𝑎𝑜 = 2

1 𝜋 1 0 𝜋
Similarly, 𝑎𝑛 = 𝜋 ∫–𝜋 𝑓(𝑥)𝑐𝑜𝑠𝑛𝑥 𝑑𝑥 = 𝜋 [∫−𝜋 𝑓(𝑥) 𝑐𝑜𝑠𝑛𝑥𝑑𝑥 + ∫0 𝑓(𝑥) 𝑐𝑜𝑠𝑛𝑥𝑑𝑥]

1 0 𝜋 1 𝜋 1 𝜋
= [∫−𝜋 0. 𝑐𝑜𝑠𝑛𝑥𝑑𝑥 + ∫0 1. 𝑐𝑜𝑠𝑛𝑥𝑑𝑥] = ∫0 1. 𝑐𝑜𝑠𝑛𝑥𝑑𝑥] = ∫0 𝑐𝑜𝑠𝑛𝑥𝑑𝑥
𝜋 𝜋 𝜋

𝑘 𝜋 𝑘 𝑘 𝑘
= 𝑛𝜋 [𝑠𝑖𝑛𝑛𝑥] = 𝑛𝜋 (𝑠𝑖𝑛𝑛𝜋 − 𝑠𝑖𝑛0) = 𝑛𝜋 (0 − 0) = 𝑛𝜋 (0) = 0
0
Hence, 𝑎𝑛 = 0
1 𝜋 1 0 𝜋 1 𝜋
Now, 𝑏𝑛 = 𝜋 ∫−𝜋 𝑓(𝑥)𝑠𝑖𝑛𝑛𝑥 𝑑𝑥 = 𝜋 [∫−𝜋 0. 𝑠𝑖𝑛𝑛𝑥𝑑𝑥 + ∫0 1. 𝑠𝑖𝑛𝑛𝑥 𝑑𝑥] = 𝜋 ∫0 1. 𝑠𝑖𝑛𝑛𝑥 𝑑𝑥

1 𝜋 −1 𝜋 −1 −1
∫ 𝑠𝑖𝑛𝑛𝑥 𝑑𝑥 = 𝑛𝜋 [𝑐𝑜𝑠𝑛𝑥] = 𝑛𝜋 (𝑐𝑜𝑠𝑛𝜋 − 𝑐𝑜𝑠0) = (𝑐𝑜𝑠𝑛𝜋 − 1)
𝜋 0 0 𝑛𝜋

−1, 𝑓𝑜𝑟 𝑜𝑑𝑑 𝑛

Bur, remember that 𝑐𝑜𝑠𝑛𝜋 = (−1)𝑛 = {
1, 𝑓𝑜𝑟 𝑒𝑣𝑒𝑛 𝑛
2
−1 , 𝑓𝑜𝑟 𝑜𝑑𝑑 𝑛
⇒ 𝒃𝒏 = (𝑐𝑜𝑠𝑛𝜋 − 1) = {𝑛𝜋
𝑛𝜋
0, 𝑓𝑜𝑟 𝑒𝑣𝑒𝑛 𝑛
2
, 𝑓𝑜𝑟 𝑜𝑑𝑑 𝑛
Hence, 𝒃𝒏 = {𝑛𝜋
0, 𝑓𝑜𝑟 𝑒𝑣𝑒𝑛 𝑛

From these, the Fourier series of 𝑓 is 𝑓(𝑥) = 𝑎𝑜 + ∑∞

𝑛=1(𝑎𝑛 𝑐𝑜𝑠𝑛𝑥 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑥)
1 2 1 2 1 2 1
= + ∑∞
𝑛=1 (0. 𝑐𝑜𝑠𝑛𝑥 + 𝑛𝜋 𝑠𝑖𝑛𝑛𝑥) = + ∑∞
𝑛=1 𝑛𝜋 𝑠𝑖𝑛𝑛𝑥 = + ∑∞
𝑛=1 𝑠𝑖𝑛𝑛𝑥
2 2 2 𝜋 𝑛
1 2 1 1
= + (𝑠𝑖𝑛𝑥 + 0 + 3 𝑠𝑖𝑛3𝑥 + 0 + 5 𝑠𝑖𝑛5𝑥 + ⋯ )
2 𝜋
1 2 1 1
Hence, 𝑓(𝑥) = + (𝑠𝑖𝑛𝑥 + 0 + 3 𝑠𝑖𝑛3𝑥 + 0 + 5 𝑠𝑖𝑛5𝑥 + ⋯ )
2 𝜋

Example 3.6: Obtain the Fourier series of 𝑓(𝑥) = 𝑥 2 over the interval (−𝜋, 𝜋) where

𝑓(𝑥 + 2𝜋) = 𝑓(𝑥)

1 𝜋 1 𝜋
Solution: we first find the Fourier coefficients, 𝑎𝑜 = 2𝜋 ∫−𝜋 𝑓(𝑥) 𝑑𝑥 = 2 2𝜋 ∫0 𝑥 2 𝑑𝑥

1 𝜋 𝜋 𝜋
= 𝜋 ∫0 𝑥 2 𝑑𝑥 (Because f is even function 𝑎𝑛𝑑 ∫−𝜋 𝑓(𝑥) 𝑑𝑥 = 2 ∫0 𝑓(𝑥) 𝑑𝑥)

Badri A, Moges B. and Teklebrhan B. 69 AKU

 APPLIED MATHEMATICS III

1 𝑥3 𝜋 1 𝜋 1 𝜋2
= 𝜋[3] = [𝑥 3 ] = (𝜋 3 − 03 ) =
0 3𝜋 0 3𝜋 3

𝜋2
Hence, 𝑎𝑜 = 3

1 𝜋 1 𝜋 1 𝜋
Similarly, 𝑎𝑛 = 𝜋 ∫–𝜋 𝑓(𝑥)𝑐𝑜𝑠𝑛𝑥 𝑑𝑥 = 𝜋 ∫–𝜋 𝑥 2 𝑐𝑜𝑠𝑛𝑥 𝑑𝑥 = 2 𝜋 ∫0 𝑥 2 𝑐𝑜𝑠𝑛𝑥 𝑑𝑥 (∴ 𝑓 𝑖𝑠 𝑒𝑣𝑒𝑛)

𝟒(−𝟏)𝒏
Now using integration by parts we have, 𝒂𝒏 = 𝒏𝟐

1 𝜋 1 𝜋
Now, 𝑏𝑛 = ∫−𝜋 𝑓(𝑥)𝑠𝑖𝑛𝑛𝑥 𝑑𝑥 = ∫−𝜋 𝑥 2 𝑠𝑖𝑛𝑛𝑥 𝑑𝑥 = 0 (∴ 𝑥 2 𝑠𝑖𝑛𝑛𝑥 𝑖𝑠 𝑜𝑑𝑑 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛 )
𝜋 𝜋

Hence, 𝑏𝑛 = 0

3.2.1 FOURIER SERIES OF FUNCTIONS OF PERIOD 𝑷 = 𝟐𝑳

So far we have considered the Fourier series expansion of functions with period 2𝜋.in many
application ,we need to find the Fourier series expansion of periodic functions with arbitrary
period ,say 2𝑙. The transition from period 𝑃 = 2𝑙 to period 𝑃 = 2𝜋 is quite simple and involves
only proportional change of scale.

Consider the periodic function f(x) with period 2l in(−𝑙, 𝑙) to change the problem to period 2𝜋,
𝜋𝑥 𝑙𝑣
Set, = , which gives 𝑥 = , 𝑡ℎ𝑢𝑠 𝑥 = ± 𝑙 corresponds to 𝑣 = ± 𝜋 and the function 𝑓(𝑥) of
𝑙 𝜋
period 2l in(−𝑙, 𝑙) may be regarded as function 𝑔(𝑣) of period 2𝜋 in (−𝜋, 𝜋).

Hence,
𝑔(𝑣) = 𝑎𝑜 + ∑∞
𝑛=1(𝑎𝑛 𝑐𝑜𝑠𝑛𝑣 + 𝑏𝑛 𝑠𝑖𝑛𝑛𝑣) (1)

Where 𝑎𝑜 ,𝑎𝑛 , 𝑎𝑛𝑑 𝑏𝑛 𝑎𝑟𝑒 𝑐𝑎𝑙𝑙𝑒𝑑 𝑡ℎ𝑒 Fourier coefficients of 𝑓 and are given by the Euler’s
Formula

1 𝜋
a) 𝑎𝑜 = 2𝜋 ∫−𝜋 𝑔(𝑣) 𝑑𝑣

1 𝜋
b) 𝑎𝑛 = 𝜋 ∫−𝜋 𝑔(𝑣)𝑐𝑜𝑠𝑛𝑣 𝑑𝑣 𝑛 = 1,2, …

1 𝜋
c) 𝑏𝑛 = 𝜋 ∫−𝜋 𝑔(𝑣)𝑠𝑖𝑛𝑛𝑣 𝑑𝑣 𝑛 = 1,2, …

𝜋𝑥
Making the inverse substitution, 𝑣 = and 𝑔(𝑣) = 𝑓(𝑥) in the above we obtain the Fourier
𝑙
series expansion

Badri A, Moges B. and Teklebrhan B. 70 AKU

 APPLIED MATHEMATICS III

𝑛𝜋𝑥 𝑛𝜋𝑥
𝑓(𝑥) = 𝑎𝑜 + ∑∞
𝑛=1(𝑎𝑛 𝑐𝑜𝑠 + 𝑏𝑛 𝑠𝑖𝑛 ) (2)
𝑙 𝑙
With coefficients,

1 𝑙
a) 𝑎𝑜 = 2𝑙 ∫−𝑙 𝑓(𝑥) 𝑑𝑥

1 𝑙 𝑛𝜋𝑥
b) 𝑎𝑛 = 𝑙 ∫−𝑙 𝑓(𝑥)𝑐𝑜𝑠 𝑑𝑥 𝑛 = 1,2, …
𝑙

1 𝑙 𝑛𝜋𝑥
c) 𝑏𝑛 = ∫−𝑙 𝑓(𝑥)𝑠𝑖𝑛 𝑑𝑥 𝑛 = 1,2, …
𝑙 𝑙
Note: we may replace the interval of integration by any interval of length 𝑃 = 2𝑙, say by the
interval (0,2𝑙)

𝑥, 𝐼𝑓 − 1 ≤ 𝑥 ≤ 0
Example 3.7: find the Fourier series for the function 𝑓(𝑥) = { where
𝑥 + 2, 𝐼𝑓0 ≤ 𝑥 ≤ 1

𝑓(𝑥 + 2) = 𝑓(𝑥)

Solution: The function is periodic with period 2 and 𝑙 = 2

1 𝜋
The Fourier coefficients are 𝑎𝑜 = 2𝑙 ∫−𝜋 𝑓(𝑥) 𝑑𝑥 2

1 𝑙 1 1
= 2𝑙 ∫−𝑙 𝑓(𝑥) 𝑑𝑥 = 2 ∫−1 𝑓(𝑥) 𝑑𝑥

1 0 1
= 2 [∫−1 𝑥 𝑑𝑥 + ∫0 (𝑥 + 2) 𝑑𝑥] -1 1
1 0 1
= 2 [∫−1 𝑥 𝑑𝑥 + ∫0 (𝑥 + 2) 𝑑𝑥]

1 𝑥2 0 1 𝑥2 1
= 2[2] + 2 [ 2 + 2𝑥] Figure3.3 graph of 𝑓
−1 0
1 02 (−1)2 1 12 02
= ( − ) + 2 ( 2 + 2 − ( 2 + 2(0)))
2 2 2

1 1 1 1 1
= 2(0 − 2) + 4 + 1 = − 4 + 4 + 1 = 1

Hence 𝑎𝑜 = 0
1 0 1
𝑎𝑛 = ∫−1 𝑓(𝑥)𝑐𝑜𝑠𝑛𝜋𝑥𝑑𝑥 = ∫−1 𝑥𝑐𝑜𝑠𝑛𝜋𝑥𝑑𝑥 + ∫0 (𝑥 + 2)𝑐𝑜𝑠𝑛𝜋𝑥𝑑𝑥

0 1 1 1 1
= ∫−1 𝑥𝑐𝑜𝑠𝑛𝜋𝑥𝑑𝑥 + ∫0 𝑥𝑐𝑜𝑠𝑛𝜋𝑥𝑑𝑥 + 2 ∫0 𝑐𝑜𝑠𝑛𝜋𝑑𝑥 = ∫−1 𝑥𝑐𝑜𝑠 𝑛𝜋𝑥𝑑𝑥 + 2 ∫0 𝑐𝑜𝑠𝑛𝜋𝑑𝑥

𝑥 sin 𝑛𝜋𝑥 cos 𝜋𝑥 2 sin 𝑛𝜋𝑥 1 1

=[ + ] 1 +[ ]0 (Using integration by parts on∫−1 𝑥𝑐𝑜𝑠 𝑛𝜋𝑥𝑑𝑥 )
𝑛𝜋 (𝑛𝜋)2 −1 𝑛𝜋

Badri A, Moges B. and Teklebrhan B. 71 AKU

 APPLIED MATHEMATICS III

sin 𝑛𝜋 cos 𝑛𝜋 −sin(−𝑛𝜋) cos(−𝑛𝜋) 𝑠𝑖𝑛 𝑛𝜋 sin 0

=[ + − ( + )] + 2 [ − ]
𝑛𝜋 (𝑛𝜋)2 𝑛𝜋 (𝑛𝜋)2 𝑛𝜋 𝑛𝜋

0 + (−1)𝑛 (−1)𝑛 0 0
=[ − (0 + )] + 2 [ − ]
(𝑛𝜋)2 (𝑛𝜋)2 𝑛𝜋 𝑛𝜋

(−1)𝑛 (−1)𝑛
= − + 2(0) = 0 − 0 + 0 = 0
(𝑛𝜋)2 (𝑛𝜋)2
1
Or for simplicity, ∫−1 𝑥𝑐𝑜𝑠 𝑛𝜋𝑥𝑑𝑥 = 0 because 𝑥𝑐𝑜𝑠 𝑛𝜋𝑥 is odd function and from section 1

1 1
∫−1(𝑜𝑑𝑑)(𝑒𝑣𝑒𝑛)𝑑𝑥 = 0 and clearly by integration 2 ∫0 𝑐𝑜𝑠𝑛𝜋𝑑𝑥 = 0

Hence 𝑎𝑛 = 0
1 0 1
Similarly, 𝑏𝑛 = ∫−1 𝑓(𝑥) sin 𝑛𝜋𝑥𝑑𝑥 = ∫−1 𝑥 sin 𝑛𝜋𝑥𝑑𝑥 + ∫0 (𝑥 + 2) sin 𝑛𝜋 𝑥𝑑𝑥

0 1 1 1 1
= ∫−1 𝑥 sin 𝑛𝜋𝑥𝑑𝑥 + ∫0 𝑥 sin 𝑛𝜋𝑥𝑑𝑥 + 2 ∫0 sin 𝑛𝜋 𝑥𝑑𝑥 = ∫−1 𝑥 sin 𝑛𝜋𝑥𝑑𝑥 + 2 ∫0 sin 𝑛𝜋 𝑥𝑑𝑥

1 1 𝑥𝑐𝑜𝑠 𝑛𝜋 sin 𝜋𝑥 1 𝑐𝑜𝑠 𝑛𝜋𝑥 1

= 2 ∫0 𝑥 sin 𝑛𝜋𝑥𝑑𝑥 + 2 ∫0 sin 𝑛𝜋 𝑥𝑑𝑥 = 2 [− + ] − 2[ ]
𝑛𝜋 (𝑛𝜋)2 −1 𝑛𝜋 0
2𝑐𝑜𝑠𝑛𝜋 2𝑐𝑜𝑠𝑛𝜋 2 2 4 2
=− − + 𝑛𝜋 = 𝑛𝜋 − 𝑛𝜋 (−1)𝑛 = 𝑛𝜋 [1 − (−1)𝑛 2]
𝑛𝜋 𝑛𝜋

6
𝑓𝑜𝑟 𝑜𝑑𝑑 𝑛
𝑛𝜋,
Thus 𝑏𝑛 = { −2
𝑓𝑜𝑟 𝑒𝑣𝑒𝑛 𝑛
𝑛𝜋,

Hence the Fourier series of 𝑓(𝑥) on (−1,1) is

∞
𝑛
𝑓(𝑥) = 𝑎𝑜 + ∑[𝑎𝑛 cos 𝑛𝜋𝑥 + 𝑏𝑛 sin 𝑛𝜋] = 𝑎𝑜 + ∑ 𝑏𝑛 sin 𝑛𝜋𝑥
𝑛=1
𝑛=1

= 𝑎𝑜 + ∑𝑛𝑛−1 𝑏𝑛 sin 𝑛𝜋 = 1 + ∑𝑛𝑛=1 𝑏𝑛 sin 𝑛𝜋𝑥

2 1 1 1 1
⟹ 𝑓(𝑥) = 1 + 𝜋 [(3 sin 𝜋𝑥) − 2 sin 2𝜋𝑥 + 3 (3sin 3𝜋𝑥) − 4 sin 4 𝜋𝑥 + 5 (sin 5𝜋𝑥) −
1
sin 6 𝜋𝑥 + − − −]
6

1 1
Further for 𝑥 = 2, 𝑓(𝑥) = 𝑥 + 2 = 2 + 2 = 5⁄2

Setting 𝑥 = 1⁄2 on both sides and we series above, we obtain

Badri A, Moges B. and Teklebrhan B. 72 AKU

 APPLIED MATHEMATICS III

5 2 1 3𝜋 1 1 3 7𝜋
= 1 + [3 sin 𝜋⁄2 − sin 𝜋 + 𝑠𝑖𝑛 − sin 2𝜋 + 3⁄5 sin 5𝜋⁄2 − sin 3𝜋 + sin − ⋯]
2 𝜋 2 2 4 6 7 2
2 3 3
=1+ [3 − 1 + − + − − −]
𝜋 5 7
5 2
or − 1 = 𝜋 [3 − 1 + 3⁄5 − 3⁄7 + ⋯ ]
2

3𝜋
⟹ = 3 − 1 + 3⁄5 − 3⁄7 + ⋯
4

𝜋 1 1 1
This gives = 1 − 3 + 5 − 7+ ….
4

Example 3.8 : Obtain the Fourier series for the periodic function

𝑓(𝑥) = 𝑒 −𝑥 − 𝑙 < 𝑥 < 𝑙, where 𝑓(𝑥 + 2𝑙) = 𝑓(𝑥)

Solution: we function f(x) is periodic with period 2 l. we forever coefficient are

1 𝑙 1 𝑙 1 1 1 𝑒 𝑙 −𝑒 −𝑙 sin ℎ 𝑙
𝑎𝑜 = 2𝑙 ∫−𝑙 𝑓(𝑥)𝑑𝑥 = 2𝑙 ∫−𝑙 𝑒 −𝑥 𝑑𝑥 = 2𝑙 [−𝑒 −𝑥 ] = 2𝑙 (𝑒 𝑙 − 𝑒 −𝑥 ) = 𝑙 ( )=
2 𝑙

sin ℎ 𝑙
⟹ 𝑎𝑜 = 𝑙

1 𝑙 𝑛𝜋𝑥 1 𝑙 𝑛𝜋𝑥
𝑎𝑛 = 𝑙 ∫−𝑙 𝑓(𝑥) cos 𝑑𝑥 = 𝑙 ∫−𝑙 𝑒 −𝑥 cos 𝑑𝑥 using integration by parts we have,
𝑙 𝑥

1 𝑙 𝑛𝜋𝑥 1 𝑒 −𝑥 𝑛𝜋𝑥 𝑛𝜋 𝑛𝜋𝑥 𝑙

𝑎𝑛 = 𝑙 ∫−𝑙 𝑒 −𝑥 cos 𝑑𝑥 = 𝑙 [ 𝑛𝜋 2 (−𝑐𝑜𝑠 + sin )]
𝑥 1+(
𝑙
) 𝑙 𝑙 𝑙 −𝑙

1 𝑒 −𝑙 𝑛𝜋 𝑒𝑙 𝑛𝜋
= [ 2 2 2 (– cos 𝜋 + sin 𝑛𝜋) − 2 2 2 (– cos 𝜋 + sin 𝑛𝜋)]
𝑙 𝑙 +𝑛 𝜋 𝑙 𝑙 +𝑛 𝜋 𝑙
𝑙2 𝑙2

1 𝑙 2 𝑒 −𝑙 𝑛𝜋 𝑙2𝑒 𝑙 𝑛𝜋
[2 (– cos 𝜋 + sin 𝑛𝜋) − (– cos 𝜋 + sin 𝑛𝜋)]
𝑙 𝑙 + (𝑛𝜋)2 𝑙 𝑙 2 + (𝑛𝜋)2 𝑙

1 𝑙2 𝑒 −𝑙 𝑙2 𝑒 −𝑙 𝑙2 𝑒 −𝑙
= 𝑙 [𝑙2 +(𝑛𝜋)2 cos 𝑛𝜋 + 𝑙2 +(𝑛𝜋)2 cos 𝑛𝜋] = 𝑙2 +(𝑛𝜋)2 [−𝑒 −𝑙 cos 𝑛 𝜋 + 𝑒 𝑙 cos 𝑛𝜋]

𝑙𝑐𝑜𝑠 𝑛𝜋 2𝑙𝑐𝑜𝑠𝑛𝜋 𝑒 𝑙 −𝑒 −𝑙 2(−1)𝑛

= 𝑙2 +(𝑛𝜋)2 (𝑒 𝑙 − 𝑒 −𝑙 ) = 𝑙2 +(𝑛𝜋)2 ( ) = 𝑙2 +(𝑛𝜋)2 sin ℎ 𝑙
2

2(−1)𝑛
⟹ 𝑎𝑛 = 𝑙2 +(𝑛𝜋)2 sin ℎ𝑙

Badri A, Moges B. and Teklebrhan B. 73 AKU

 APPLIED MATHEMATICS III

1 𝑙 𝑛𝜋𝑥 1 𝑙 𝑛𝜋𝑥
Similarly 𝑏𝑛 = 𝑙 ∫−𝑙 𝑓(𝑥) sin 𝑙
𝑑𝑥 = ∫ 𝑒 −𝑥 sin
𝑙 −𝑙 𝑙
𝑑𝑥

1 𝑒 −𝑙 𝑛𝜋𝑥 𝑛𝜋 𝑛𝜋𝑥 𝑙
= 𝑙[ 𝑛𝜋 (− sin − 𝑐𝑜𝑠 )]
1+( )2 𝑙 𝑙 𝑙 −𝑙
𝑙

1 𝑒 −𝑙 𝑛𝜋 𝑒𝑙 𝑛𝜋
= 𝑙[ 𝑛𝜋 (− sin 𝑛𝜋 − 𝑐𝑜𝑠𝑛𝜋) − 𝑛𝜋 (𝑠𝑖𝑛𝑛𝜋 − cos 𝑛𝜋)]
1+( )2 𝑙 1+( )2 𝑙
𝑙 𝑙

1 𝑙 2 𝑒 −𝑙 −𝑛𝜋 𝑙2𝑒 𝑙 𝑛𝜋
= (2 2
( cos 𝑛𝜋) + 2 2
( cos 𝑛𝜋))
𝑙 𝑙 + (𝑛𝜋) 𝑙 𝑙 + (𝑛𝜋) 𝑙

−𝑛𝜋𝑒 −𝑙 cos 𝑛𝜋+𝑛𝜋𝑒 𝑙 cos 𝑛𝜋 𝑛𝜋 cos 𝑛𝜋 2𝑛𝜋 cos 𝑛𝜋 𝑒 𝑙 −𝑒 −𝑙 2(−1)𝑛 𝑛𝜋

= = (𝑒 𝑙 − 𝑒 −𝑙 ) = ( )= 𝑠𝑖𝑛ℎ𝑙
𝑙2 +(𝑛𝜋)2 𝑙2 +𝑛2 𝜋2 𝑙2 +(𝑛𝜋)2 2 𝑙2 +𝑛2 𝜋2

2(−1)𝑛 𝑛𝜋
⟹ 𝑏𝑛 = 𝑠𝑖𝑛ℎ𝑙
𝑙2 +𝑛2 𝜋2

Hence the Fourier series or 𝑓(𝑥) on (−𝑙, 𝑙) is

∞
𝑛𝜋𝑥 𝑛𝜋𝑥
𝑓(𝑥) = 𝑎𝑜 + ∑ (𝑎𝑛 cos + 𝑏𝑛 sin )
𝑙 𝑥
𝑛=1

∞
sin ℎ𝑙 2(−1)𝑛 𝑛𝜋𝑥 2(−1)𝑛 𝑛𝜋𝑥
⟹ 𝑒 −𝑥 = +∑( 2 2 2
sin ℎ𝑙. cos + 2 2
sinh𝑙 sin )
𝑥 𝑙 +𝑛 𝜋 𝑙 𝑙 +𝑛 𝜋 𝑙
𝑛=1

∞
sin ℎ𝑙 (−1)𝑛 𝑛𝜋𝑥 𝑛𝜋𝑥
= + 2 sin ℎ𝑙 (∑ 2 (𝑙 cos + 𝑛𝜋 𝑠𝑖𝑛 ))
𝑙 𝑙 + 𝑛2 𝜋 2 𝑙 𝑙
𝑛=1

1 1 𝜋𝑥 1 2𝜋𝑥 1 3𝜋𝑥
= sin ℎ𝑙 [( − 2𝑙 ( 2 cos − cos + cos )
𝑙 𝑙 + 𝜋2 𝑙 𝑙 2 + 22 𝜋 2 𝑙 𝑙 2 + 32 𝜋 2 𝑙
1 𝜋𝑥 2 2𝜋𝑥 3 3𝜋𝑥
− 2𝜋 ( 2 2 sin − 2 2 2
sin + 2 2 2
sin … . . ))]
𝑙 𝜋 𝑙 𝑙 +2 𝜋 𝑙 𝑙 +2 𝜋 𝑙

Badri A, Moges B. and Teklebrhan B. 74 AKU

 APPLIED MATHEMATICS III

Fourier integrals
Fourier integrals extend the concept or Fourier series to non periodic functions defined for all x.
A non periodic function which cannot be represented as a Fourier series over the entire real line
may be represented in an integral form. In many practical problems we came across functions
defined on −∞ < 𝑥 < ∞ that are non periodic e.g 𝑓(𝑥) = 𝑒 −𝑥2 𝑓(𝑥)

Figure3.4 𝑓(𝑥) = 𝑒 −𝑥2

We cannot expand such functions in Fourier series since they are not periodic; however we can
consider such function to be periodic, but with infinite period.

Definition (Fourier integral): The integral representation of the form

∞
𝑓(𝑥) = ∫0 [𝐴(𝜔) cos 𝜔𝑥 + 𝛽(𝜔) sin 𝜔𝑥]𝑑𝜔

1 ∞ 1 ∞
Where 𝐴(𝜔) = ∫−∞ 𝑓(𝑢) cos 𝜔𝑢𝑑𝑢 𝑎𝑛𝑑 𝐵(𝜔) = ∫−∞ 𝑓(𝑢) sin 𝜔𝑢𝑑𝑢
𝜋 𝜋

Are the Fourier integrals coefficients is called the Fourier integral representation of 𝑓(𝑥)

The sufficient conductions for which we Fourier integral representation is valid are,

1. 𝑓(𝑥) is piecewise cortisones on every integral [– 𝑙, 𝑙]

∞
2. 𝑓(𝑥) is absolutely integrals on the real axis, that is ∫−∞/𝑓(𝑥)/𝑑𝑥 converges
3. At every 𝑥 on the real line 𝑓(𝑥) has left and right-hand derivatives

Theorem(Fourier integral theorem): If f(x) satisfies the conditions 1 to 3 stated above, then the
Fourier integral or 𝑓 converges to at every point 𝑥 at which 𝑓 is continuous, and to the mean
value [𝑓(𝑥+0) + 𝑓(𝑥-0)]/2 at every point 𝑥 at which 𝑓 is discontinuous, where 𝑓(𝑥+) and 𝑓(𝑥-)
are the right and left hand limits respectively.

1, −1 ≤ 𝑥 ≤ 1
Example 3.9: find the Fourier integrals representation of 𝑓(𝑥) = { and hence
0, 𝑓𝑜𝑟 |𝑥|>|
∞ sin 𝜔 𝜋
prove that ∫0 𝑑𝜔 =
𝜔 2

Badri A, Moges B. and Teklebrhan B. 75 AKU

 APPLIED MATHEMATICS III

Solution: 𝑓(𝑥) is piecewise smooth and absolutely integrals over (−∞, ∞) . Thus 𝑓(𝑥) has
Fourier integral representation.

The Fourier integral coefficients of 𝑓(𝑥) are 1 𝑓(𝑥)

-1 1 𝑥

Figure 3.5 the graph of 𝑓

1 ∞ 1 1 sin 𝜔𝑢 1 2 sin 𝜔
𝐴(𝜔) = 𝜋 ∫−∞ 𝑓(𝑢) cos 𝜔𝑢𝑑𝑢 = 𝜋 ∫−1 cos 𝜔 𝑢𝑑𝑢 = [ ] = 𝜋𝜔
𝜋𝜔 −1
1 ∞ 1 1
and 𝐵(𝜔) = 𝜋 ∫−∞ 𝑓(𝑢) sin 𝜔𝑢𝑑𝑢 = 𝜋 ∫−1 sin 𝜔 𝑢𝑑𝑢 = 0 (because sin 𝜔 𝑢 𝑖𝑠 𝑜𝑑𝑑 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛)

Hence the Fourier integral of 𝑓(𝑥) is

∞ 2 ∞ cos 𝜔𝑥 sin 𝜔
𝑓(𝑥) = ∫0 [𝐴(𝜔) cos 𝜔𝑥 + 𝛽(𝜔) sin 𝜔𝑥]𝑑𝜔 = 𝜋 ∫0 𝑑𝜔
𝜔

Since 𝑥 = ±1 are the points of discontinuity of 𝑓(𝑥), thus at 𝑥 = ±1

2 ∞ cos 𝜔𝑥 sin 𝜔 1 1
𝑓(𝑥) = 𝜋 ∫0 𝑑𝑤 = 2 [𝑓(𝑥 + 0) + 𝑓(𝑥 − 0)] = 2 for 𝑥 = ±1
𝜔

𝜋
𝑓𝑜𝑟 − 1 < 𝑥 < 1
2,
2 ∞ 𝑐𝑜𝑠𝜔𝑥𝑠𝑖𝑛𝜔 𝜋 ∞ sin 𝜔 𝜋
Thus, ∫ 𝑑𝑤 ={ 𝑓𝑜𝑟 𝑥 = ± 1 and setting𝑥 = 0, we have ∫0 𝑑𝜔 =
𝜋 0 𝜔 4, 𝜔 2
0, 𝑓𝑜𝑟 1𝑥1 > 1

𝑒 −𝑥 , 𝑥 > 0
Example 3.10: find the Fourier integral representation of 𝑓(𝑥) = {
0, 𝑥 ≤ 0

and find the value of the repulsing integral when,(a) 𝑥 < 0, (b) 𝑥 = 0 (c)𝑥 > 0 also derive that
∞ 𝑑𝑤 𝜋
∫0 =
1+𝑤 2 2

Solution: the given function f(x) is piecewise smooth and is absolutely inferable over(−∞, ∞),
∞ ∞
Since ∫−∞|𝑓(𝑥)| 𝑑𝑥 = ∫0 𝑒 −𝑥 𝑑𝑥 = 1, 𝑡ℎ𝑎𝑡 𝑖𝑠 𝑏𝑦 𝑢𝑠𝑖𝑛𝑔 𝑖𝑚𝑝𝑟𝑜𝑝𝑒𝑟 𝑖𝑛𝑡𝑒𝑔𝑟𝑎𝑙𝑠

∞ 𝑏 𝑏
∫0 𝑒 −𝑥 𝑑𝑥 = lim ∫0 𝑒 −𝑥 𝑑𝑥 = lim [−𝑒 −𝑥 ] 0 = lim [−𝑒 −𝑏 + 𝑒 0 ] = lim (1 − 𝑒 −𝑏 )
𝑏→∞ 𝑏→∞ 𝑏→∞ 𝑏→∞

= lim 1 − lim 𝑒 −𝑏 = 1−0= 1

𝑏→∞ 𝑏→∞

Now let us find the Fourier integral coefficients

Badri A, Moges B. and Teklebrhan B. 76 AKU

 APPLIED MATHEMATICS III

1 ∞ 1 ∞
𝐴(𝑤) = 𝜋 ∫−∞ 𝑓(𝑢) cos 𝜔𝑢 𝑑𝑢 = 𝜋 ∫0 𝑒 −𝑢 cos 𝜔𝑢 𝑑𝑢 (𝑓𝑜𝑟 𝑥 > 0, 𝑓(𝑥) = 𝑒 −𝑥 , 𝑥 ≤ 0 , 𝑓(𝑥) =
sin 𝜔𝑢
0) Now using integration by pans, let 𝑡 = 𝑒 −𝑢 , 𝑑𝑡 = −𝑒 −𝑢 𝑑𝑢, let 𝑑𝑣 = cos 𝜔𝑢, 𝑣 = 𝑤

𝑒 −𝑢 sin 𝜔𝑢 1
∫ 𝑡𝑑𝑣 = 𝑡𝑣 − ∫ 𝑣𝑑𝑡 ⟹ ∫ 𝑒 −𝑢 cos 𝜔𝑢𝑑𝑢 = 𝜔
+ 𝑤 ∫ 𝑒 −𝑢 sin 𝜔𝑢𝑑𝑢 ………………. (1)

Again using integration by parts on ∫ 𝑒 −𝑢 sin 𝜔𝑢𝑑𝑢, we have

−cos 𝜔𝑢
𝑙𝑒𝑡 𝑡 = 𝑒 −𝑢 , 𝑑𝑡 = −𝑒 −𝑢 𝑑𝑢 and 𝑙𝑒𝑡 𝑑𝑣 = sin 𝜔𝑢 , 𝑣 = 𝑣 = 𝑤

cos 𝜔𝑢 1
Then, ∫ 𝑒 −𝑢 sin 𝜔𝑢𝑑𝑢 = −𝑒 −𝑢 − 𝑤 ∫ 𝑒 −𝑢 cos 𝜔𝑢𝑑𝑢………………………………(2)
𝑤

Now substituting (2) into (1), we have ,

−𝑢
𝑒 −𝑢 sin 𝜔𝑢 1 −𝑒 −𝑢 cos 𝜔𝑢 1
∫𝑒 𝑐𝑜𝑠𝜔𝑢𝑑𝑢 = + [ − ∫ 𝑒 −𝑢 cos 𝜔𝑢𝑑𝑢]
𝜔 𝜔 𝜔 𝜔
𝑒 −𝑢 sin 𝜔𝑢 𝑒 −𝑢 cos 𝜔𝑢 1
= − − 𝜔2 ∫ 𝑒 −𝑢 cos 𝜔𝑢𝑑𝑢
𝜔 𝜔2

1 −𝑢
𝑒 −𝑢 sin 𝜔𝑢 𝑒 −𝑢 cos 𝜔𝑢
= (1 + ) ∫ 𝑒 cos 𝜔𝑢𝑑𝑢 = −
𝜔2 𝜔 𝜔2
∞
1 + 𝜔2 −𝑢
𝑒 −𝑢 sin 𝜔𝑢 𝑒 −𝑢 cos 𝜔𝑢
= ∫ 𝑒 cos 𝜔𝑢𝑑𝑢 = −
𝜔2 𝜔 𝜔2
0

∞ 𝜔2 𝑒 −𝑢 sin 𝜔𝑢 −𝑒 −𝑢 cos 𝜔𝑢 ∞
⇒ ∫0 𝑒 −𝑢 cos 𝜔𝑢𝑑𝑢 = 1+𝜔2 [ − ]
𝜔 𝜔2 0
𝜔2 𝑒 −𝑢 sin 𝜔𝑢 −𝑒 −𝑢 cos 𝜔𝑢 ∞
= 1+𝜔2 [ − ]
𝜔 𝜔2 0
𝑒 −𝑢 sin 𝜔𝑢 −𝑒 −𝑢 cos 𝜔𝑢 𝑒 −𝑢 sin 𝜔𝑢 −𝑒 −𝑢 cos 𝜔𝑢 1
As 𝑢 → ∞, − → 0 and as 𝑢 → 0, − → 1+𝜔2
𝜔 𝜔2 𝜔 𝜔2

1
When u tends to infinity, it becomes zero, and when u tend to zero u becomes 1+𝑤2

∞ 1
⇒ ∫0 𝑒 −𝑢 cos 𝜔𝑢𝑑𝑢 = 1+𝜔2

1 ∞ 1 ∞ 1
so that 𝐴(𝜔) = 𝜋 ∫−∞ 𝑓(𝑢) cos 𝜔𝑢𝑑𝑢 = 𝜋 ∫0 𝑒 −𝑢 cos 𝜔𝑢𝑑𝑢 = 𝜋(1+𝜔2)

1 ∞ 1 ∞ 𝜔
similarly,𝐵(𝜔) = ∫−∞ 𝑓(𝑢) sin 𝜔𝑢𝑑𝑢 = ∫0 𝑒 −𝑢 sin 𝜔𝑢𝑑𝑢 = (use integration by parts)
𝜋 𝜋 𝜋(1+𝜔2 )

thus, the Fourier integral representation of 𝑓(𝑥) is

Badri A, Moges B. and Teklebrhan B. 77 AKU

 APPLIED MATHEMATICS III

∞ 1 𝑐𝑜𝑠𝜔𝑥+𝜔 sin 𝜔𝑥
𝑓(𝑥) = ∫0 [𝐴(𝜔) cos 𝜔𝑥 + 𝛽(𝜔) sin 𝜔𝑥]𝑑𝜔 = 𝜋 ∫ 𝑑𝜔
1+𝜔 2

1 ∞ cos 𝜔𝑥+𝜔𝑠𝑖𝑛𝜔𝑥
(a) For 𝑥 < 0, 𝑓(𝑥) = 𝜋 ∫0 𝑑𝜔 = 0
1+𝜔 2

(b) at the point of discontinuity, 𝑥 = 0

∞
1 cos 𝜔𝑥 + 𝜔𝑠𝑖𝑛 𝜔𝑥 1 1 1
∫ 𝑑𝜔 = [𝑓(𝑥 + 0) + 𝑓(𝑥 − 0] = [1 + 0] =
𝜋 1 + 𝜔2 2 2 2
0

1 ∞ cos 𝜔𝑥+𝜔𝑠𝑖𝑛𝜔𝑥
(c) For 𝑥 > 0, 𝑓(𝑥) = 𝜋 ∫0 𝑑𝜔 = 𝑒 −𝑥
1+𝜔2

0, 𝑥 < 0
∞ cos 𝜔𝑥+𝑤 sin 𝜔𝑥 𝜋
Thus, ∫0 1+𝜔 2
𝑑𝜔 = { 2, 𝑥 = 0
𝜋𝑒 −𝑥 , 𝑥 > 0

Setting 𝑥 = 0 in the above we have,

∞ cos 𝜔𝑥+𝜔 sin 𝜔𝑥 ∞ cos 0+𝜔 sin 0 ∞ 𝑑𝜔
∫0 𝑑𝑤 = ∫0 𝑑𝜔 = ∫0 = 𝜋⁄2
1+𝜔2 1+𝜔2 1+𝜔 2

Exercises 3.2

1. Find the Fourier series representation of each of the following

1
(a) 𝑓(𝑥) = 2 (𝜋 − 𝑥) in 𝜋 < 𝑥 < 𝜋 (b) 𝑓(𝑥) = 𝑥 2 in 𝜋 < 𝑥 < 𝜋

𝑥, 0≤𝑥≤𝜋 𝜋2 1 1
(c) 𝑓(𝑥) = { and deduce that 8 = 32 + 52 + ⋯
2𝜋 − 𝑥, 𝜋 ≤ 𝑥 ≤ 2𝜋

0, −2 < 𝑥 ≤ −1
2
(d) 𝑓(𝑥) = 1 − 𝑥 over the interval (1,-1) (e) 𝑓(𝑥) = { 𝑘, −1 < 𝑥 < 1 , 𝑓(𝑥) = (𝑥 + 4)
0, 1 ≤ 𝑥 < 2

2. in each of the following drive the Fourier integral representation, at which points, if any ,does the
Fourier integral fail to converge to f(x)? to what value does the integral converge at those points,
𝜋 𝜋
100, 0 ≤ 𝑥 ≤ 2 ( )𝑐𝑜𝑠𝑥, |𝑥| ≤
2 2
(a) 𝑓(𝑥) = { (b) 𝑓(𝑥) = { 𝜋
0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 0, |𝑥| >
2

𝑏𝑥⁄ , |𝑥| ≤ 𝑎, 𝑎, 𝑏 > 0 0, 𝑥<0

(c) 𝑓(𝑥) = { 𝑎 (d) 𝑓(𝑥) = {𝑘, 0 ≤ 𝑥 ≤ 𝜋
0, |𝑥| > 𝑎 0, 𝑥>𝜋

Badri A, Moges B. and Teklebrhan B. 78 AKU

 APPLIED MATHEMATICS III

3.3The complex Fourier series and integrals

3.3.1 Complex Fourier series

Overview:

In this section, we are going to introduce the complex Fourier series and the complex Fourier
integral representation of real functions together with some examples.

Section Objectives:

At the end of this subtopic, students will be able to:

 Define and understand about complex Fourier series of periodic functions;

 Understand about complex Fourier integral representation of real functions
To simplify the calculations it is some times convenient to work in terms of complex
numbers even when the parameters under reference are real. In this context we study the
complex of the Fourier series of a real function 𝑓(𝑥), this form is of especial interest in
the study of electrical circuits.

Definition (Complex Fourier series): - let f(x) be a real periodic function of period 2𝑙 over the
interval (−𝑙. 𝑙).then the complex Fourier series representation of 𝑓 is defined as.
𝑖𝑛𝜋𝑥
𝑓(𝑥) = lim ∑𝑘𝑛=−𝑘 𝑐𝑛 𝑒 𝑙 for −𝑙 < 𝑥 < 𝑙
𝑘⟶∞

𝑙 −𝑖𝑛𝜋𝑥
1
Where 𝑐𝑛 = 2𝑙 ∫−𝑙 𝑓(𝑥) 𝑒 𝑙 𝑑𝑥 , 𝑛 = 0, ±1, , ±2 are the complex Fourier coefficients.

Note: in Complex Fourier series, at points of continuity 𝑓(𝑥) the series converges to 𝑓(𝑥) while
at point of discontinuity it converges to the midpoint

Example 3.11: find the complex Fourier series representation of the function

0, 0< 𝑥 ≤1
𝑓(𝑥) = { When 𝑓(𝑥) = 𝑓(𝑥 + 4)
1, 1< 𝑥 <4

Solution: The function 𝑓(𝑥) is periodic with period 4 defined on the internal(0,4), with 2𝑙 =
4,𝑙 = 2. Thus the complex Fourier coefficient 𝑐𝑛 is given by:

4 −𝑖𝑛𝜋𝑥 4 −𝑖𝑛𝜋𝑥
1 1
𝑐𝑛 = 4 ∫0 𝑓(𝑥)𝑒 𝑙 𝑑𝑥 = 4 ∫0 𝑒 2 𝑑𝑥

Badri A, Moges B. and Teklebrhan B. 79 AKU

 APPLIED MATHEMATICS III

1 4
For 𝑛 = 0, we get 𝑐𝑜 = 4 ∫1 𝑑𝑥 = 3⁄4

For all 𝑐𝑛, except 𝑛 = 0

−𝑖𝑛𝜋𝑥 4 −𝑖𝑛𝜋𝑥 −𝑖𝑛𝜋𝑥 −𝑖𝑛𝜋
1 4 1 1 −2 4 1 −2 𝑖 𝑖2𝑛𝜋
𝑐𝑛 = 4 ∫0 𝑓(𝑥)𝑒 𝑙 𝑑𝑥 = 4 ∫0 𝑒 2 𝑑𝑥 = 4 [𝑖𝑛𝜋 𝑒 2 ] = 4 [(𝑖𝑛𝜋 𝑖 (𝑒 − 𝑒 2 )]
1
𝑖 −𝑛𝜋𝑥⁄
= 2𝜋𝑛 [1 − 𝑒 2]

Hence the complex Fourier series representation or f(x) is

−𝑖𝑛𝜋𝑥⁄ 𝑖𝑛𝜋𝑥
𝑖
𝑓(𝑥) = lim ∑𝑘𝑛=−𝑘 2𝜋𝑛 (1 − 𝑒 2) 𝑒 2
𝑘⟶∞

−𝑖𝑛𝜋𝑥⁄ 𝑖𝑛𝜋𝑥
3 𝑖
𝑓(𝑥) = 4 + lim ∑𝑘𝑛=−𝑘 2𝜋𝑛 (1 − 𝑒 2 ) 𝑒 2 (𝑛 ≠ 0)
𝑘⟶∞

Example 3.12: find the complex Fourier series representation of the function

𝑓(𝑥) = 𝑒,−𝑥 − 𝜋 < 𝑥 < 𝜋, 𝑓(𝑥) = 𝑓(𝑥 + 2𝜋)

Solution: The function 𝑓(𝑥) is periodic with period2𝜋, defined on the internal (−𝜋, 𝜋). Here 𝑙 =
𝜋. Thus the complex Fourier coefficients are:
𝜋 𝜋
1 1 −1 𝜋
𝑐𝑛 = ∫ 𝑒 −𝑥 𝑒 𝑖𝑛𝑥 𝑑𝑥 = ∫ 𝑒 −(1+𝑖𝑛)𝑥 𝑑𝑥 = [𝑒 −𝑖(1+𝑖𝑛)𝑥 ]
2𝜋 2𝜋 2𝜋(1 + 𝑖𝑛) −𝜋
−𝜋 −𝜋

−1 −1
= [𝑒 −(1+𝑖𝑛)𝜋 − 𝑒 (1+𝑖𝑛)𝜋 ] = (𝑒 −𝜋 𝑒 −𝑖𝑛𝜋 − 𝑒 𝜋 𝑒 𝑖𝑛𝜋 )
2𝜋(1 + 𝑖𝑛) 2𝜋(1 + 𝑖𝑛)

−1
= [𝑒 −𝜋 (cos 𝑛𝜋 − 𝑖 sin 𝑛𝜋) − 𝑒 𝜋 (cos 𝑛𝜋 + 𝑖 sin 𝑛𝜋)]
2𝜋(1 + 𝑖𝑛)
(1−𝑖𝑛) (1−𝑖𝑛)𝑠𝑖𝑛ℎ𝜋
= 2𝜋(1+𝑛2 ) [(𝑒 𝜋 − 𝑒 −𝜋 ) cos 𝑛𝜋] = (−1)𝑛 𝜋(1+𝑛2 )

Hence, the complex Fourier series is;

𝑠𝑖𝑛ℎ𝜋 𝑛 1−𝑖𝑛 𝑖𝑛𝑥

𝑓(𝑥) = lim ∑𝑘𝑛=−𝑘(−1) ( )𝑒
𝜋 𝑘⟶∞ 1+𝑛2

The Complex Fourier Integral Representation

Analogous to the complex form of the Fourier series the Fourier integral can be expressed in the
equivalent complex form this complex form provides the necessary plate form to develop the
Fourier transform which are highly developed as a methodology like the Laplace transform.

Badri A, Moges B. and Teklebrhan B. 80 AKU

 APPLIED MATHEMATICS III

Definition (complex Fourier integral) the integral representation or the form

∞
𝑓(𝑥) = ∫−∞ 𝑐(𝜔)𝑒 𝑖𝜔𝑥 𝑑𝜔

1 ∞
Where, 𝑐(𝜔) = 2𝜋 ∫−∞ 𝑓(𝑢)𝑒 −𝑖𝜔𝑢 𝑑𝑢

is the complex Fourier integral coefficient is called the complex Fourier integral representation of 𝑓
in the real line

Example 3.13: If 𝑓(𝑥) = 𝑒 −𝑎|𝑥| for all real 𝑥 and with 𝑎 > 0, a positive constant Then find the
complex Fourier integral representation or 𝑓.

𝑒 −𝑎𝑥 , 𝑓𝑜𝑟 𝑥 < 0 𝑥 𝑓𝑜𝑟 𝑥 > 0

Solution: The function is { 𝑎𝑥 as |𝑥| = { , 𝑎 > 0 being a constant
𝑒 , 𝑓𝑜𝑟 𝑥 < 0 −𝑥 𝑓𝑜𝑟 𝑥 < 0

Obviously, 𝑓(𝑥) is piecewise smooth and is absolutely integrable over the interval(−∞, ∞).

The complex Fourier integral coefficient of 𝑓 is given by

∞ 0 ∞
1 1
𝑐(𝜔) = ∫ 𝑓(𝑢)𝑒 −𝑖𝜔𝑢 𝑑𝑢 = [ ∫ 𝑒 𝑎𝑢 𝑒 −𝑖𝜔𝑢 𝑑𝑢 + ∫ 𝑒 −𝑎𝑢 𝑒 −𝑖𝜔𝑢 𝑑𝑢]
2𝜋 2𝜋
−∞ −∞ 0

0 ∞
1 1 𝑒 (𝑎−𝑖𝜔)𝑢 0 1 𝑒 −(𝑎+𝑖𝜔)𝑢 ∞
= [ ∫ 𝑒 (𝑎−𝑖𝜔)𝑢 𝑑𝑢 + ∫ 𝑒 (𝑎+𝑖𝜔)𝑢 𝑑𝑢] = [ ] + [ ]
2𝜋 2𝜋 𝑎 − 𝑖𝜔 −∞ 2𝜋 −(𝑎 + 𝑖𝜔) 0
−∞ 0

𝑒 (𝑎−𝑖𝜔)𝑢 1 𝑒 (𝑎−𝑖𝜔)𝑢
From this as 𝑢 → 0, → 𝑎−𝑖𝜔 and as 𝑢 → −∞ →0
𝑎−𝑖𝜔 𝑎−𝑖𝜔

𝑒 −(𝑎+𝑖𝜔)𝑢 𝑒 −(𝑎+𝑖𝜔)𝑢 −1
Similarly, as 𝑢 → ∞ , → 0 and as 𝑢 → 0, →
−(𝑎+𝑖𝜔) −(𝑎+𝑖𝜔) 𝑎+𝑖𝜔

1 1 1 𝑎 𝑎
= [ + ]= = 𝑐(𝜔) =
2𝜋 𝑎 + 𝑖𝜔 𝑎 − 𝑖𝜔 𝜋(𝑎2 + 𝜔 2 ) 𝜋(𝑎2 + 𝜔 2 )
𝑎
⟹ 𝑐(𝜔) =
𝜋(𝑎2 + 𝜔2)

Thus the complex Fourier integral representation or 𝑓(𝑥) is

Badri A, Moges B. and Teklebrhan B. 81 AKU

 APPLIED MATHEMATICS III

∞ ∞ ∞
𝑎 𝑎 𝑎
𝑓(𝑥) = 𝑒 −𝑎|𝑥| = ∫ 𝑐(𝜔)𝑒 𝑖𝜔𝑥 𝑑𝜔 = ∫ 𝑒 𝑖𝜔𝑥
𝑑𝜔 = ∫ 𝑒 𝑖𝜔𝑥 𝑑𝜔
𝜋(𝑎2 + 𝜔 2 ) 𝜋 𝑎2 + 𝜔 2
−∞ −∞ −∞

Example3.14: find the complex Fourier integral representation of 𝑓(𝑥) =

𝑠𝑖𝑛𝜋𝑥 , |𝑥| ≤ 5
{ and determine what this integrals converges to.
0 |𝑥| > 5

Solution: Clearly 𝑓 is piecewise continuing and absolutely integrals on (−∞, ∞) (i.e over the
real line)

So, to find the complex Fourier integral, we first find the complex Fourier integral coefficient
c(𝜔) ,that is
∞ 5 5
1 1 1
𝑐(𝜔) = ∫ 𝑓(𝑢)𝑒 −𝑖𝜔𝑢 𝑑𝑢 = ∫ 𝑠𝑖𝑛𝜋𝑢 𝑒 −𝑖𝜔𝑢 𝑑𝑢 = ∫ 𝑒 −𝑖𝜔𝑢 sin 𝜋𝑢 𝑑𝑢
2𝜋 2𝜋 2𝜋
−∞ −5 −5

5
Now using integration by parts on∫−5 𝑒 −𝑖𝜔𝑢 sin 𝜋𝑢 𝑑𝑢, we let 𝑡 = 𝑒 −𝑖𝜔𝑢 , 𝑑𝑡 = 𝑖𝜔𝑒 𝑖𝜔𝑢 𝑑𝑢
− cos 𝜋𝑢
𝑑𝑣 = sin 𝜋𝑢 , 𝑣 =
𝜋

5 −𝑒 −𝑖𝜔𝑢 cos 𝜋𝑢 𝑖𝜔 5
∫−5 𝑒 −𝑖𝜔𝑢 𝑠𝑖𝑛𝜋𝑢 𝑑𝑢 = ( 𝜋
− 𝜋
∫−5 𝑒 −𝑖𝜔𝑢 cos 𝜋 𝑢𝑑𝑢) (1)

5
Again using integration by party on ∫−5 𝑒 −𝑖𝑤𝑢 cos 𝜋𝑢 𝑑𝑢, let t = 𝑒 −𝑖𝜔𝑢 , 𝑑𝑡 = −𝑖𝜔𝑒 𝑖𝜔𝑢 𝑑𝑢

sin𝜋𝑢
𝑑𝑣 = cos 𝜋𝑢 , 𝑣 = 𝜋

5 𝑒 −𝑖𝑤𝑢 sin𝜋𝑢 𝑖𝑤 5
⟹ ∫−5 𝑒 −𝑖𝑤𝑢 cos 𝜋𝑢 𝑑𝑢 = + ∫−5 𝑒 −𝑖𝑤𝑢 sin 𝜋 𝑢𝑑𝑢 (2)
𝜋 𝜋

Now subsisting (2) into (1), we have

5 −𝑒 −𝑖𝑤𝑢 cos 𝜋𝑢 𝑖𝑤 𝑒 −𝑖𝑤𝑢 sin𝜋𝑢 𝑖𝑤 5 5

∫−5 𝑒 −𝑖𝜔𝑢 sin 𝜋𝑢 𝑑𝑢 = [ 𝜋
− 𝜋
( 𝜋
+ 𝜋
∫−5 𝑒 −𝑖𝑤𝑢 sin 𝜋 𝑢𝑑𝑢 )] −5

−𝑒 −𝑖𝑤𝑢 cos 𝜋𝑢 𝑖𝑤 𝑒 −𝑖𝑤𝑢 sin𝜋𝑢 5 𝑤2 5

=[ − ( )] −5 + 𝜋2 ∫−5 𝑒 −𝑖𝑤𝑢 sin 𝜋 𝑢𝑑𝑢
𝜋 𝜋 𝜋

−𝑒 −𝑖𝑤𝑢 cos 𝜋𝑢 𝑖𝑤𝑒 −𝑖𝑤𝑢 sin𝜋𝑢 5 𝑤2 5

[ − ] −5 + 𝜋2 ∫−5 𝑒 −𝑖𝑤𝑢 sin 𝜋 𝑢𝑑𝑢
𝜋 𝜋2

1 5 1 −𝑒 −𝑖𝑤𝑢 cos 𝜋𝑢 𝑖𝑤𝑒 −𝑖𝑤𝑢 sin𝜋𝑢 5 𝑤2 5

⟹ ∫ 𝑒 −𝑖𝑤𝑢 sin 𝜋𝑢 𝑑𝑢 = 2𝜋 [
2𝜋 −5 𝜋
− 𝜋2
] −5 + 𝜋2 ∫−5 𝑒 −𝑖𝑤𝑢 sin 𝜋 𝑢𝑑𝑢

𝑤2 5 1 −𝑒 −𝑖𝑤𝑢 cos 𝜋𝑢 𝑖𝑤𝑒 −𝑖𝑤𝑢 sin𝜋𝑢 5

(1 − 𝜋2 ) ∫−5 𝑒 −𝑖𝑤𝑢 sin 𝜋𝑢 𝑑𝑢 = 2𝜋 [ − ] −5
𝜋 𝜋2

Badri A, Moges B. and Teklebrhan B. 82 AKU

 APPLIED MATHEMATICS III

5 −𝜋 2 −𝑒 −𝑖𝑤 cos 𝜋𝑢 𝑖𝑤𝑒 −𝑖𝑤𝑢 sin 𝜋𝑢 5

⟹ ∫−5 𝑒 −𝑖𝑤𝑢 sin 𝜋𝑢 𝑑𝑢 = 𝑤2 −𝜋2 [ − ] −5
𝜋 𝜋2

5 −𝜋2 −𝑒 −𝑖𝑤 cos 𝜋𝑢 𝑖𝑤𝑒 −𝑖𝑤𝑢 sin 𝜋𝑢 5

∫−5 𝑒 −𝑖𝑤𝑢 sin 𝜋𝑢 𝑑𝑢 = 𝑤2 −𝜋2 [ 𝜋
− 𝜋2
] −5

−1 5
= 𝑤2 −𝜋2 [−𝜋𝑒 −𝑖𝑤𝑢 cos 𝜋𝑢 −𝑖𝑤𝑒 −𝑖𝑤𝑢 sin 𝜋𝑢] −5

−1
[−𝜋𝑒 −𝑖5𝑤 cos 5𝜋 −𝑖𝑤𝑒 −𝑖5𝑤 sin 5𝜋 + 𝜋𝑒 𝑖5𝑤 cos(−5)𝜋 + 𝑖𝑤𝑒 𝑖5𝑤 sin(−5)𝜋]
𝑤 2 −𝜋 2

−𝜋
= [𝑒 −𝑖5𝑤 −𝑒 𝑖5𝑤 ]
𝑤2 − 𝜋2
−𝜋
= [cos 5𝑤 − 𝑖 sin 5𝑤 − 𝑐𝑜𝑠5𝑤 − 𝑖𝑠𝑖𝑛 5𝑤]
(𝑤 2 − 𝜋 2 )

𝜋 2𝜋𝑖 𝑠𝑖𝑛 5𝑤
(2𝑖𝑠𝑖𝑛 5𝑤) =
(𝑤 2 2
−𝜋 ) 𝑤2 − 𝜋2
5 2𝜋𝑖 𝑠𝑖𝑛 5𝑤
From this ∫−5 𝑒 −𝑖𝜔𝑢 sin 𝜋𝑢 𝑑𝑢 = 𝑤 2 −𝜋 2

1 ∞ 1 2𝜋𝑖 𝑠𝑖𝑛 5𝑤 𝑖 𝑠𝑖𝑛 5𝑤

Thus, 𝑐(𝑤) = 2𝜋 ∫−∞ 𝑓(𝑢)𝑒 −𝑖𝑤𝑢 𝑑𝑢 = 2𝜋 ( )=
𝑤 2 −𝜋 2 𝑤 2 −𝜋2

∞ 5 sin 5𝑤
Hence 𝑓(𝑥) = ∫−∞ 𝑐(𝑤)𝑒 −𝑖𝑤𝑢 𝑑𝑢 = 𝑖 ∫−5 (𝑤2 −𝜋2 ) 𝑒 −𝑖𝑤𝑥 𝑑𝑢

is complex Fourier integral representation of 𝑓.

Exercises 3.3

1. In each of the following find the complex Fourier series representation of 𝑓(𝑥) on the given interval.

Badri A, Moges B. and Teklebrhan B. 83 AKU

 APPLIED MATHEMATICS III

−1, −1 < 𝑥 < 0

(a) 𝑓(𝑥) = { 𝑓(𝑥 + 2) = 𝑓(𝑥)
1, 0 ≤ 𝑥 < 1

(b) 𝑓(𝑥) = 𝑒 𝑥 , 0 < 𝑥 < 1 , 𝑓(𝑥 + 1) = 𝑓(𝑥)

(c) 𝑓(𝑥) = 𝑒 −|𝑥| , −2 < 𝑥 < 2, 𝑓(𝑥 + 4) = 𝑓(𝑥)

2. In each of the following problems, the complex Fourier integral of the function and determine what this
integral converges to.

(a) (𝑥) = 𝑥𝑒 |𝑥| , for all real 𝑥

𝑐𝑜𝑠𝜋𝑥, |𝑥| ≤ 2
(b) 𝑓(𝑥) = {
0, |𝑥| > 2

Badri A, Moges B. and Teklebrhan B. 84 AKU

 APPLIED MATHEMATICS III

Unit Summary:

 Fourier series are infinite series designed to represent general periodic functions in terms
of simple ones, namely, 𝑐𝑜𝑠𝑖𝑛𝑒𝑠 and 𝑠𝑖𝑛𝑒𝑠.
 A function f (x) is called a periodic function of period 𝑝 if f (x) is defined for 𝑥, and

𝑓(𝑥 + 𝑝) = 𝑓(𝑥) and the smallest period 𝑝 is called the fundamental period.

 If f (x) has period 𝑝, it also has the period2𝑝 and in general for any integer 𝑛 ≥ 1
𝑓(𝑥 + 𝑛𝑝) = 𝑓(𝑥) 𝑓𝑜𝑟 𝑎𝑙𝑙 𝑥
 In Fourier series representations, even and odd functions are very important in finding the
Fourier coefficients 𝑎0 ,𝑎𝑛 and 𝑏𝑛
 A function 𝑓(𝑥) is called an Even function if 𝑓(−𝑥) = 𝑓(𝑥) for all x in the domain of 𝑓
 A function 𝑓(𝑥) is called an odd function if 𝑓(−𝑥) = −𝑓(𝑥) for all x in the domain of 𝑓
𝑙 𝑙
For any even function 𝑓(𝑥), ∫−𝑙 𝑓(𝑥) 𝑑𝑥 = 2∫0 𝑓(𝑥) 𝑑𝑥 for any odd function
𝑙
𝑓(𝑥), ∫−𝑙 𝑓(𝑥) 𝑑𝑥 = 0

 If 𝑓(𝑥) is a periodic function of period 𝑃 = 2𝑙 and integrable over the interval (−𝑙. 𝑙).
Then, the Fourier series expansion of 𝑓 is;

𝑛𝜋𝑥 𝑛𝜋𝑥
𝑓(𝑥) = 𝑎𝑜 + ∑∞
𝑛=1(𝑎𝑛 𝑐𝑜𝑠 + 𝑏𝑛 𝑠𝑖𝑛 )
𝑙 𝑙
With coefficients,

1 𝑙
a) 𝑎𝑜 = 2𝑙 ∫−𝑙 𝑓(𝑥) 𝑑𝑥

1 𝑙 𝑛𝜋𝑥
b) 𝑎𝑛 = 𝑙 ∫−𝑙 𝑓(𝑥)𝑐𝑜𝑠 𝑑𝑥 𝑛 = 1,2, …
𝑙

1 𝑙 𝑛𝜋𝑥
c) 𝑏𝑛 = 𝑙 ∫−𝑙 𝑓(𝑥)𝑠𝑖𝑛 𝑑𝑥 𝑛 = 1,2, …
𝑙
 Fourier series are powerful tools for problems involving functions that are periodic or are
of interest on a finite interval only.
 Fourier integrals extend the concept or Fourier series to non periodic functions defined
for all 𝑥.
 The Fourier integral representation of 𝑓(𝑥) can be defined as;
∞
𝑓(𝑥) = ∫0 [𝐴(𝜔) cos 𝜔𝑥 + 𝛽(𝜔) sin 𝜔𝑥]𝑑𝜔

1 ∞ 1 ∞
Where 𝐴(𝜔) = 𝜋 ∫−∞ 𝑓(𝑢) cos 𝜔𝑢𝑑𝑢 𝑎𝑛𝑑 𝐵(𝜔) = 𝜋 ∫−∞ 𝑓(𝑢) sin 𝜔𝑢𝑑𝑢

 The sufficient conductions for which we Fourier integral representation is valid are,

Badri A, Moges B. and Teklebrhan B. 85 AKU

 APPLIED MATHEMATICS III

1. 𝑓(𝑥) is piecewise cortisones on every integral [– 𝑙, 𝑙]

∞
2. 𝑓(𝑥) is absolutely integrals on the real axis, that is ∫−∞/𝑓(𝑥)/𝑑𝑥 converges
3. At every 𝑥 on the real line 𝑓(𝑥) has left and right-hand derivatives

 If f(x) be a real periodic function of period 2𝑙 over the interval (−𝑙. 𝑙).then the complex
Fourier series representation of 𝑓 is defined as.
𝑖𝑛𝜋𝑥
𝑓(𝑥) = lim ∑𝑘𝑛=−𝑘 𝑐𝑛 𝑒 𝑙 for −𝑙 < 𝑥 < 𝑙
𝑘⟶∞

𝑙 −𝑖𝑛𝜋𝑥
1
Where 𝑐𝑛 = 2𝑙 ∫−𝑙 𝑓(𝑥) 𝑒 𝑙 𝑑𝑥 , 𝑛 = 0, ±1, , ±2 are the complex Fourier coefficients

 The complex Fourier integral representation of a function 𝑓 is the integral representation

of the form,
∞
𝑓(𝑥) = ∫−∞ 𝑐(𝜔)𝑒 𝑖𝜔𝑥 𝑑𝜔

1 ∞
Where, 𝑐(𝜔) = 2𝜋 ∫−∞ 𝑓(𝑢)𝑒 −𝑖𝜔𝑢 𝑑𝑢 is the complex Fourier integral coefficient.

Badri A, Moges B. and Teklebrhan B. 86 AKU

 APPLIED MATHEMATICS III

Miscellaneous Exercises

1. Determine whether the following functions are even or odd or neither

(a) 𝑒 𝑥 , 𝑒 −|𝑥| , 𝑥 3 𝑐𝑜𝑠𝑛𝜋𝑥, 𝑥 2 𝑡𝑎𝑛𝜋𝑥, 𝑠𝑖𝑛ℎ𝑥 − 𝑐𝑜𝑠ℎ𝑥

(a) 𝑠𝑖𝑛2 𝑥, 𝑠𝑖𝑛(𝑥 2 ), 𝑙𝑛𝑥, 𝑥⁄(𝑥 2 + 1), 𝑥𝑐𝑜𝑡𝑥

(c) Sums and products of even functions (d) Sums and products of odd functions

(e) Absolute values of odd functions (f) product of an odd and an even functions

2. Find the Fourier series representation of each of the following

−𝜋, −𝜋 < 𝑥 < 0

(a) 𝑓(𝑥) = { (b) 𝑓(𝑥) = 𝜋𝑠𝑖𝑛𝜋𝑥,0 < 𝑥 < 1, 𝑓(𝑥 + 1) = 1
𝑥, 0<𝑥<𝜋
(−1)𝑛
(c) 𝑓(𝑥) = 𝑒 −𝑎𝑥 In the interval (−𝜋, 𝜋) and deduce that 𝑐𝑠𝑐ℎ 𝜋 = ∑∞
𝑛=2 𝑛2 +1

1 1 1 1 𝜋2
(d) 𝑓(𝑥) = 𝑥 − 𝑥 2 , −𝜋 < 𝑥 < 𝜋 and deduce that 12 − 22 + 32 − 42 + ⋯ = 12

1, −𝜋 < 𝑥 < 0
(e) 𝑓(𝑥) = { 𝑓(𝑥 + 2𝜋) = 𝑓(𝑥)
−1, 0<𝑥<𝜋

(f) 𝑓(𝑥) = 𝑥𝑠𝑖𝑛𝑥, −𝜋 < 𝑥 < 𝜋 , 𝑓(𝑥 + 2𝜋) = 𝑓(𝑥)

4𝑥 3
1+ ,−2 < 𝑥 ≤ 0
3
(g) 𝑓(𝑥) = { 4𝑥 3 𝑓(𝑥 + 3) = 𝑓(𝑥)
1− , 0≤𝑥<2
3

3. Find the complex Fourier series representation of each of the following

𝜋
0, −𝜋 < 𝑥 < − 2
𝜋 𝜋 0, 0 < 𝑥 < 1
(a) 𝑓(𝑥) = 1, − 2 < 𝑥 < (b) 𝑓(𝑥) = { 2
𝜋
2 𝑘 ,1 < 𝑥 < 8
{ 0, − 2 < 𝑥 < 𝜋

𝜋, − 𝜋 < 𝑥 < 0 𝑥, 0 < 𝑥 < 1

(c) 𝑓(𝑥) = { (d) 𝑓(𝑥) = {
𝜋 − 𝑥, 0 ≤ 𝑥 ≤ 𝜋 0, 1 < 𝑥 < 2
0, −𝑙 <𝑥 ≤0
4. Find the Fourier series representation of 𝑓(𝑥) = {
𝐸𝑠𝑖𝑛𝑤𝑥, 0 < 𝑥 < 𝑙

Badri A, Moges B. and Teklebrhan B. 87 AKU

 APPLIED MATHEMATICS III

5. Let 𝑓 be aperiodic function of period 2𝜋 such that 𝑓(𝑥) = 𝜋 2 − 𝑥 2 for 𝑥 ∈ (−𝜋, 𝜋), then
show that
2𝜋 2 −4
𝜋2 − 𝑥2 = + ∑∞ 𝑛
𝑛=1 𝑛2 (−) 𝑐𝑜𝑠𝑛𝑥
3

6. Show for 𝑥 ∈ (−𝜋, 𝜋) that

−6 6 𝜋2
(a) 3𝑥 = ∑∞
𝑛=1 (−1)𝑛 sin 𝑛𝑥 (b) 𝑥 3 = ∑∞ 𝑛
𝑛=1 2 (−1) (𝑛3 − ) sin 𝑛𝑥
𝑛 𝑛

7. Using the Fourier integral representation, show that

𝜋
∞ 1−𝑐𝑜𝑠𝜋𝜔 , 0<𝑥≤𝜋
(a) ∫0 sin(𝑥𝜔) 𝑑𝜔 = {2
𝜔
0, 𝑥>𝜋
1
∞ sin 𝜋𝜔 sin 𝜔𝑥 𝜋 sin 𝑥, 0 ≤ 𝑥 ≤ 𝜋
(b) ∫0 𝑑𝜔 = {2
1−𝜔 2
0, 𝑥>𝜋
∞ 1, 0 < 𝑎 < 1
8. If ∫0 𝑓(𝑥) sin 𝑎𝑥 𝑑𝑥 = { then find 𝑓(𝑥).
0, 𝑎>1

9. Find the complex Fourier integral representation of

𝜋
𝑐𝑜𝑠𝑥, 0≤𝑥≤ 2
𝜋
(a) 𝑓(𝑥) = 𝑠𝑖𝑛𝑥, − 2 < 𝑥 < 0 (b) 𝑓(𝑥) = 𝑘𝑥𝑒 |𝑥| for all real 𝑥
𝜋
{ 0, |𝑥| >
2

Badri A, Moges B. and Teklebrhan B. 88 AKU

 APPLIED MATHEMATICS III

References

Allan pinkas, Fourier Series and Integral transform, Cambridge university press, 1997
Alan Jeffrey, Advanced engineering mathematics, RR Donnelley & Sons, Inc, 2002
Abramowitz, M. and I. A. Stegun (eds.), Handbook of Mathematical Functions. 10th
Courant, R., Differential and Integral Calculus. 2 vols. Hoboken, NJ: Wiley, 1988.
Churchill, R. V., Operational Mathematics. 3rd ed.New York: McGraw-Hill, 1972.
Erwin Kreyzing, Advanced engineering mathematics, 10th ed, wiley, 2000
G.B. Foland, Fourier Analysis and its applications, Wadsworth and Brooks/Cole, Pacific
Grove,CA, 1992
Graham, R. L. et al., Concrete Mathematics. 2nd ed. Reading, MA: Addison-Wesley,
1994.
Hanna, J. R. and J. H. Rowland, Fourier Series, Transforms, and Boundary Value Problems.
2nd ed.New York: Wiley, 2008.
Jerri, A. J., The Gibbs Phenomenon in Fourier Analysis, Splines, and Wavelet
Approximations. Boston: Kluwer, 1998.
Szegö, G., Orthogonal Polynomials. 4th ed. Reprinted. New York: American
Mathematical Society, 2003.
Tolstov, G. P., Fourier Series. New York: Dover, 1976
Thomas, G. et al., Thomas’ Calculus, Early Transcendental Update. 10th ed. Reading,
MA: Addison-Wesley, 2003.
W. Brown and R. V. Churchill, Fourier Series and Boundary Value Problems, 5th ed.,McGraw-Hill,
New York, 1993
Zygmund, Trigonometric Series, 2nd ed. (Volumes I and II combined),Cambridge University
Press, Cambridge, UK, 1988

Badri A, Moges B. and Teklebrhan B. 89 AKU

 APPLIED MATHEMATICS III

Unit-four
Fourier and Laplace Transformation
Introduction

An integral transform is a transformation that produces from a given function a new function,
that depends on a different variable and appears in the form of an integral. These transformations
are mainly employed as a tool to solve certain initial and boundary value problems in ordinary
and partial differential equations arising in many areas of science and engineering. Fourier
transforms are integral transforms which are of vital importance from the applications view point
in solving initial and boundary value problems.

In this chapter we will discuss three transforms: the Fourier cosine transform, the Fourier sine
transform; the first two being real and the later one complex. These transforms are obtained from
the corresponding Fourier integral. We will also see Laplace transforms, inverse Laplace
transform, differentiation and integration of Laplace transforms convolution and integral
equations

Unit Objectives:

At the end of this unit each student should able to:

 Know representation of the Fourier cosine and sine transforms of a function;

 Understand properties of Fourier cosine and sine transforms
 Realize and learn about Fourier transformation and its properties.
 understand and Learn about Laplace transformation of exponentially ordered functions
 Understand and develop the inverse Laplace transformation of function.
 Know and understand about how to differentiate the Laplace transform of a function.
 Develop and understand about integration of the Laplace transform
 Find and understand convolution and convolution theorem
 Understand about integral equations

Badri A, Moges B. and Teklebrhan B. 90 AKU

 APPLIED MATHEMATICS III

4.1 Fourier Cosine and Fourier Sine Transforms

Overview:
In this section, we are going to deal with the Fourier Cosine and Fourier Sine Transforms with
consideration of various examples

Section Objectives:

At the end of this subtopic, students will be able to:

 Define Fourier Cosine and Fourier Sine Transforms;

 Understand properties of Fourier Cosine and Fourier Sine Transforms ;

The Fourier Cosine and Sine Transforms can be considered as a special cases of the Fourier
transform 𝑜𝑓 𝑓(𝑥) when 𝑓(𝑥) is even or odd function over the real axis

Definition: If 𝑓(𝑥) is piecewise continuous on each finite interval [0, 𝑙] and absolutely integrable
over the positive real axis so that its Fourier Transform 𝐹(𝑤) exists then the Fourier cosine and
Fourier sine transforms of f(x) denoted by 𝐹𝐶 (𝑤) 𝑜𝑟𝑓𝑐 ^ 𝑎𝑛𝑑 𝐹𝑠 (𝑤) 𝑜𝑟 𝑓𝑠 ^ respectively is defined
as;

2 ∞
𝐹𝑐 (𝑤) = √𝜋 ∫0 𝑓(𝑥) 𝑐𝑜𝑠𝑤𝑥𝑑𝑥,

2 ∞
𝐹𝑠 (𝑤) = √𝜋 ∫0 𝑓(𝑥) 𝑠𝑖𝑛𝑤𝑥𝑑𝑥

1, 0 ≤ 𝑥 ≤ 𝑎
Example 4.1: find the Fourier cosine and sine transform of 𝑓(𝑥) = {
0, 𝑥>𝑎

2 ∞
Solution; by definition 𝐹𝐶 (𝑤) = √𝜋 ∫0 𝑓(𝑥) 𝑐𝑜𝑠𝑤𝑥𝑑𝑥

2 ∞ 2 𝑎 2 𝑠𝑖𝑛𝑤𝑥 𝑎 2 𝑠𝑖𝑛𝑎𝑤
= √𝜋 ∫0 1 𝑐𝑜𝑠𝑤𝑥𝑑𝑥 = √𝜋 ∫0 𝑐𝑜𝑠𝑤𝑥𝑑𝑥 = √𝜋 [ ] = √𝜋 𝑤
𝑤 0

2 𝑠𝑖𝑛𝑎𝑤
Hence, 𝐹𝐶 (𝑤) = √
𝜋 𝑤

2 ∞
Similarly, 𝐹𝑠 (𝑤) = √𝜋 ∫0 𝑓(𝑥) 𝑠𝑖𝑛𝑤𝑥𝑑𝑥

Badri A, Moges B. and Teklebrhan B. 91 AKU

 APPLIED MATHEMATICS III

2 ∞ 2 ∞ 2 −𝑐𝑜𝑠𝑤𝑥 𝑎 2 1−𝑐𝑜𝑠𝑎𝑤
= √𝜋 ∫0 1 𝑠𝑖𝑛𝑤𝑥𝑑𝑥 = √𝜋 ∫0 𝑠𝑖𝑛𝑤𝑥 𝑑𝑥 = √𝜋 [ ] = √𝜋 ( 𝑤 )
𝑤 0

2 1−𝑐𝑜𝑠𝑎𝑤
Hence, 𝐹𝑠 (𝑤) = √𝜋 ( )
𝑤

𝑐𝑜𝑠𝑥, 0 ≤ 𝑥 ≤ 𝑎
Example 4.2: find the Fourier cosine and sine transform of 𝑓(𝑥) = {
0, 𝑥>0

2 ∞ 2 𝑎
Solution; by definition 𝐹𝑐 (𝑤) = √𝜋 ∫0 𝑓(𝑥) 𝑐𝑜𝑠𝑤𝑥𝑑𝑥 = √𝜋 ∫0 𝑐𝑜𝑠𝑥 𝑐𝑜𝑠𝑤𝑥𝑑𝑥

2 𝑎1
= √𝜋 ∫0 2 (cos(1 − 𝑤)𝑥 + cos(1 + 𝑤)𝑥)𝑑𝑥 (𝑎𝑠 𝑐𝑜𝑠𝑥𝑐𝑜𝑠𝑦 = 1⁄2 (cos(𝑥 − 𝑦) + cos(𝑥 + 𝑦))

1 2 𝑎 1 2 𝑎
= 2 √𝜋 ∫0 (cos(1 − 𝑤)𝑥𝑑𝑥 + 2 √𝜋 ∫0 (cos(1 + 𝑤)𝑥𝑑𝑥

1 2 sin(1−𝑤)𝑥 𝑎 1 2 sin(1+𝑤)𝑥 𝑎 1 2 sin(1−𝑤)𝑎 1 2 sin(1+𝑤)𝑎

= 2 √𝜋 [ ]0 + 2 √𝜋 [ ]0 = √4 𝜋 ( ) + √4 𝜋 ( )
1−𝑤 1+𝑤 1−𝑤 1+𝑤

1 sin a(1−𝑤) sin a(1+𝑤)

= ( +( )
√2𝜋 1−𝑤 1+𝑤

2 ∞ 2 𝑎
𝐹𝑠 (𝑤) = √𝜋 ∫0 𝑓(𝑥) 𝑠𝑖𝑛𝑤𝑥𝑑𝑥 = √𝜋 ∫0 𝑐𝑜𝑠𝑥 𝑠𝑖𝑛𝑤𝑥𝑑𝑥

2 𝑎1
= √𝜋 ∫0 2 (sin(1 − 𝑤) 𝑥 + sin(1 + 𝑤)𝑥)𝑑𝑥

1 2 𝑎 1 2 𝑎1
= 2 √𝜋 ∫0 (sin(1 − 𝑤) 𝑥𝑑𝑥 + 2 √𝜋 ∫0 2 (sin(1 + 𝑤) 𝑥𝑑𝑥

1 2 −cos(1−𝑤)𝑥 𝑎 1 2 cos(1+𝑤)𝑥 𝑎
= 2 √𝜋 [ ] + 2 √𝜋 [ 1+𝑤 ]
1−𝑤 0 0

1 2 −cos(1−𝑤)𝑎 cos0 1 2 −cos(1+𝑤)𝑎 cos0

= 2 √𝜋 [ + 1−𝑤] + 2 √𝜋 [ + 1+𝑤]
1−𝑤 1+𝑤

1 2 −cos(1−𝑤)𝑎 1 cos(1+𝑤)𝑎 1
√ ( + 1−𝑤 + − 1+𝑤)
2 𝜋 1−𝑤 1+𝑤

1 cos(1+𝑤)𝑎 cos(1−𝑤)𝑎 1 1
= ( − + 1−𝑤 − 1+𝑤)
√2𝜋 1+𝑤 1−𝑤

Badri A, Moges B. and Teklebrhan B. 92 AKU

 APPLIED MATHEMATICS III

1 cos(1+𝑤)𝑎 cos(1−𝑤)𝑎 2w
= ( − + 𝑤2 −1)
√2𝜋 1∓𝑤 1−𝑤

2 w 1 cos(1+𝑤)𝑎 cos(1−𝑤)𝑎
= √𝜋 𝑤2 −1 + ( − +)
√2𝜋 1+𝑤 1−𝑤

2 w 1 cos(1+𝑤)𝑎 cos(1−𝑤)𝑎
Hence, 𝐹𝑠 (𝑤) = √𝜋 𝑤2 −1 + ( − )
√2𝜋 1+𝑤 1−𝑤

Like Fourier transform the Fourier cosine and sine transforms also satisfy certain properties
which are useful from application point of view.

Property1 (Linearity): for any two functions 𝑓(𝑥) and 𝑔(𝑥) whose Fourier cosine and sine
transform exist and for any constants a and b

(a) 𝑓𝑐 ^ [𝑎𝑓(𝑥) + 𝑏𝑔(𝑥)] = 𝑎𝑓𝑐 ^ [𝑓(𝑥)] + 𝑏𝑓𝑐 ^ [𝑔(𝑥)] and

(𝑏) 𝑓𝑠 ^ [𝑎𝑓(𝑥) + 𝑏𝑔(𝑥)] = 𝑎𝑓𝑠 ^ [𝑓(𝑥)] + 𝑏𝑓𝑠 ^ [𝑔(𝑥)]

Proof:

2 ∞
(a) By definition 𝑓𝑐 ^ [𝑎𝑓(𝑥) + 𝑏𝑔(𝑥)] = √𝜋 ∫0 (𝑎𝑓(𝑥) + 𝑏𝑔(𝑥))𝑐𝑜𝑠𝑤𝑥𝑑𝑥

2 ∞ 2 ∞
= √𝜋 ∫0 𝑎𝑓(𝑥)𝑐𝑜𝑠𝑤𝑥𝑑𝑥 + √𝜋 ∫0 𝑏𝑔(𝑥)𝑐𝑜𝑠𝑤𝑥𝑑𝑥

2 ∞ 2 ∞
= 𝑎√ ∫ 𝑓(𝑥)𝑐𝑜𝑠𝑤𝑥𝑑𝑥 + 𝑏√ ∫ 𝑔(𝑥)𝑐𝑜𝑠𝑤𝑥𝑑𝑥 = 𝑎𝑓𝑐 ^ [𝑓(𝑥)] + 𝑏𝑓𝑐 ^ [𝑔(𝑥)]
𝜋 0 𝜋 0

Hence, 𝑓𝑐 ^ [𝑎𝑓(𝑥) + 𝑏𝑔(𝑥)] = 𝑎𝑓𝑐 ^ [𝑓(𝑥)] + 𝑏𝑓𝑐 ^ [𝑔(𝑥)]

2 ∞
(b) Similarly, by definition 𝑓𝑠 ^ [𝑎𝑓(𝑥) + 𝑏𝑔(𝑥)] = √𝜋 ∫0 (𝑎𝑓(𝑥) + 𝑏𝑔(𝑥))𝑠𝑖𝑛𝑤𝑥𝑑𝑥

2 ∞ 2 ∞
= √𝜋 ∫0 𝑎𝑓(𝑥)𝑠𝑖𝑛𝑤𝑥𝑑𝑥 + √𝜋 ∫0 𝑏𝑔(𝑥)𝑠𝑖𝑛𝑤𝑥𝑑𝑥

2 ∞ 2 ∞
= 𝑎√𝜋 ∫0 𝑓(𝑥)𝑠𝑖𝑛𝑤𝑥𝑑𝑥 + 𝑏√𝜋 ∫0 𝑔(𝑥)𝑠𝑖𝑛𝑤𝑥𝑑𝑥 = 𝑎𝑓𝑠 ^ [𝑓(𝑥)] + 𝑏𝑓𝑠 ^ [𝑔(𝑥)]

Hence, 𝑓𝑠 ^ [𝑎𝑓(𝑥) + 𝑏𝑔(𝑥)] = 𝑎𝑓𝑠 ^ [𝑓(𝑥)] + 𝑏𝑓𝑠 ^ [𝑔(𝑥)]

Badri A, Moges B. and Teklebrhan B. 93 AKU

 APPLIED MATHEMATICS III

Property2 (shifting 𝑤 by 𝑤0 and scaling 𝑥 𝑏𝑦 𝑎)

If 𝐹𝑐 (𝑤) and 𝐹𝑠 (𝑤) are the Fourier cosine and sine transforms of 𝑓(𝑥),then
1
a) 𝑓𝑐 ^ [cos(𝑤0 𝑥) 𝑓(𝑥)] = 2 [𝐹𝑐 (𝑤 + 𝑤0 ) + 𝐹𝑐 (𝑤 − 𝑤0 )]

1
b) 𝑓𝑐 ^ [sin(𝑤0 𝑥) 𝑓(𝑥)] = 2 [𝐹𝑠 (𝑤 + 𝑤0 ) + 𝐹𝑠 (𝑤 − 𝑤0 )]

1
c) 𝑓𝑠 ^ [cos(𝑤0 𝑥) 𝑓(𝑥)] = 2 [𝐹𝑠 (𝑤 + 𝑤0 ) + 𝐹𝑠 (𝑤 − 𝑤0 )]

1
d) 𝑓𝑠 ^ [sin(𝑤0 𝑥) 𝑓(𝑥)] = 2 [𝐹𝑐 (𝑤 − 𝑤0 ) + 𝐹𝑐 (𝑤 + 𝑤0 )]

1
e) 𝑓𝑐 ^ [𝑓(𝑎𝑥)] = 𝑎 𝐹𝑐 (𝑤⁄𝑎), 𝑎 > 0

1
f) 𝑓𝑠 ^ [𝑓(𝑎𝑥)] = 𝑎 𝐹𝑠 (𝑤⁄𝑎), 𝑎 > 0

These results follow directly from the definitions of the Fourier cosine and sine transforms

For example let us proof (b)

Proof;

2 ∞
By definition 𝑓𝑐 ^ [sin(𝑤0 𝑥) 𝑓(𝑥)] = √𝜋 ∫0 sin(𝑤0 𝑥) cos(𝑤0 𝑥) 𝑓(𝑥) 𝑑𝑥

But recall that from the trigonometric identities,

1 1
sin 𝑤0 𝑥 𝑐𝑜𝑠𝑤𝑥 = 2 [sin(𝑤0 + 𝑤)𝑥 + sin(𝑤0 − 𝑤)𝑥] = 2 [sin(𝑤0 + 𝑤)𝑥 −sin(𝑤 − 𝑤0 )𝑥]

2 ∞
Thus, 𝑓𝑐 ^ [sin(𝑤0 𝑥) 𝑓(𝑥)] = √ ∫0 sin(𝑤0 𝑥) cos(𝑤0 𝑥) 𝑓(𝑥) 𝑑𝑥
𝜋

1 2 ∞
= 2 [√𝜋 ∫0 𝑠𝑖𝑛(𝑤0 + 𝑤)𝑓(𝑥)𝑥 − sin(𝑤 − 𝑤0 )𝑓(𝑥)𝑥 𝑑𝑥]

1 2 ∞ 2 ∞
= 2 [√𝜋 ∫0 𝑠𝑖𝑛(𝑤0 + 𝑤)𝑓(𝑥)𝑥𝑑𝑥 − √𝜋 ∫0 𝑠𝑖𝑛(𝑤 − 𝑤0 )𝑓(𝑥)𝑥𝑑𝑥

1
=2 [ 𝐹𝑠 (𝑤 + 𝑤0 ) + 𝐹𝑠 (𝑤 − 𝑤0 )

Hence proved.

2 ∞ 1 2 ∞ 𝑤 1
(c) By definition 𝑓𝑐 ^ [𝑓(𝑎𝑥)] = √𝜋 ∫0 𝑓(𝑎𝑥) 𝑐𝑜𝑠𝑤𝑥𝑑𝑥 = 𝑎 √𝜋 ∫0 𝑓(𝑥) 𝑐𝑜𝑠 𝑎 𝑥𝑑𝑥 = 𝑎 𝐹𝑐 (𝑤⁄𝑎)

Badri A, Moges B. and Teklebrhan B. 94 AKU

 APPLIED MATHEMATICS III

(a),(c).(d) and (f) are left as an exercise for you.

Property 3(Fourier cosine and sine transforms of derivatives)

Let 𝑓(𝑥) and 𝑓′(𝑥) be continuous and absolutely integrable on the interval[0, ∞) and 𝑓′′(𝑥) be
piecewise continuous on every subinterval[0, 𝑙), then

2
a) 𝑓𝑐 ^ [𝑓′(𝑥)] = 𝑤𝐹𝑠 (𝑤) − √𝜋 𝑓(0)

b) 𝑓𝑠 ^ [𝑓′(𝑥)] = −𝑤𝐹𝑐 (𝑤)

2
c) 𝑓𝑐 ^ [𝑓′′(𝑥)] = −𝑤 2 𝐹𝑐 (𝑤) − √𝜋 𝑓′(0)

2
d) 𝑓𝑠 ^ [𝑓′′(𝑥)] = −𝑤 2 𝐹𝑠 (𝑤) + 𝑤√𝜋 𝑓(0)

Proof

2 ∞ 2 ∞
(a) By definition 𝑓𝑐 ^ [𝑓′(𝑥)] = √𝜋 ∫0 𝑓′(𝑥) 𝑐𝑜𝑠𝑤𝑥𝑑𝑥 = √𝜋 [[𝑓(𝑥)𝑐𝑜𝑠𝑤𝑥] +
0
∞
𝑤 ∫0 𝑓(𝑥) 𝑠𝑖𝑛𝑤𝑥𝑑𝑥]

2
= w 𝑤𝐹𝑠 (𝑤) − √𝜋 𝑓(0), this by assuming that 𝑓(𝑥) → 0 as 𝑥 → ∞

2
Hence 𝑓𝑐 ^ [𝑓′(𝑥)] = 𝑤𝐹𝑠 (𝑤) − √𝜋 𝑓(0)

The result (b) can be proved on the similar lines as in (a)

Proof of (c)

2 ∞
By definition 𝑓𝑐 ^ [𝑓′′(𝑥)] = √𝜋 ∫0 𝑓′′(𝑥) 𝑐𝑜𝑠𝑤𝑥𝑑𝑥

2 ∞ ∞
= √𝜋 [[𝑓 ′ (𝑥)𝑐𝑜𝑠𝑤𝑥 + 𝑤𝑓(𝑥)𝑠𝑖𝑛𝑤𝑥] − 𝑤 2 ∫0 𝑓(𝑥) 𝑐𝑜𝑠𝑤𝑥𝑑𝑥]
0

2
=−𝑤 2 𝐹𝑐 (𝑤) − √𝜋 𝑓′(0), this is by assuming that 𝑓(𝑥), 𝑓′(𝑥) → 0 as 𝑥 → ∞

The result (d) can be proved on the similar lines as in (c)

Badri A, Moges B. and Teklebrhan B. 95 AKU

 APPLIED MATHEMATICS III

Example 4.3:find the Fourier cosine and sine transform of 𝑓(𝑥) = 𝑒 −𝑎𝑥 , 𝑥 ≥ 0, 𝑎 > 0, by using
the Fourier cosine and sine transforms of derivatives.

Solution; here 𝑓(𝑥) = 𝑒 −𝑎𝑥 , this gives 𝑓′(𝑥) = −𝑎𝑒 −𝑎𝑥 and 𝑓 ′′ (𝑥) = 𝑎2 𝑒 −𝑎𝑥

Thus, by property 3, part c, we have

𝑓𝑐 ^ [𝑓′′(𝑥)] = 𝑓𝑐 ^ [𝑎2 𝑒 −𝑎𝑥 ] = 𝑎2 𝑓𝑐 ^ [𝑒 −𝑎𝑥 ] = 𝑎2 𝐹𝑐 (𝑤) (1)

Where 𝐹𝑐 (𝑤) denotes the Fourier cosine transform of 𝑓(𝑥) = 𝑒 −𝑎𝑥

2 2
Also, 𝑓𝑐 ^ [𝑓′′(𝑥)] = −𝑤 2 𝐹𝑐 (𝑤) − √𝜋 𝑓 ′ (0) = −𝑤 2 𝐹𝑐 (𝑤) + 𝑎√𝜋, since 𝑓 ′ (0) = 𝑎 (2)

From (1) and (2), we have

2 2 𝑎
𝑎2 𝐹𝑐 (𝑤) = −𝑤 2 𝐹𝑐 (𝑤) + 𝑎√𝜋 or 𝐹𝑐 (𝑤) = √𝜋 𝑤2 +𝑎2

2 𝑎
Hence 𝑓𝑐 ^ [𝑓(𝑥)] = 𝑓𝑐 ^ [𝑒 −𝑎𝑥 ] = √𝜋 𝑤2 +𝑎2

Similarly, to find the Fourier sine transform,

^
Consider, 𝑓𝑠 ^ [𝑓′′(𝑥)] = 𝑓𝑠 ^ [𝑎2 𝑒 −𝑎𝑥 ] = 𝑎2 𝑓𝑠 [𝑒 −𝑎𝑥 ] (3)

Where 𝐹𝑠 (𝑤) denotes the Fourier sine transform of 𝑓(𝑥) = 𝑒 −𝑎𝑥

2 2
Also, 𝑓𝑠 ^ [𝑓′′(𝑥)] = −𝑤 2 𝐹𝑠 (𝑤) + 𝑤√𝜋 𝑓(0) = −𝑤 2 𝐹𝑠 (𝑤) + 𝑤√𝜋 (4)

From (3) and (4),we have

^ 2 2 𝑤
𝑎2 𝑓𝑠 [𝑒 −𝑎𝑥 ] = −𝑤 2 𝐹𝑠 (𝑤) + 𝑤√𝜋 or 𝐹𝑠 (𝑤) = √𝜋 𝑤2 +𝑎2

2 𝑤
Hence 𝑓𝑠 ^ [𝑓(𝑥)] = 𝑓𝑠 ^ [𝑒 −𝑎𝑥 ] = √𝜋 𝑤2 +𝑎2

Badri A, Moges B. and Teklebrhan B. 96 AKU

 APPLIED MATHEMATICS III

Exercises 4.1

1. Find the Fourier cosine and sine transform of each of the following

𝑐𝑜𝑠𝑥, 0 ≤ 𝑥 ≤ 𝑎 ∞
a) 𝑓(𝑥) = 𝑒 −𝑥 , 𝑥 > 0 (b). 𝑓(𝑥) = { c. ∫0 𝑓(𝑥)𝑔(𝑥)
0, 𝑥>𝑎

2. Find the Fourier cosine and sine transform of each of the following functions

a) 𝑓(𝑥) = 1,0 < 𝑥 < 1 b) 𝑓(𝑥) = −1, 1 < 𝑥 < 2

3. Explain why the following functions have neither Fourier cosine transform nor Fourier sine
transform

a) 𝑓(𝑥) = 1 b) 𝑓(𝑥) = 𝑒 𝑥

Badri A, Moges B. and Teklebrhan B. 97 AKU

 APPLIED MATHEMATICS III

4.2 Fourier transformation

Overview:
In this section, we are going to deal with the Fourier Transforms and its properties with various
examples

Section Objectives:

At the end of this subtopic, students will be able to:

 Define and identify Fourier Transforms of function;

 Know and Understand about properties of Fourier Transforms ;

Fourier transforms of a function f (𝑥) can be derived from the complex Fourier integral
representation of 𝑓(𝑥) on the real line, that is,

Recall the complex Fourier integral representation of 𝑓(𝑥) on the real line
∞ 1 ∞ ∞
𝑓(𝑥) = ∫−∞ 𝑐(𝑤)𝑒 𝑖𝑤𝑥 𝑑𝑤 = 2𝜋 ∫−∞ ∫−∞ 𝑓(𝑢)𝑒 −𝑖𝑤(𝑢−𝑥) 𝑑𝑢𝑑𝑤

1 ∞
Where, 𝑐(𝑤) = 2𝜋 ∫−∞ 𝑓(𝑢)𝑒 −𝑖𝑤𝑢 𝑑𝑢 and taking 𝜔 = 𝑤

1 ∞ ∞ 1 ∞ 1 ∞
⟹ 𝑓(𝑥) = 2𝜋 ∫−∞ ∫−∞ 𝑓(𝑢)𝑒 −𝑖𝑤(𝑢−𝑥) 𝑑𝑢𝑑𝑤 = ∫ [ ∫ 𝑓(𝑢) 𝑒 −𝑖𝑤𝑢 ] 𝑒 𝑖𝑖𝑤𝑥 (1)
√2𝜋 −∞ √2𝜋 −∞

Here, the expression in the bracket, a function of 𝑤 denoted by 𝐹(𝑤) is called the Fourier
Transform of 𝑓 and since 𝑢 is a dummy variable, we replace 𝑢 by 𝑥 and have
1 ∞
𝐹(𝑤) = ∫ 𝑓(𝑢) 𝑒 −𝑖𝑤𝑥 𝑑𝑥 so that (1) becomes
√2𝜋 −∞

1 ∞ ∞ 1 ∞
𝑓(𝑥) = 2𝜋 ∫−∞ ∫−∞ 𝑓(𝑢)𝑒 −𝑖𝑤(𝑢−𝑥) 𝑑𝑢𝑑𝑤 = ∫ 𝐹(𝑤) 𝑒 −𝑖𝑤𝑥 𝑑𝑤 and is called the inverse
√2𝜋 −∞
Fourier Transform of 𝐹(𝑤).

Other common notations used for Fourier transform of 𝑓(𝑥) are 𝑓^ (𝑤) or ℱ(𝑓(𝑥)).

In general we define the Fourier transform of a function f as follows

Definition (Fourier transform): The Fourier transform denoted by 𝐹(𝑤) 𝑜𝑟 ℱ(𝑓(𝑥))of a function
𝑓(𝑥) is defined as
1 ∞
𝐹(𝑤) = ∫ 𝑓(𝑥) 𝑒 −𝑖𝑤𝑥 𝑑𝑥
√2𝜋 −∞

Badri A, Moges B. and Teklebrhan B. 98 AKU

 APPLIED MATHEMATICS III

The sufficient conditions for the existence of the Fourier transform of 𝑓(𝑥)are:

1. 𝑓(𝑥) is piecewise continuous on every finite interval;

2. 𝑓(𝑥) is absolutely integrable on the real axis.

𝑘, 0 < 𝑥 < 𝑎
Example 4.4: find the Fourier transform 𝑓(𝑥) = {
0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
1 ∞ 1 𝑎
Solution: By definition ℱ(𝑓(𝑥)) = ∫ 𝑓(𝑥) 𝑒 −𝑖𝑤𝑥 𝑑𝑥 = √2𝜋 ∫0 𝑘 𝑒 −𝑖𝑤𝑥 𝑑𝑥
√2𝜋 −∞

𝑘 𝑎 𝑘 𝑒 −𝑖𝑤𝑥 𝑎 𝑘
= ∫ 𝑘 𝑒 −𝑖𝑤𝑥 𝑑𝑥 = √2𝜋 [
√2𝜋 0 −𝑤
] = 𝑖𝑤√2𝜋 (1 − 𝑒 −𝑖𝑤𝑎 )
0
𝑘
Hence, 𝐹(𝑤) = 𝑖𝑤√2𝜋 (1 − 𝑒 −𝑖𝑤𝑎 )

1, |𝑥| ≤ 𝑎
Example 4.5 : find the Fourier Transform of 𝑓(𝑥) = {
0, |𝑥| > 𝑎
1 ∞ 1 𝑎
Solution: By definition ℱ(𝑓(𝑥)) = ∫ 𝑓(𝑥) 𝑒 −𝑖𝑤𝑥 𝑑𝑥
√2𝜋 −∞
= ∫ 1 𝑒 −𝑖𝑤𝑥 𝑑𝑥
√2𝜋 −𝑎

1 𝑎 1 𝑒 −𝑖𝑤𝑥 𝑎 1 𝑒 𝑖𝑤𝑎 −𝑒 −𝑖𝑤𝑎

= ∫ 𝑒 −𝑖𝑤𝑥 𝑑𝑥 = √2𝜋 [
√2𝜋 −𝑎 −𝑖𝑤
]
−𝑎
= 𝑤√2𝜋 [ 𝑖
]

1 𝑐𝑜𝑠𝑤𝑎+𝑖𝑠𝑖𝑛𝑤𝑎−𝑐𝑜𝑠𝑤𝑎+𝑖𝑠𝑖𝑛𝑤𝑎
= [ ]
𝑤√2𝜋 𝑖

1 2𝑖𝑠𝑖𝑛𝑤𝑎 2𝑠𝑖𝑛𝑤𝑎 2 𝑠𝑖𝑛𝑤𝑎

= 𝑤√2𝜋 = =
𝑖 𝑤√2𝜋 √2𝜋 𝑤

4 𝑠𝑖𝑛𝑤𝑎 2 𝑠𝑖𝑛𝑤𝑎
= √2𝜋 = √𝜋
𝑤 𝑤

2 𝑠𝑖𝑛𝑤𝑎
Hence: 𝐹(𝑤) = √𝜋 𝑤

Example 4.6: find the Fourier transform of 𝑓(𝑥) = 𝑒 −|𝑥|

𝑒 𝑥 , −∞ < 𝑥 ≤ 0
Solution: The function can also be written as 𝑓(𝑥) = 𝑒 −|𝑥| = { (by the
𝑒 −𝑥 , 0 < 𝑥 < ∞
definition of absolute value)
1 ∞
Now by the definition of Fourier transform, 𝐹(𝑤) = ∫ 𝑓(𝑥) 𝑒 −𝑖𝑤𝑥 𝑑𝑥
√2𝜋 −∞

1 0 ∞
⇒ 𝐹(𝑤) = [∫−∞ 𝑒 𝑥 𝑒 −𝑖𝑤𝑥 𝑑𝑥 + ∫0 𝑒 −𝑥 𝑒 −𝑖𝑤𝑥 𝑑𝑥]
√2𝜋

Badri A, Moges B. and Teklebrhan B. 99 AKU

 APPLIED MATHEMATICS III

1 0 ∞
= [∫−∞ 𝑒 (1−𝑖𝑤)𝑥 𝑑𝑥 + ∫0 𝑒 −(1−𝑖𝑤𝑥) 𝑑𝑥]
√2𝜋

1 0 ∞
= [∫−∞ 𝑒 (1−𝑖𝑤)𝑥 𝑑𝑥 + ∫0 𝑒 −(1−𝑖𝑤𝑥) 𝑑𝑥]
√2𝜋

1 𝑒 (1−𝑖𝑤)𝑥 0 1 𝑒 −(1+𝑖𝑤)𝑥 ∞ 1 1 1 1
= [ (1−𝑖𝑤) ] − [ (1+𝑖𝑤) ] = [(1−𝑖𝑤)] + [(1+𝑖𝑤)]
√2𝜋 −∞ √2𝜋 0 √2𝜋 √2𝜋

𝑒 (1−𝑖𝑤)𝑥 1 𝑒 −(1+𝑖𝑤)𝑥 1
Here, as x→0 (1−𝑖𝑤)
→ (1−𝑖𝑤) and as x→ −∞ (1+𝑖𝑤)
→ (1+𝑖𝑤)

𝑒 −(1+𝑖𝑤)𝑥 𝑒 (1−𝑖𝑤)𝑥
Similarly, as x→ ∞ (1+𝑖𝑤)
→ 0 and as x→ 0 (1−𝑖𝑤)
→0

So, the above expression becomes;

1 𝑒 (1−𝑖𝑤)𝑥 0 1 𝑒 −(1+𝑖𝑤)𝑥 ∞ 1 1 1
[ (1−𝑖𝑤) ] − [ (1+𝑖𝑤) ] = [(1−𝑖𝑤) − (1+𝑖𝑤)]
√2𝜋 −∞ √2𝜋 0 √2𝜋

1 1+𝑖𝑤+1−𝑖𝑤 1 2 2 1
= [ ]= = √𝜋 1+𝑤2
√2𝜋 1+𝑤 2 √2𝜋 1+𝑤 2

2 1
Hence, 𝐹(𝑤) = √𝜋 1+𝑤2

2
Example 4.7: find the Fourier transform of 𝑓(𝑥) = 𝑒 −𝑎𝑥 , 𝑎 > 0
1 ∞ 1 ∞ 2
Solution by definition ℱ(𝑓(𝑥)) = ∫ 𝑓(𝑥) 𝑒 −𝑖𝑤𝑥 𝑑𝑥 = √2𝜋 ∫−∞ 𝑒 −𝑎𝑥 𝑒 −𝑖𝑤𝑥 𝑑𝑥
√2𝜋 −∞

𝑖𝑤 2 𝑖𝑤 2
1 ∞ −(𝑎𝑥 2 +𝑖𝑤𝑥) 1 ∞ −[(√𝑎𝑥+ ) +( ) ]
= ∫−∞
𝑒 𝑑𝑥 = ∫−∞
𝑒 2√𝑎 2√𝑎 𝑑𝑥
√2𝜋 √2𝜋

−𝑤2 𝑖𝑤 2
1 ∞ −((√𝑎𝑥+ ) )
= 𝑒 4𝑎 ∫−∞ 𝑒 2√𝑎 𝑑𝑥
√2𝜋

Now using integration by substitution, we have

𝑖𝑤
Let 𝑡 = √𝑎𝑥 + 2 , then 𝑑𝑡 = √𝑎𝑑𝑥
√𝑎

−𝑤2 𝑖𝑤 2 −𝑤2
1 ∞ −((√𝑎𝑥+ ) ) 1 ∞ 2 1
⟹ ℱ(𝑓(𝑥)) = 𝑒 4𝑎 ∫−∞ 𝑒 2√𝑎 𝑑𝑥 = 𝑒 4𝑎 ∫−∞ 𝑒 −𝑡 𝑑𝑡
√2𝜋 √2𝜋 √𝑎

−𝑤2 −𝑤2 −𝑤2 −𝑤2 −𝑤2

1 ∞ 2 1 √𝜋 𝜋 1
=
√2𝜋𝑎
𝑒 4𝑎 ∫−∞ 𝑒 −𝑡 𝑑𝑡 =
√2𝜋𝑎
𝑒 4𝑎 . √𝜋 =
√2𝜋𝑎
𝑒 4𝑎 = √2𝜋𝑎 𝑒 4𝑎 =
√2𝑎
𝑒 4𝑎

∞ 2 ∞ 2
This is because ∫−∞ 𝑒 −𝑡 𝑑𝑡 = 2 ∫0 𝑒 −𝑡 𝑑𝑡 =  1  2  = √𝜋

Badri A, Moges B. and Teklebrhan B. 100 AKU

 APPLIED MATHEMATICS III

−𝑤2
2 1
Hence ℱ(𝑓(𝑥)) = ℱ(𝑒 −𝑎𝑥 ) = 𝑒 4𝑎
√2𝑎

4.2.1 Properties of Fourier Transform

The properties of Fourier transform help to simplify the calculations involving Fourier transform
and to obtain some results which are otherwise difficult to obtain.

Property 1: We state the following theorem

Theorem (Linearity Theorem): For any functions 𝑓(𝑥) and 𝑔(𝑥)Whose Fourier Transform exist
and for any constants 𝑎, 𝑏

ℱ[𝑎𝑓(𝑥) + 𝑏𝑔(𝑥)] = 𝑎ℱ(𝑓(𝑥)) + 𝑏ℱ(𝑔(𝑥))

Proof:
1 ∞
By definition ℱ[𝑎𝑓(𝑥) + 𝑏𝑔(𝑥)] = ∫ (𝑎𝑓(𝑥) + 𝑏𝑔(𝑥))𝑒 −𝑖𝑤𝑥 𝑑𝑥
√2𝜋 −∞

1 ∞ 1 ∞
= ∫ 𝑎𝑓(𝑥))𝑒 −𝑖𝑤𝑥 𝑑𝑥 + √2𝜋 ∫−∞ 𝑏𝑔(𝑥)𝑒 −𝑖𝑤𝑥 𝑑𝑥
√2𝜋 −∞

1 ∞ ∞
= (∫−∞ 𝑎𝑓(𝑥))𝑒 −𝑖𝑤𝑥 𝑑𝑥 + ∫−∞ 𝑏𝑔(𝑥)𝑒 −𝑖𝑤𝑥 𝑑𝑥)
√2𝜋

1 ∞ ∞
= (𝑎 ∫−∞ 𝑓(𝑥))𝑒 −𝑖𝑤𝑥 𝑑𝑥 + 𝑏 ∫−∞ 𝑔(𝑥)𝑒 −𝑖𝑤𝑥 𝑑𝑥)
√2𝜋

𝑎 ∞ 𝑏 ∞
= ∫ 𝑓(𝑥))𝑒 −𝑖𝑤𝑥 𝑑𝑥 + √2𝜋 ∫−∞ 𝑔(𝑥)𝑒 −𝑖𝑤𝑥 𝑑𝑥)
√2𝜋 −∞

1 ∞ 1 ∞
=𝑎 ∫ 𝑓(𝑥))𝑒 −𝑖𝑤𝑥 𝑑𝑥 + 𝑏 √2𝜋 ∫−∞ 𝑔(𝑥)𝑒 −𝑖𝑤𝑥 𝑑𝑥)
√2𝜋 −∞

= 𝑎ℱ(𝑓(𝑥)) + 𝑏ℱ(𝑔(𝑥))

Hence, Fourier Transform is Linear

Property 2: Fourier Transforms of Derivatives; it is stated as follows

Theorem (Transform of derivatives): if 𝑓(𝑥) is continuous function of 𝑥 with 𝑓(𝑥) ⟶ 0 as

|𝑥| ⟶ ∞ and 𝑓′(𝑥) is absolutely integrable over (−∞, ∞),then

a) ℱ(𝑓′(𝑥)) = 𝑖𝑤ℱ[𝑓(𝑥)]

b) ℱ (𝑓 (𝑛) (𝑥)) = (𝑖𝑤)𝑛 ℱ[𝑓(𝑥)]

and this holds for all 𝑛 such that the derivatives 𝑓 (𝑟) (𝑥), 𝑟 = 1,2, … , 𝑛 satisfies the sufficient
conditions for the existence of the Fourier transforms

Badri A, Moges B. and Teklebrhan B. 101 AKU

 APPLIED MATHEMATICS III

Proof:
1 ∞
(a) By, definition ℱ(𝑓′(𝑥)) = ∫ 𝑓′(𝑥))𝑒 −𝑖𝑤𝑥 𝑑𝑥,
√2𝜋 −∞
integrating by parts we obtain

1 ∞ ∞
ℱ(𝑓′(𝑥)) = [(𝑓(𝑥)𝑒 −𝑖𝑤𝑥 ) −∞ − (−𝑖𝑤) ∫−∞ 𝑓(𝑥))𝑒 −𝑖𝑤𝑥 𝑑𝑥]
√2𝜋

1 ∞ ∞
= [(𝑓(𝑥)𝑒 −𝑖𝑤𝑥 ) −∞ + 𝑖𝑤 ∫−∞ 𝑓(𝑥))𝑒 −𝑖𝑤𝑥 𝑑𝑥
√2𝜋

Since 𝑓(𝑥) ⟶ 0 as |𝑥| ⟶ ∞, therefore

ℱ(𝑓′(𝑥)) = 𝑖𝑤ℱ[𝑓(𝑥)]

(b) The repeated application of result (a) gives result (b) provided that the desired conditions are
satisfied at each step.
2
Example 4.8: find the Fourier transform of 𝑓(𝑥) = 𝑥𝑒 −𝑎𝑥 , 𝑎 > 0

Solution; by property 2 part (a), we have

2 2 1 2 1 2
ℱ(𝑓(𝑥)) = ℱ[𝑥𝑒 −𝑎𝑥 ] = ℱ[𝑥𝑒 −𝑎𝑥 ] = ℱ [− 2𝑎 (𝑒 −𝑎𝑥 )′ ] = − 2𝑎 ℱ[(𝑒 −𝑎𝑥 )′ ]

1 2
= − 2𝑎 (𝑖𝑤)ℱ[𝑒 −𝑎𝑥 ], using differentiability

−𝑤2
−𝑖𝑤 1 2
= ( 𝑒 4𝑎 ) Refer example (find the Fourier transform of 𝑓(𝑥) = 𝑒 −𝑎𝑥 , 𝑎 > 0)
2𝑎 √2𝑎

−𝑤2
−𝑖𝑤
= 2𝑎√2𝑎 𝑒 4𝑎

−𝑤2
−𝑖𝑤
Hence ℱ(𝑓(𝑥)) = 2𝑎√2𝑎 𝑒 4𝑎

𝑑𝑛
Example 4.9: show that (a) ℱ[𝑥 𝑛 𝑓(𝑥)] = 𝑖 𝑛 𝑑𝑤𝑛 [𝐹(𝑤)]

𝑑𝑚
(b) ) ℱ[𝑥 𝑚 𝑓 (𝑛) (𝑥)(𝑥)] = 𝑖 𝑚+𝑛 𝑑𝑤𝑚 [𝑤 𝑛 𝐹(𝑤)]

Solution: (a) By the definition of Fourier transform, we have

1 ∞
𝐹(𝑤) = ∫ 𝑓(𝑥) 𝑒 −𝑖𝑤𝑥 𝑑𝑥
√2𝜋 −∞

Differentiating it w.r.t 𝑤 and using Leibnitz rule to differentiate under the integral sign, we have
𝑑 1 𝑑 ∞ −𝑖 ∞
𝑑𝑤
[𝐹(𝑤)] = ∫ 𝑓(𝑥) 𝑒 −𝑖𝑤𝑥 𝑑𝑥
√2𝜋 𝑑𝑤 −∞
= ∫ 𝑥𝑓(𝑥) 𝑒 −𝑖𝑤𝑥 𝑑𝑥
√2𝜋 −∞

Badri A, Moges B. and Teklebrhan B. 102 AKU

 APPLIED MATHEMATICS III

𝑑 1 ∞
or −𝑖 𝑑𝑤 [𝐹(𝑤)] = ∫ 𝑥𝑓(𝑥) 𝑒 −𝑖𝑤𝑥 𝑑𝑥 = ℱ(𝑥𝑓(𝑥))
√2𝜋 −∞

𝑑 𝑑
⟹ ℱ(𝑥1 𝑓(𝑥)) = (−𝑖)1 𝑑𝑤 [𝐹(𝑤)] = −𝑖 𝑑𝑤 [𝐹(𝑤)]

The repeated applications of the differentiation w.r.t 𝑤 leads to the desired result
𝑑𝑛
ℱ(𝑥 𝑛 𝑓(𝑥)) = 𝑖 𝑛 𝑑𝑤𝑛 [𝐹(𝑤)]

𝑑𝑚
(b) Consider ℱ[𝑥 𝑚 𝑓 (𝑛) (𝑥)(𝑥)] = 𝑖 𝑚 𝑑𝑤𝑚 [𝑓 (𝑛) (𝑥)] ( this is by using part (a))

𝑑𝑚 𝑑𝑚
= 𝑖 𝑚 𝑑𝑤𝑚 [(𝑖𝑤)𝑛 𝐹(𝑤)] = 𝑖 𝑚 . 𝑖 𝑛 𝑑𝑤𝑚[𝑤 𝑛 𝐹(𝑤)]

This because by the above Theorem (transforms of derivatives)

𝑑𝑚 𝑑𝑚
⟹ ℱ[𝑥 𝑚 𝑓 (𝑛) (𝑥)(𝑥)] = 𝑖 𝑚 . 𝑖 𝑛 𝑑𝑤𝑚 [𝑤 𝑛 𝐹(𝑤)] = 𝑖 𝑚+𝑛 𝑑𝑤𝑚 [𝑤 𝑛 𝐹(𝑤)]

𝑑𝑚
⟹ ℱ[𝑥 𝑚 𝑓 (𝑛) (𝑥)(𝑥)] = 𝑖 𝑚+𝑛 𝑑𝑤𝑚 [𝑤 𝑛 𝐹(𝑤)]

Provided 𝑓(𝑥) and its successive derivatives the requisite condition

Example 4.10: using the property of the Fourier transform of derivatives, find the Fourier
2
transform of 𝑓(𝑥) = 𝑒 −𝑎𝑥 , 𝑎 > 0.

Solution: clearly 𝑓(𝑥) satisfies the requisite conditions of continuity and absolute inerrability
over the real axis for the existence of Fourier transform.

It is easy to see that 𝑓(𝑥) satisfies the differential equation 𝑓′(𝑥) + 2𝑎𝑓(𝑥) = 0

Taking the Fourier transform of this, we have

ℱ[𝑓 ′ (𝑥)] + ℱ[2𝑎𝑥𝑓(𝑥)] = ℱ[𝑓 ′ (𝑥)] + 2𝑎ℱ[𝑥𝑓(𝑥)] = 0

This gives,

𝑖𝑤𝐹(𝑤) + 2𝑎(𝑖𝐹′(𝑤))= 𝑤𝐹(𝑤) + 2𝑎𝐹′(𝑤) = 0 (by the above example part (a))

⟹ ℱ[𝑓 ′ (𝑥)] + ℱ[2𝑎𝑥𝑓(𝑥)] = 2𝑎𝐹 ′ (𝑤) + 𝑤𝐹(𝑤) = 0

Where 𝐹(𝑤) is the Fourier transform of 𝑓(𝑥)

Rewriting the above equation as,

𝐹′ (𝑤) 1
= − 2𝑎 𝑤
𝐹(𝑤)

Badri A, Moges B. and Teklebrhan B. 103 AKU

 APPLIED MATHEMATICS III

Integrating the above equation w.r.t 𝑤, we obtain

−𝑤2
[ ]
𝐹(𝑤) = 𝐴𝑒𝑥𝑝 4𝑎 , where 𝐴 is an arbitrary constant, to determine 𝐴 we have 𝐹(0) = 𝐴and also
at 𝑤 = 0

1 ∞ 2 1 √𝜋 1 1
𝐹(0) = ∫ 𝑒 −𝑎𝑥
√2𝜋 −∞
𝑑𝑥 =
√2𝜋 √𝑎
. =
√2𝑎
, from this 𝐴 =
√2𝑎

−𝑤2 −𝑤2
1 1
Thus, 𝐹(𝑤) = 𝑒𝑥𝑝[ 4𝑎 ] = 𝑒 4𝑎 ,𝑎>0
√2𝑎 √2𝑎

Property3: let us state this result as the following theorem

Theorem (Shifting and Scaling): If 𝑓(𝑥) has Fourier transform 𝐹(𝑤),then

(a) ℱ[𝑓(𝑥 − 𝑥0 )] = 𝑒 −𝑖𝑤𝑥0 𝐹(𝑤); Shifting on the 𝑥-axis by 𝑥0

1
(b) ℱ[𝑓(𝑎𝑥)] = 𝑎 𝐹(𝑤 ⁄𝑎),𝑎 > 0; Scaling 𝑥 by 𝑎

(c) ℱ[𝑒 𝑖𝑤0 𝑥 𝑓(𝑥)] = 𝐹(𝑤 − 𝑤0 ); Shifting 𝑤 by 𝑤0

Proof: The results follow immediately from the definition of Fourier transform (you try!)
2
Example 4.11: find the Fourier transform of 𝑓(𝑥) = 𝑒 −𝑎(𝑥−5) , 𝑎 > 0

Solution: By the shifting property in the above theorem part (a) with𝑥0 = 5 , we have
−𝑤2
2 2 1
ℱ[𝑒 −𝑎(𝑥−5) ] = 𝑒 −𝑖5𝑤 ℱ[𝑒 −𝑎𝑥 ] = 𝑒 −𝑖5𝑤 𝑒 4𝑎 (By the above example (refer))
√2𝑎

𝑤2
1
= 𝑒 −( 4𝑎 +𝑖5𝑤)
√2𝑎

Example 4.12: find the Fourier transform of 𝑓(𝑥) = 4𝑒 −|𝑥| − 5𝑒 −3|𝑥+2|

Solution: using the linearity property, we have

ℱ[𝑓(𝑥)] = ℱ[4𝑒 −|𝑥| − 5𝑒 −3|𝑥+2| ] = ℱ[4𝑒 −|𝑥| ] − ℱ[5𝑒 −3|𝑥+2| ] = 4ℱ[𝑒 −|𝑥| ] − 5ℱ[𝑒 −3|𝑥+2| ]

⟹ ℱ[𝑓(𝑥)] = 4ℱ[𝑒 −|𝑥| ] − 5𝑒 2𝑖𝑤 ℱ[𝑒 |−3𝑥| ] (By using 𝑥-shifting)

5
= 4ℱ[𝑒 −|𝑥| ] − 3 𝑒 2𝑖𝑤 ℱ[𝑒 −|𝑥| ]𝑤→𝑤 (Using scaling)
⁄3

1 2 5 1 2 1 8 30𝑒 2𝑖𝑤
=4 . 1+𝑤2 − 3 𝑒 2𝑖𝑤 . . 𝑤 = [1+𝑤2 − ] (Refer the previous example)
√2𝜋 √2𝜋 1+( )2 √2𝜋 9+𝑤 2
3

Badri A, Moges B. and Teklebrhan B. 104 AKU

 APPLIED MATHEMATICS III

Exercises 4.2

1. Find the Fourier transform of each of the following

𝑒 𝑥 , |𝑥| < 𝑎 𝑎 − |𝑥|, |𝑥| < 𝑎

(a) 𝑓(𝑥) = { (b) 𝑓(𝑥) = {
0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 0, |𝑥| > 𝑎

(c) 𝑓(𝑥) = 𝑢(𝑥 + 1) − 𝑢(𝑥 − 1), where 𝑢(𝑥) is the unit –step function

𝑠𝑖𝑎𝑥 1, |𝑥| ≤ 𝑎
(d) 𝑓(𝑥) = ,𝑎 >0 (e) 𝑓(𝑥) = {
𝑥 0, |𝑥| > 𝑎

2. In each of the following find the Fourier transform of 𝑓(𝑥)

𝑒 𝑘𝑥 , 𝑥 < 0 (𝑘 > 0) 𝑥, 0 < 𝑥 < 𝑎

(a) 𝑓(𝑥) = { (b) 𝑓(𝑥) = {
0, 𝑥 > 0 0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

−1, −1 < 𝑥 < 0

|𝑥| , − 1 < 𝑥 < 1 𝑥𝑒 −𝑥 , −1 < 𝑥 < 0
(c) 𝑓(𝑥) = { (d) 𝑓(𝑥) = { 1, 0 < 𝑥 < 1 (e) 𝑓(𝑥) = {
0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

(𝑓) 𝑓(𝑥) = 𝑒 −𝑥 (−∞ < 𝑥 < ∞)

4.3 Laplace transform

Overview:
In this section, we are going to deal with the Laplace Transform and inverse Laplace transform
by considering various examples

Badri A, Moges B. and Teklebrhan B. 105 AKU

 APPLIED MATHEMATICS III

Section Objectives:

At the end of this subtopic, students will be able to:

 Define and identify the Laplace transform of a function;

 Understand and Know about the inverse Laplace transform of 𝑓 ;

The Laplace transforms which transforms a function f of one variable (t) into function F of
another variable (s) is named in honor of the French mathematician and Astronomer Pierre-
Simon Marquis de Laplace (1749–1827).

Integral Transform; If f (x, y) is a function of two variables, then a definite integral of f with
respect to one of the variables leads to a function of the other variable. For example, by holding y
2 𝑏
constant, we see that∫1 2𝑥𝑦 2 𝑑𝑥 = 3𝑦 2 . Similarly, a definite integral such as∫𝑎 𝐾(𝑠, 𝑡)𝑓(𝑡)𝑑𝑡
transforms a function f of the variable t into function F of the variable s.

We are particularly interested in an integral transform, where the interval of integration is the
unbounded interval [0, ∞). If f (t) is defined fort ≥0, then the improper integral is defined as a
limit.
∞ 𝑏
∫0 𝐾(𝑠, 𝑡)𝑓(𝑡)𝑑𝑡 = lim ∫0 𝐾(𝑠, 𝑡)𝑓(𝑡)𝑑𝑡 (1)
𝑏→∞

If the limit in (1) exists, then we say that the integral exists or is convergent; if the limit does not
exist, the integral does not exist and is divergent. The limit in (1) will, in general, exist for only
certain values of the variable s.

Definition: The function K(s, t) in (1) is called the kernel of the transform. The choice
𝐾(𝑠, 𝑡) = 𝑒 −𝑠𝑡 as the kernel gives us an especially important integral Transform.

Let us now introduce the main definition of Laplace transform

Definition (Laplace Transform): let f be a function defined for t ≥ 0. Then the integral
∞
ℒ{𝑓(𝑡)} = 𝐹(𝑠) = ∫0 𝑒 −𝑠𝑡 𝑓(𝑡)𝑑𝑡 (2)
is said to be the Laplace transform of f, provided that the integral converges

Badri A, Moges B. and Teklebrhan B. 106 AKU

 APPLIED MATHEMATICS III

Example 4.13: Evaluate ℒ{1}

∞ 𝑏 −𝑒 −𝑠𝑡 𝑏 −𝑒 −𝑠𝑏 +1
Solution: by definition ℒ{1} = ∫0 𝑒 −𝑠𝑡 (1)𝑑𝑡 lim ∫0 𝑒 −𝑠𝑡 𝑑𝑡 = lim [ ] = lim ( )
𝑏→∞ 𝑏→∞ 𝑠 0 𝑏→∞ 𝑠
−𝑒 −𝑠𝑏 1 1 1
= lim + lim = 0 + 𝑠 = 𝑠 , whenever 𝑠 > 0
𝑏→∞ 𝑠 𝑏→∞ 𝑠

Example 4.14: Evaluate ℒ{𝑡}

∞
Solution: from the Definition of Laplace transform we have ℒ{𝑓(𝑡)} = ∫0 𝑒 −𝑠𝑡 𝑓(𝑡)𝑑𝑡 =
∞ 𝑏
∫0 𝑒 −𝑠𝑡 𝑡𝑑𝑡 = lim ∫0 𝑒 −𝑠𝑡 𝑡𝑑𝑡
𝑏→∞
𝑏
Now let as use integration by parts on ∫0 𝑒 −𝑠𝑡 𝑡𝑑𝑡
𝑒 −𝑠𝑡
Let 𝑢 = 𝑡 ⇒ 𝑑𝑢 = 𝑑𝑡 and let 𝑑𝑣 = 𝑒 −𝑠𝑡 𝑑𝑡 ⇒ 𝑣 = − 𝑠
∞ 𝑒 −𝑠𝑡 ∞ 1 ∞ 𝑒 −𝑠𝑡 ∞ 1
⇒ ∫0 𝑒 −𝑠𝑡 𝑡𝑑𝑡 = [−𝑡 𝑠 ] + 𝑠 ∫0 𝑒 −𝑠𝑡 𝑑𝑡 = = [−𝑡 𝑠 ] + 𝑠 ℒ{1} (by the above example)
0 0
𝑒 −𝑠𝑡 𝑒 −𝑠𝑡 𝑒 −𝑠𝑡 ∞
As t→ ∞,= −𝑡 → 0 and as 𝑡 → 0, 𝑡 → 0 ⇒ [−𝑡 ] = 0
𝑠 𝑠 𝑠 0

∞ 1 1 1 1
Hence, ℒ{𝑡} = ∫0 𝑒 −𝑠𝑡 𝑡𝑑𝑡 = ℒ{1} = 𝑠 . 𝑠 = 𝑠2 when ever 𝑠 > 0
𝑠

Example 4.15: Evaluate (𝒂) ℒ{𝑒 −3𝑡 } (𝑏) ℒ{𝑒 5𝑡 }

Solution: In each case we use the definition of Laplace transform

∞ ∞ −𝑒 −(𝑠+3)𝑡 ∞ 1
For a) ℒ{𝑒 −3𝑡 } = ∫0 𝑒 −𝑠𝑡 𝑒 −3𝑡 𝑑𝑡 = ∫0 𝑒 −(𝑠+3)𝑡 𝑑𝑡 = [ ]0 = 𝑠+3
𝑠+3

−𝑒 −(𝑠+3)𝑡 −𝑒 −(𝑠+3)𝑡 1
From this observe that as t→ ∞, → 0 𝑎𝑛𝑑 as 𝑡 → 0, → 𝑠+3
𝑠+3 𝑠+3

1
Hence, ℒ{𝑒 −3𝑡 } = 𝑠+3, whenever 𝑠 > −3

Example 4.16: evaluate ℒ{𝑠𝑖𝑛2𝑡}

Solution: By the definition of Laplace transform we obtain,

∞
ℒ{𝑠𝑖𝑛2𝑡} = ∫0 𝑒 −𝑠𝑡 𝑠𝑖𝑛2𝑡𝑑𝑡 , now using integration by parts, we have

Badri A, Moges B. and Teklebrhan B. 107 AKU

 APPLIED MATHEMATICS III

𝑒 −𝑠𝑡
Let 𝑢 = 𝑠𝑖𝑛2𝑡 𝑑𝑢 = 2𝑐𝑜𝑠2𝑡 and let 𝑑𝑣 = 𝑑𝑣 = 𝑒 −𝑠𝑡 , 𝑣 = − 𝑠

∞ 𝑒 −𝑠𝑡 𝑠𝑖𝑛2𝑡 ∞ 2 ∞ −𝑠𝑡 2 ∞

⟹ ℒ{𝑠𝑖𝑛2𝑡} = ∫0 𝑒 −𝑠𝑡 𝑠𝑖𝑛2𝑡𝑑𝑡 = [− ] + 𝑠 ∫0 𝑒 𝑐𝑜𝑠2𝑡𝑑𝑡 = 𝑠 ∫0 𝑒 −𝑠𝑡 𝑐𝑜𝑠2𝑡𝑑𝑡
𝑠 0
𝑒 −𝑠𝑡 𝑠𝑖𝑛2𝑡 𝑒 −𝑠𝑡 𝑠𝑖𝑛2𝑡
Because as 𝑠 ⟶ ∞, − ⟶ 0 and as 𝑠 ⟶ 0, − ⟶0
𝑠 𝑠

∞ 2 ∞
⟹ ℒ{𝑠𝑖𝑛2𝑡} = ∫0 𝑒 −𝑠𝑡 𝑠𝑖𝑛2𝑡𝑑𝑡 = 𝑠 ∫0 𝑒 −𝑠𝑡 𝑐𝑜𝑠2𝑡𝑑𝑡 (1)

∞
Now integrating ∫0 𝑒 −𝑠𝑡 𝑐𝑜𝑠2𝑡𝑑𝑡 by parts we have,

𝑒 −𝑠𝑡
Let 𝑢 = 𝑐𝑜𝑠2𝑡, 𝑑𝑢 = −2𝑠𝑖𝑛2𝑡 and let = 𝑒 −𝑠𝑡 , 𝑣 = − 𝑠

∞ −𝑒 −𝑠𝑡 𝑐𝑜𝑠2𝑡 ∞ 2 ∞ −𝑠𝑡 1 2 ∞

∫0 𝑒 −𝑠𝑡 𝑐𝑜𝑠2𝑡𝑑𝑡 = [[ 𝑠
] −
0 𝑠 ∫0
𝑒 𝑠𝑖𝑛2𝑡𝑑𝑡 ] = 𝑠 − 𝑠 ∫0 𝑒 −𝑠𝑡 𝑠𝑖𝑛2𝑡𝑑𝑡

−𝑒 −𝑠𝑡 𝑐𝑜𝑠2𝑡 −𝑒 −𝑠𝑡 𝑐𝑜𝑠2𝑡 1

Because 𝑠 ⟶ ∞, ⟶ 0 and as 𝑠 ⟶ 0, ⟶
𝑠 𝑠 𝑠

∞ 1 2 ∞
∫0 𝑒 −𝑠𝑡 𝑐𝑜𝑠2𝑡𝑑𝑡 = 𝑠 − 𝑠 ∫0 𝑒 −𝑠𝑡 𝑠𝑖𝑛2𝑡𝑑𝑡 , 𝑠 > 0 (2)

Now substituting (2) into (1),we have

∞ 2 1 2 ∞ 2 4 ∞
ℒ{𝑠𝑖𝑛2𝑡} = ∫0 𝑒 −𝑠𝑡 𝑠𝑖𝑛2𝑡𝑑𝑡 = 𝑠 (𝑠 − 𝑠 ∫0 𝑒 −𝑠𝑡 𝑠𝑖𝑛2𝑡𝑑𝑡) = − 𝑠2 ∫0 𝑒 −𝑠𝑡 𝑠𝑖𝑛2𝑡𝑑𝑡
𝑠2

∞ 4 ∞ 2
⟹ ℒ{𝑠𝑖𝑛2𝑡} = ∫0 𝑒 −𝑠𝑡 𝑠𝑖𝑛2𝑡𝑑𝑡 + 𝑠2 ∫0 𝑒 −𝑠𝑡 𝑠𝑖𝑛2𝑡𝑑𝑡 = 𝑠2

4 ∞ 2
= (1 + 𝑠2 ) ∫0 𝑒 −𝑠𝑡 𝑠𝑖𝑛2𝑡𝑑𝑡 = 𝑠2

𝑠2 +4 ∞ 2
=( ) ∫0 𝑒 −𝑠𝑡 𝑠𝑖𝑛2𝑡𝑑𝑡 = 𝑠2
𝑠2

∞ 𝑠2 2 2
⟹ ℒ{𝑠𝑖𝑛2𝑡} = ∫0 𝑒 −𝑠𝑡 𝑠𝑖𝑛2𝑡𝑑𝑡 = 𝑠2 +4 (𝑠2 ) = 𝑠2 +4

2
Hence ℒ{𝑠𝑖𝑛2𝑡} = 𝑠2 +4 , 𝑠 > 0

Property (𝓛 is a linear transform): for a linear combination of functions we can write

∞ ∞ ∞
∫0 𝑒 −𝑠𝑡 [𝛼𝑓(𝑡) + 𝛽𝑔(𝑡)]𝑑𝑡 = 𝛼 ∫0 𝑒 −𝑠𝑡 𝑓(𝑡)𝑑𝑡 + 𝛽 ∫0 𝑒 −𝑠𝑡 𝑔(𝑡)𝑑𝑡

When ever both integrals converge for 𝑠 > 𝑐. Hence it follows that

Badri A, Moges B. and Teklebrhan B. 108 AKU

 APPLIED MATHEMATICS III

ℒ{ 𝛼𝑓(𝑡) + 𝛽𝑔(𝑡)} = 𝛼ℒ{𝑓(𝑡)} + 𝛽ℒ{𝑔(𝑡)} = 𝛼𝐹(𝑠) + 𝛽𝐺(𝑠)

Example 4.17: Evaluate (𝑎) ℒ{1 + 5𝑡} (𝑏) ℒ{4𝑒 5𝑡 − 10𝑠𝑖𝑛2𝑡}

Solution: By using linearity property and the above examples we have

a) For 𝑠 > 0
1 1
ℒ{1 + 5𝑡} = ℒ{1} + ℒ{5𝑡} = ℒ{1} + 5ℒ{𝑡} = 𝑠 + 𝑠2 (By 𝓛 linearity and above examples)

b) 𝑓𝑜𝑟 𝑠 > 5

ℒ{4𝑒 5𝑡 − 10𝑠𝑖𝑛2𝑡} = ℒ{4𝑒 5𝑡 } − ℒ{10𝑠𝑖𝑛2𝑡} = 4ℒ{𝑒 5𝑡 } − 10ℒ{𝑠𝑖𝑛2𝑡}

4 20
ℒ{4𝑒 5𝑡 − 10𝑠𝑖𝑛2𝑡} = 4ℒ{𝑒 5𝑡 } − 10ℒ{𝑠𝑖𝑛2𝑡} = −
𝑠−5 𝑠2 +4

This is because 𝓛 linearity and above examples

We state the generalization of some of the preceding examples by means of the next theorem.
From this point on we shall also refrain from stating any restrictions on s; it is understood that s
is sufficiently restricted to guarantee the convergence of the appropriate Laplace transform.

Theorem: Transforms of Some Basic Functions

1 𝑘
(a) ℒ{1} = 𝑠 (d) ℒ{𝑠𝑖𝑛𝑘𝑡} = 𝑠2 +𝑘 2
𝑛! 𝑠
(b) ℒ{𝑡 𝑛 } = 𝑠𝑛+1 ,𝑛 = 1,2,3, .. (e) ℒ{𝑐𝑜𝑠𝑘𝑡} = 𝑠2 +𝑘 2
1 𝑘
(c) ℒ{𝑒 𝑎𝑡 } = 𝑠−𝑎 (f) ℒ{𝑠𝑖𝑛ℎ 𝑘𝑡} = 𝑠2 −𝑘 2
1 𝑠
(d) ℒ{𝑒 −𝑎𝑡 } = 𝑠+𝑎 (g) ℒ{𝑐𝑜𝑠ℎ 𝑘𝑡} = 𝑠2 −𝑘 2
These can be proved by the direct application of the definition of Laplace transform

Sufficient Conditions for Existence of ℒ{𝑓(𝑡)}: The integral that define the Laplace transform
∞
does not have to converge (i.e. if ℒ{𝑓(𝑡)} = 𝐹(𝑠) = ∫0 𝑒 −𝑠𝑡 𝑓(𝑡)𝑑𝑡 doesn’t converge), then the
2
Laplace transform of 𝑓 doesn’t exist). For example, neither ℒ{1⁄𝑡 } nor ℒ{𝑒 𝑡 }exists. Sufficient
conditions guaranteeing the existence of ℒ{𝑓(𝑡)} are that f be piecewise continuous on [0,∞)
and that f be of exponential order for 𝑡 ≥ 𝑇. Recall that a function f is piecewise continuous on
[0,∞) if, in any interval 0 ≤ a ≤t ≤b, there are at most a finite number of 𝑝𝑜𝑖𝑛𝑡𝑠 𝑡𝑘 , 𝑘 =
1, 2, . . . , 𝑛 (𝑡𝑘−1 < 𝑡𝑘 ) at which f has finite discontinuities and is continuous on each open
interval (𝑡𝑘−1 , 𝑡𝑘 ). See Figure 4.1 below. The concept of exponential order is defined in the
following manner.

Badri A, Moges B. and Teklebrhan B. 109 AKU

 APPLIED MATHEMATICS III

Figure 4.1 piece wise continuous function

Definition (Exponential order): a function 𝑓 is said to be of exponential order 𝑐 if there exists

constants 𝑐, 𝑀 > 0 𝑎𝑛𝑑 𝑇 > 0 such that |𝑓(𝑡)| ≤ 𝑀𝑒 𝑐𝑡 for all 𝑡 > 𝑇

Example 4.18: 𝑓(𝑡) = 𝑡 , 𝑓(𝑡) = 𝑒 −𝑡 𝑎𝑛𝑑 𝑓(𝑡) = 2𝑐𝑜𝑠𝑡 Is exponential order of 𝑐 = 1 for
all 𝑡 > 𝑇.

Since we have respectively|𝑓(𝑡)| = |𝑡| ≤ 𝑒 𝑡 , |𝑓(𝑡)| = |𝑒 −𝑡 | and |𝑓(𝑡)| = |2𝑐𝑜𝑠𝑡 | ≤ 2𝑒 𝑡 .

Theorem (sufficient conditions for the existence of Laplace transforms): if f is piecewise

continuous on [0, ∞) and exponential order 𝑐,then ℒ{𝑓(𝑡)} exists for𝑠 > 𝑐.

0, 0 ≤ 𝑡 < 3
Example 4.19: Evaluate ℒ{𝑓(𝑡)} where 𝑓(𝑡) = {
2, 𝑡≥3

Solution: The function 𝑓 is piecewise continuous and of exponential order for 𝑡 > 0
∞ 3 ∞ 2𝑒 −𝑠𝑡 ∞ 2𝑒 −3𝑠
Now, ℒ{𝑓(𝑡)} = ∫0 𝑒 −𝑠𝑡 𝑓(𝑡)𝑑𝑡 = ∫0 𝑒 −𝑠𝑡 (0)𝑑𝑡 + ∫3 𝑒 −𝑠𝑡 (2)𝑑𝑡 = 0 + [ ] = ,𝑠 > 0
−𝑠 3 𝑠

2𝑒 −3𝑠
Hence, ℒ{𝑓(𝑡)} = ,𝑠 > 0
𝑠

Inverse Laplace Transform

Badri A, Moges B. and Teklebrhan B. 110 AKU

 APPLIED MATHEMATICS III

If 𝐹(𝑠) represents the Laplace transform of a function 𝑓(𝑡),that is ℒ{𝑓(𝑡)} = 𝐹(𝑠) we then say
𝑓(𝑡) is the inverse Laplace transform of 𝐹(𝑠) and write

𝑓(𝑡) = ℒ −1 {𝐹(𝑠)}.

Example 4.20: Evaluate the inverse Laplace transform of each of the following
1 1 1
a.𝐹(𝑠) = b). 𝐹(𝑠) = 𝑠2 c. 𝐹(𝑠) = 𝑠+3
𝑠

1
Solution: 𝑎) 𝑓(𝑡) = ℒ −1 {𝐹(𝑠)} = ℒ −1 {𝑠 } = 1

1
b) 𝑓(𝑡) = ℒ −1 {𝐹(𝑠)} = ℒ −1 {𝑠2 } = 𝑡

1
c) 𝑓(𝑡) = ℒ −1 {𝐹(𝑠)} = ℒ −1 {𝑠+3 } = 𝑒 −3𝑡

Theorem (some inverse transforms)

1 𝑛!
(a) 1 = ℒ −1 {𝑠 } (b) 𝑡 𝑛 = ℒ −1 {𝑠𝑛+1 } , 𝑛 = 1,2,3, …

1 𝑘
(c) 𝑒 𝑎𝑡 = ℒ −1 {𝑠−𝑎 } (d) 𝑠𝑖𝑛𝑘𝑡 = ℒ −1 {𝑠2 +𝑘 2 }

𝑠 𝑘
(e) 𝑐𝑜𝑠𝑘𝑡 = ℒ −1 {𝑠2 +𝑘 2 } (f) 𝑠𝑖𝑛ℎ𝑘𝑡 = ℒ −1 {𝑠2 −𝑘 2 }

𝑠
(g) 𝑐𝑜𝑠ℎ𝑘𝑡 = ℒ −1 {𝑠2 −𝑘 2 }

1 1
Example 4.21: Evaluate a) ℒ −1 {𝑠5 } b) ℒ −1 {𝑠2 +7}

Solution: (a) By the above theorem and identifying that 𝑛 + 1 = 5 𝑜𝑟 𝑛 = 4 and then
multiplying and dividing 4!, we have,
1 1 4! 1 1 1 1 4! 1 1
ℒ −1 {𝑠5 } = ℒ −1 {𝑠4+1 } = ℒ −1 {4! 𝑠4+1 } = ℒ −1 {4! 𝑠4+1 } = 4! ℒ −1 {𝑠4+1 } = 4! 𝑡 4 = 24 𝑡 4
4!

1 1 √7 1 √7 1
b) ℒ −1 {𝑠2 +7} = ℒ −1 { 2 2}
= ℒ −1 {𝑠2 +(√7)2 } = 𝑠𝑖𝑛√7𝑡
√7 𝑠 +(√7) √7 √7

1
Here, we have fixed up the expression𝑠2 +7 by multiplying and dividing by √7

𝓛−𝟏 is a linear transform: The inverse Laplace transform is also a linear transform, that is for
constants 𝛼 and 𝛽 and for some functions 𝐹 and 𝐺 that are transforms of 𝑓 and 𝑔 respectively,
then

ℒ −1 {𝛼𝐹(𝑠) + 𝛽𝐺(𝑆)} = 𝛼ℒ −1 {𝐹(𝑠)} + 𝛽ℒ −1 {𝐺(𝑠)}

Badri A, Moges B. and Teklebrhan B. 111 AKU

 APPLIED MATHEMATICS III

−2𝑠+6
Example 4.22: Evaluate ℒ −1 { 𝑠2 +4 }

Solution: we first rewrite the given function of s as two expressions by means of term wise
division and the use linearity ofℒ −1 .
−2𝑠+6 −2𝑠 6 −2𝑠 −6 𝑠 1
ℒ −1 { 𝑠2 +4 } = ℒ −1 {𝑠2 +4 + 𝑠2 +4} = ℒ −1 {𝑠2 +4} + ℒ −1 {𝑠2 +4} = −2ℒ −1 {𝑠2 +4} +6ℒ −1 {𝑠2 +4}

𝑠 6 2
= −2ℒ −1 {𝑠2 +4} + 2 ℒ −1 {𝑠2 +4} (by linearity and fixing the second expression)

= −2𝑐𝑜𝑠2𝑡 + 3𝑠𝑖𝑛2𝑡
𝑠+3
Example 4.23: Evaluate ℒ −1 {𝑠2 −7𝑠+12}

𝑠+3 𝑠+3 𝑠+3 𝑠+3

Solution: ℒ −1 {𝑠2 −7𝑠+12} = ℒ −1 {𝑠2 −4𝑠−3𝑠+12} = ℒ −1 {𝑠(𝑠−4)−3(𝑠−4)} = ℒ −1 {(𝑠−4)(𝑠−3)}

𝑠+𝑠 𝑠+𝑠
Now we decompose 𝑠2 −7𝑠+12 = (𝑠−4)(𝑠−3) into sum of partial fractions

𝑠+3 𝑠+3 𝐴 𝐵 𝐴(𝑆−3)+𝐵(𝑆−4)

That is = (𝑠−4)(𝑠−3) = + (𝑠−3) =
𝑠2 −7𝑠+12 (𝑠−4) (𝑠−4)(𝑠−3)

𝑠+3 𝐴(𝑆−3)+𝐵(𝑆−4)
From this (𝑠−4)(𝑠−3) = (𝑠−4)(𝑠−3)
⟹ 𝐴(𝑠 − 3) + 𝐵(𝑆 − 4) = 𝑠 + 3

⟹ 𝐴𝑠 − 3𝐴 + 𝐵𝑆 − 4𝐵 = 𝑠 + 3 ⟹ 𝐴𝑠 + 𝐵𝑆 − 3𝐴 − 4𝐵 = 𝑠 + 3

A+B =1
⟹ (A+B) + (-3A-4B) = 𝑠 + 3 ⟹ { ⟹ 𝐴 = 7 𝑎𝑛𝑑 𝐵 = −6
−3A − 4B = 3
𝑠+3 𝑠+3 𝐴 𝐵 7 −6
From this 𝑠2 −7𝑠+12 = (𝑠−4)(𝑠−3) = (𝑠−4)
+ (𝑠−3) = (𝑠−4)
+ (𝑠−3)

𝑠+3 7 −6 7 −6
Now ℒ −1 {𝑠2 −7𝑠+12} = ℒ −1 {(𝑠−4) + (𝑠−3)} = ℒ −1 {𝑠−4} + ℒ −1 {𝑠−3}

1 1
= 7ℒ −1 {𝑠−4} − 6ℒ −1 {𝑠−3} = 7𝑒 4𝑡 − 6𝑒 3𝑡

Transforms of derivative

Badri A, Moges B. and Teklebrhan B. 112 AKU

 APPLIED MATHEMATICS III

As was pointed out in the introduction to this chapter, Laplace transform is used to solve
differential equations. To that end we need to evaluate quantities such as ℒ{𝑑𝑦⁄𝑑𝑡} and
ℒ{𝑑 2 𝑦⁄𝑑 2 𝑡}.

For example, if 𝑓′ is continuous for 𝑡 ≥ 0, then integration by parts gives

∞ ∞
ℒ{𝑓 ′ (𝑡)} = ∫0 𝑒 −𝑠𝑡 𝑓′(𝑡)𝑑𝑡 = [𝑒 −𝑠𝑡 𝑓(𝑡)]∞
0
+ 𝑠 ∫0 𝑒 −𝑠𝑡 𝑓(𝑡)𝑑𝑡 = −𝑓(0) + 𝑠 ℒ{𝑓(𝑡)}

or ℒ{𝑓 ′ (𝑡)} = 𝑠𝐹(𝑠) − 𝑓(0) (1)

here we have assumed that 𝑒 −𝑠𝑡 𝑓(𝑡) → 0 𝑎𝑠 𝑡 → ∞,similarly with the aid of (1),
∞ ∞
ℒ{𝑓 ′′ (𝑡)} = ∫0 𝑒 −𝑠𝑡 𝑓 ′′ (𝑡)𝑑𝑡 = [𝑒 −𝑠𝑡 𝑓′(𝑡)]∞
0
+ 𝑠 ∫0 𝑒 −𝑠𝑡 𝑓′(𝑡)𝑑𝑡 = −𝑓 ′ (0) + 𝑠 ℒ{𝑓′(𝑡)}

⟹ ℒ{𝑓 ′′ (𝑡)} = 𝑠[𝑠𝐹(𝑠) − 𝑓(0)] − 𝑓 ′ (0) (By (1))

⟹ ℒ{𝑓 ′′ (𝑡)} = 𝑠 2 𝐹(𝑠) − 𝑠𝑓(0) − 𝑓 ′ (0) (2)

In like manner it can be shown that

ℒ{𝑓 ′′ ′(𝑡)} = 𝑠 3 𝐹(𝑠) − 𝑠 2 𝑓(0) − 𝑠𝑓 ′ (0) − 𝑓 ′′ (0) (3)

In general, The results (1), (2) and (3) can be generalized in the following theorem

Theorem (Transform of a Derivative)

If 𝑓, 𝑓′, … , 𝑓 (𝑛−1) are continuous on [0,∞) and are of exponential order and if 𝑓 (𝑛) (𝑡) is piece
wise continuous on [0,∞), then

ℒ{𝑓 (𝑛) (𝑡)} = 𝑠 𝑛 𝐹(𝑠) − 𝑠 𝑛−1 𝑓(0) − 𝑠𝑓 𝑛−2 (0) − … − 𝑓 𝑛−1 (0),

Where, F(s) = ℒ{𝑓(𝑡)}.

In solving ODEs it is apparent from the general result given in the above theorem (Transform of
a Derivative) that ℒ{𝑑𝑛 𝑦⁄𝑑𝑛 𝑡} depends on 𝑌(𝑠) = ℒ{𝑦(𝑡)} and 𝑛 − 1 derivatives of 𝑦(𝑡)
evaluated at 𝑡 = 0.this property make the Laplace ideally suited for solving linear initial-value
problems in which the differential equation has constant coefficients such a differential is simply
a linear combination of terms 𝑦, 𝑦′, 𝑦′′,…,𝑦 (𝑛) :

𝑑𝑛 𝑦 𝑑𝑛−1 𝑦
𝑎𝑛 𝑑𝑡 𝑛 + 𝑎𝑛−1 𝑑𝑡 𝑛−1 + … + 𝑎0 𝑦 = 𝑔(𝑡), 𝑦(0) = 𝑦0,𝑦 ′ (0) = 𝑦1 , … , 𝑦 (𝑛−1) (0) = 𝑦𝑛−1

Where 𝑎𝑖 , 𝑖 = 0,1, … , 𝑛 and 𝑦0 , 𝑦1 , … , 𝑦𝑛−1 are constants.

Badri A, Moges B. and Teklebrhan B. 113 AKU

 APPLIED MATHEMATICS III

By the linearity property the Laplace transform of this linear combination is a linear combination
of Laplace transforms:
𝑑𝑛 𝑦 𝑑𝑛−1 𝑦
𝑎𝑛 ℒ{ 𝑑𝑡 𝑛 } + 𝑎𝑛−1 { 𝑑𝑡 𝑛−1 } + … + 𝑎0 ℒ{𝑦} = ℒ{𝑔(𝑡)}

𝑑𝑛 𝑦 𝑑𝑛−1 𝑦
𝑎𝑛 𝑑𝑡 𝑛 + 𝑎𝑛−1 𝑑𝑡 𝑛−1 + … + 𝑎0 𝑦 = 𝑔(𝑡) (4)

Then by the above theorem (Transform of a Derivative)

𝑎𝑛 [𝑠 𝑛 𝑌(𝑠) − 𝑠 𝑛−1 𝑦(0) − 𝑠 𝑛−2 𝑦(0) − … − 𝑦 (𝑛−1) (0)] + 𝑎𝑛−1 [𝑠 𝑛−1 𝑌(𝑠) − 𝑠 𝑛−2 𝑦(0) − ⋯ −
𝑦 (𝑛−2) (0)] + 𝑎0 𝑌(𝑠) = 𝐺(𝑠). (5)

Where ℒ{𝑦(𝑡)}= 𝑌(𝑠) and ℒ{𝑔(𝑡)}= 𝐺(𝑠).

In other words,

The Laplace transform of a linear differential equation with constant coefficient becomes an
algebraic equation in 𝑌(𝑠).

If we solve the general transformed equation (5) for the symbol𝑌(𝑠), we first obtain

𝑃(𝑠)𝑌(𝑠) = 𝑄(𝑠) + 𝐺(𝑠) and then write

𝑄(𝑠) 𝐺(𝑠)
𝑌(𝑠) = + 𝑃(𝑠) (6)
𝑃(𝑠)

Where 𝑃(𝑠) = 𝑎𝑛 𝑠 𝑛 + 𝑎𝑛−1 𝑠 𝑛−1 + ⋯ + 𝑎0, 𝑄(𝑠) is a polynomial in s of degree less than or
equal to 𝑛 − 1 consisting of the various products of the coefficient 𝑎𝑖 , 𝑖 = 0,1, … , 𝑛 and the
prescribed initial conditions 𝑦0 , 𝑦1 , … , 𝑦𝑛−1 and 𝐺(𝑠) is the Laplace transform of 𝑔(𝑡).Typically,
we put the two terms in (6) over the least common denominator and then decompose the
expression into two or more partial fractions. Finally, the solution 𝑦(𝑡) of the original initial-
value problem is 𝑦(𝑡) = ℒ −1 {𝑌(𝑠)}, where the inverse transform is done term by term. Let us
summarize the procedure in the following diagram

Badri A, Moges B. and Teklebrhan B. 114 AKU

 APPLIED MATHEMATICS III

Find the unknown Apply Laplace Transformed DE

𝑦(𝑡) that satisfies the DE Transform ℒ Becomes an algebraic
and Initial conditions equation in Y(s)

Apply inverse Solve transformed

Solution 𝑦(𝑡)
Laplace transformℒ −1 equation for Y(s)
of original IVP

Figure 4.2 Steps in solving an IVP by the Laplace transform

Example 4.24: Use the Laplace Transform to solve the initial-value problem
𝑑𝑦
+ 3𝑦 + 13𝑠𝑖𝑛2𝑡, 𝑦(0) = 6
𝑑𝑡

Solution: We first take the transform of each member of the differential equation:
𝑑𝑦 𝑑𝑦
ℒ { 𝑑𝑡 + 3𝑦 + 13𝑠𝑖𝑛2𝑡} = ℒ { 𝑑𝑡 } + 3ℒ{ 𝑦} + 13ℒ{ 𝑠𝑖𝑛2𝑡} (7)

𝑑𝑦
From (1), ℒ { 𝑑𝑡 } = 𝑠𝑌(𝑠) − 𝑦(0) = 𝑠𝑌(𝑠) − 6 and we know that ℒ{ 𝑠𝑖𝑛2𝑡} = 2⁄(𝑠 2 + 4)

𝑑𝑦
So, ℒ { 𝑑𝑡 } + 3ℒ{ 𝑦} + 13ℒ{ 𝑠𝑖𝑛2𝑡} = 𝑠𝑌(𝑠) − 6 + 3𝑌(𝑠) = 26⁄(𝑠 2 + 4)

26
Or (𝑠 + 3)𝑌(𝑠) = 6 + ,we get
𝑠2 +4

Solving this last equation for 𝑌(𝑠)

6 26 6𝑠2 +50
𝑌(𝑠) = 𝑠+3 + (𝑠+3)(𝑠2 +4) = (𝑠+3)(𝑠2 +4) (7)

Since the quadratic polynomial 𝑠 2 + 4 does not factor using real numbers, its assumed numerator
in the partial fraction decomposition is a linear polynomial in𝑠:

6𝑠2 +50 𝐴 𝐵𝑠+𝐶

(𝑠+3)(𝑠2 +4)
= 𝑠+3 + .
𝑠2 +4

Putting the right-hand side of the equality over a common denominator and equating numerators

gives 6𝑠 2 + 50 = 𝐴(𝑠 2 + 4) + (𝐵𝑠 + 𝐶)(𝑠 + 3).setting 𝑠 = −3 then immediately yields 𝐴 = 8

Badri A, Moges B. and Teklebrhan B. 115 AKU

 APPLIED MATHEMATICS III

Since the denominator has no more real zeros, we equate the coefficients of 𝑠 2 and s:

6 = 𝐴 + 𝐵 and 0 = 3𝐵 + 𝐶. Using the value of 𝐴 in the first equation gives 𝐵 = −2, and then
using in the second equation gives 𝐶 = 6.Thus

6𝑠2 +50 8 −2𝑠+6 8 −2𝑠 6

𝑌(𝑠) = (𝑠+3)(𝑠2 +4) = 𝑠+3 + = 𝑠+3 + + 𝑠2 +4
𝑠2 +4 𝑠2 +4

8 −2𝑠 6
Or 𝑌(𝑠) = 𝑠+3 + + 𝑠2 +4 (8)
𝑠2 +4

Taking the inverse Laplace transform in (8), we have

1 𝑠 1
𝑦(𝑡) = 8ℒ −1 { } − 2ℒ −1 { 2 } + 6ℒ −1 { 2 }
𝑠+3 𝑠 +4 𝑠 +4

Or 𝑦(𝑡) = 8𝑒 −3𝑡 − 2𝑐𝑜𝑠 2𝑡 + 3𝑠𝑖𝑛 2𝑡.

Hence, the solution of the initial value problem is (𝑡) = 8𝑒 −3𝑡 − 2𝑐𝑜𝑠 2𝑡 + 3𝑠𝑖𝑛 2𝑡.

Example 4.25: Solve the Second-Order IVB

𝑦 ′′ − 3𝑦 ′ + 2𝑦 = 𝑒 −4𝑡 , 𝑦(0) = 1, 𝑦 ′ (0) = 5

Solution: proceeding as in the example above, we transform the DE. We take the sum of the
transforms of each term, use the given initial conditions and then solve for 𝑌(𝑠):
𝑑2 𝑦 𝑑𝑦
ℒ { 𝑑𝑡 2 } − 3ℒ { 𝑑𝑡 } + 2ℒ{ 𝑦} = ℒ{ 𝑒 −4𝑡 }

1
𝑠 2 𝑌(𝑠) − 𝑠𝑦(𝑠) − 𝑦 ′ (0) − 3[𝑠𝑌(𝑠) − 𝑦(0)] = 𝑠+4

1
(𝑠 2 − 3𝑠 + 2)𝑌(𝑠) = 𝑠 + 2 +
𝑠+4

𝑠+2 1 𝑠2 +6𝑠+9
𝑌(𝑠) = + (𝑠2 +3𝑠+2)(𝑠+4) = (𝑠−1)(𝑠−2)(𝑠+4)
𝑠2 −3𝑠+2

Then decomposition into partial fraction yields

𝑠2 +6𝑠+9 𝐴 𝐵 𝐶 𝐴(𝑆−2)(𝑆+4)+𝐵(𝑆−1)(𝑆+4)+𝐶(𝑆−1)(𝑆−2)
(𝑠−1)(𝑠−2)(𝑠+4)
= 𝑠−1 + 𝑠−2 + 𝑠+4 = (𝑠−1)(𝑠−2)(𝑠+4)

from this we have

𝐴(𝑆 − 2)(𝑆 + 4) + 𝐵(𝑆 − 1)(𝑆 + 4) + 𝐶(𝑆 − 1)(𝑆 − 2) = 𝑠 2 + 6𝑠 + 9

Letting 𝑠 = 2 to get the value of 𝐵, we have,

𝐴(2 − 2)(2 + 4) + 𝐵(2 − 1)(2 + 4) + 𝐶(2 − 1)(2 − 2) = 22 + 6(2) + 9

Badri A, Moges B. and Teklebrhan B. 116 AKU

 APPLIED MATHEMATICS III

25
6𝐵 = 25 or 𝐵 = 6

Letting 𝑠 = −4 to get the value of 𝐶, we have,

𝐴(−4 − 2)(−4 + 4) + 𝐵(−4 − 1)(−4 + 4) + 𝐶(−4 − 1)(−4 − 2) = (−4)2 + 6(−4) + 9

1
30𝐶 = 1 or 𝐶 = 30

Letting 𝑠 = 1 to get the value of 𝐴, we have,

𝐴(1 − 2)(1 + 4) + 𝐵(1 − 1)(1 + 4) + 𝐶(1 − 1)(1 − 2) = 12 + 6(1) + 9

16
−5𝐴 = 16 or 𝐴 = − 5

𝑠2 +6𝑠+9 𝐴 𝐵 𝐶 16 1 25 1 1 1
Thus, 𝑌(𝑠) = (𝑠−1)(𝑠−2)(𝑠+4) = 𝑠−1 + 𝑠−2 + 𝑠+4 = − (𝑠−1) + (𝑠−2) + 𝐶 = 30 (𝑠+4)
5 6

16 1 25 1 1 1
Or 𝑌(𝑠) = − (𝑠−1) + (𝑠−2) + 𝐶 = 30 (𝑠+4)
5 6

Taking the inverse Laplace transform on both sides, we have

16 1 25 1 1 1
ℒ −1 {𝑌(𝑠)} = ℒ −1 {− (𝑠−1)} + ℒ −1 { 6 (𝑠−2)} + ℒ −1 {30 (𝑠+4)}
5

16 1 25 1 1 1
𝑦(𝑡) = − ℒ −1 {𝑠−1} + ℒ −1 {𝑠−2} + 30 ℒ −1 {𝑠+4}
5 6

16 25 1
Or 𝑦(𝑡) = ℒ −1 {𝑌(𝑠)} = − 𝑒𝑡 + 𝑒 2𝑡 + 30 𝑒 −4𝑡
5 6

Hence, the solution of the initial-value problem is

16 25 1
𝑦(𝑡) = − 𝑒𝑡 + 𝑒 2𝑡 + 30 𝑒 −4𝑡
5 6

Badri A, Moges B. and Teklebrhan B. 117 AKU

 APPLIED MATHEMATICS III

Operational Properties of Laplace Transform

It is not convenient to use Definition of Laplace transform each time we wish to find the Laplace
Transform of a function f (t). For example, the integration by parts involved in evaluating,
say, ℒ{𝑒 𝑡 𝑡 2 𝑠𝑖𝑛3𝑡}is formidable, to say the least. In this section we present several labor-saving
Operational properties of the Laplace transform that enable us to build up a more extensive list of
Transforms without having to resort to the basic definition and integration.

1. Translation on The s-Axis

Evaluating transforms such as ℒ{𝑒 5𝑡 𝑡 3 } and ℒ{𝑒 −2𝑡 𝑐𝑜𝑠4𝑡} is straightforward provided that we
know (and we do know) ℒ{𝑡 3 } and ℒ{𝑐𝑜𝑠4𝑡}.in general, if we know the Laplace transform of a
function 𝑓, ℒ{𝑓(𝑡)} = 𝐹(𝑠), it is possible to compute the Laplace transform of an exponential
multiple of 𝑓,that is ℒ{𝑒 𝑎𝑡 𝑓(𝑡)},with no additional effort other than translating, or shifting, the
transform 𝐹(𝑠) to 𝐹(𝑠 − 𝑎).this result is known as the first translation theorem or first
shifting theorem.

Theorem (First translation theorem):

If ℒ{𝑓(𝑡)} = 𝐹(𝑠) and 𝑎 is any real number, then

ℒ{𝑒 𝑎𝑡 𝑓(𝑡)} = 𝐹(𝑠 − 𝑎).

Proof: By definition
∞ ∞
ℒ{𝑒 𝑎𝑡 𝑓(𝑡)} = ∫0 𝑒 −𝑠𝑡 𝑒 𝑎𝑡 𝑓(𝑡)𝑑𝑡 = ∫0 𝑒 −(𝑠−𝑎)𝑡 𝑓(𝑡)𝑑𝑡 = 𝐹(𝑠 − 𝑎).
Hence the proof

If we consider 𝑠 as areal variable, then the graph of 𝐹(𝑠 − 𝑎) is the graph of 𝐹(𝑠) shifted on the
𝑠-axis by amount|𝑎|. If 𝑎 > 0, the graph of 𝐹(𝑠) is shifted 𝑎 units to the right, where as if 𝑎 < 0,
is shifted |𝑎| units to the left as shown in the Fig 4.3.

Fig 4.3 Shift on 𝑠-axis

Badri A, Moges B. and Teklebrhan B. 118 AKU

 APPLIED MATHEMATICS III

Remark: For emphasis it is sometimes useful to use the symbolism

ℒ{𝑒 𝑎𝑡 𝑓(𝑡)} = ℒ{𝑓(𝑡)}𝑠→𝑠−𝑎

Where 𝑠 → 𝑠 − 𝑎 means that in the Laplace transform 𝐹(𝑠) of 𝑓(𝑡),we replace the symbol 𝑠
when ever it appears by 𝑠 − 𝑎.

Example 4.25: Evaluate each of the following

(a) ℒ{𝑒 5𝑡 𝑡 3 } (b) ℒ{𝑒 −2𝑡 𝑐𝑜𝑠4𝑡}
Solution: (a) By the first translation theorem, we have
3! 6
ℒ{𝑒 5𝑡 𝑡 3 } = ℒ{𝑡 3 } = =
s  s 5 𝑠4 s  s 5 (𝑠−5)4

(b) Similarly, by the first translation theorem

𝑠+2
ℒ{𝑒 −2𝑡 𝑐𝑜𝑠4𝑡} = ℒ{𝑐𝑜𝑠4𝑡} = ℒ{𝑐𝑜𝑠4𝑡} = (𝑠+2)2 +16
s  s  (2) ss2

Inverse form of the first translation theorem: To compute the inverse of 𝐹(𝑠 − 𝑎), we must
recognize 𝐹(𝑠),find f(t) by taking the inverse Laplace transform 𝐹(𝑠),and then multiply f(t) by
the exponential function 𝑒 𝑎𝑡 .this procedure can be summarized symbolically in the following
manner:

ℒ −1 {𝐹(𝑠 − 𝑎)} = ℒ −1 {𝐹(𝑠) } = 𝑒 𝑎𝑡 𝑓(𝑡) (1)

s sa
Where 𝑓(𝑡) = ℒ −1 {𝐹(𝑠)}.
Example 4.26: Solve the Initial-Value Problem
𝑦 ′′ − 6𝑦 ′ + 9𝑦 = 𝑡 2 𝑒 3𝑡 , 𝑦(0) = 2, 𝑦 ′ (0) = 17.
Solution: Using linearity and the initial conditions we simplify and then solve for
𝑌(𝑠) = ℒ{𝑓(𝑡)}:
ℒ{𝑦′′} − 6 ℒ{𝑦′} + ℒ{𝑦} = ℒ{𝑡 2 𝑒 3𝑡 }
Now, we use transform of derivatives together with the first translation theorem:
2
𝑠 2 𝑌(𝑠) − 𝑠𝑦(0) − 𝑦 ′ (0) − 6[𝑠𝑌(𝑠) − 𝑦(0)] + 9𝑌(𝑠) = (𝑠−3)3
2
(𝑠 2 − 6𝑠 + 9)𝑌(𝑠) = 2𝑠 + 5 + (𝑠−3)3
2
(𝑠 − 3)2 𝑌(𝑠) = 2𝑠 + 5 + (𝑠−3)3
2𝑠+5 2
𝑌(𝑠) = +
(𝑠−3)2 (𝑠−3)5
2𝑠+5
From this, decomposition into partial fraction on yields
(𝑠−3)2

Badri A, Moges B. and Teklebrhan B. 119 AKU

 APPLIED MATHEMATICS III

2 11 2
𝑌(𝑠) = 𝑠−3 + (𝑠−3)2 + (𝑠−3)5
Thus, taking the inverse Laplace transform on both sides gives
2 1 1
𝑦(𝑡) = ℒ −1 {𝑠−3} + 11ℒ −1 {(𝑠−3)2 }+2ℒ −1 {(𝑠−3)5 }
2 1 2 4!
𝑦(𝑡) = ℒ −1 {𝑠−3} + 11ℒ −1 {(𝑠−3)2 }+4! ℒ −1 {(𝑠−3)5 }

1 1 4!
𝑦(𝑡) = 2ℒ −1 {𝑠−3} + ℒ −1 {𝑠2 }+ℒ −1 {𝑠5 } (by the translation theorem)
s  s 3 s  s 3
1 4
𝑦(𝑡) = 2𝑒 3𝑡 + 11𝑡𝑒 3𝑡 + 12 𝑡𝑒 3𝑡 .
1 4
Hence, 𝑦(𝑡) = 2𝑒 3𝑡 + 11𝑡𝑒 3𝑡 + 12 𝑡𝑒 3𝑡 is the solution to the IVP
Example 4.27: Solve the Initial-Value Problem
𝑦 ′′ + 4𝑦 ′ + 6𝑦 = 1 + 𝑒 −𝑡 , 𝑦(0) = 0, 𝑦 ′ (0) = 0.
Solution: ℒ{𝑦′′} + ℒ{𝑦′} + 6ℒ{𝑦} = ℒ{1} + ℒ{𝑒 −𝑡 }
1 1
𝑠 2 𝑌(𝑠) − 𝑠𝑦(0) − 𝑦 ′ (0) + 4[𝑠 𝑌(𝑠) − 𝑦(0)] + 6 𝑌(𝑠) = + 𝑠+1
𝑠
2𝑠+1
(𝑠 2 + 4𝑠 + 6) 𝑌(𝑠) = 𝑠(𝑠+1)
2𝑠+1
𝑌(𝑠) = 𝑠(𝑠+1)(𝑠2 +4𝑠+6)
Since the quadratic term in the denominator does not factor into real linear factors, the partial
fraction decomposition for 𝑌(𝑠) is found to be
1⁄6 1⁄3 𝑠⁄2+5⁄3
𝑌(𝑠) = + 𝑠+1 − 𝑠2 +4𝑠+6
𝑠
Taking the Laplace transform on each side and completing the square on 𝑠 2 + 4𝑠 + 6
1 1 1 1 1 𝑠+2 2 √2
𝑦(𝑡) = ℒ −1 {𝑌(𝑠)} = 6 ℒ −1 {𝑠 } + 3 ℒ −1 {𝑠+1} − 2 ℒ −1 {(𝑠+2)2 +2} − 3√2 ℒ −1 {(𝑠+2)2 +2}
1 1 1 √2 −2𝑡
= 6 + 3 𝑒 −𝑡 − 2 𝑒 −2𝑡 𝑐𝑜𝑠√2𝑡 − 𝑒 𝑠𝑖𝑛√2
3

Exercises 4.3

1. Define each of the following

(a) Laplace transform (b) Exponentially ordered function (c) Inverse Laplace transform
(c) Piecewise continuous function

2. Using the definition of Laplace transform, Find ℒ{𝑓(𝑡)}

Badri A, Moges B. and Teklebrhan B. 120 AKU

 APPLIED MATHEMATICS III

(a). 𝑓 (𝑡) = 𝑡 2 + 6𝑡 − 3 (b). 𝑓(𝑡) = −4𝑡 2 + 16𝑡 + 5 (c). 𝑓(𝑡) = 1 + 𝑒 4𝑡 (d). 𝑓(𝑡) = 1 + 𝑒 4𝑡
(f) 𝑓(𝑡) = 4𝑡 2 − 5 𝑠𝑖𝑛 3𝑡 (g) 𝑓(𝑡) = 𝑒 𝑡 𝑠𝑖𝑛ℎ 𝑡 (h) 𝑓(𝑡) = (2𝑡 − 1)3 (i) 𝑓(𝑡) = 𝑠𝑖𝑛(4𝑡 + 5)

3. Find the Laplace transform of each of the following

−1,0 ≤ 𝑡 < 1 4, 0 ≤ 𝑡 < 1 2𝑡 + 1, 0 ≤ 𝑡 < 1
(a) 𝑓(𝑡) = { (b). 𝑓(𝑡) = { (c). 𝑓(𝑡) = {
1, 𝑡 ≥ 1 0 , 𝑡≥2 0 , 𝑡≥1
4. Find the inverse Laplace transform of each of the following
1 𝑠+1 𝑠+1 1
(a) ℒ −1 { } (b) ℒ −1 {𝑠2 −4𝑠} (𝑐) ℒ −1 {𝑠2 −2𝑠−3} (d)ℒ −1 {𝑠2 +𝑠−20}
𝑠2 +3𝑠
5. Use the Laplace transform to solve the given initial-value problem.
𝑑𝑦 𝑑𝑦
(a) − 𝑦 = 5,𝑦(0) = 0 (b) 2 𝑑𝑡 + 𝑦 = 0,𝑦(0) = −3 (c) 𝑦′ + 6𝑦 = 𝑒 4𝑡 , 𝑦(0) = 2
𝑑𝑡
(d) 𝑦 ′′ − 4𝑦 ′ = 4𝑒 3𝑡 − 3𝑒 −𝑡 , 𝑦(0) = 1, 𝑦 ′ (0) = −1
6. Evaluate each of the following
(a) ℒ{𝑡𝑒 10𝑡 } (b). ℒ{𝑡 3 𝑒 −2𝑡 } (c). ℒ{𝑒 −2𝑡 cos 4𝑡} (d) ℒ{𝑒 𝑡 sin 3𝑡}
7. Use the Laplace transform to solve the given initial-value problem
(a) 𝑦 ′′ + 2𝑦 ′ + 𝑦 = 0, 𝑦(0) = 1, 𝑦 ′ (0) = 1 (b) 𝑦 ′′ − 4𝑦 ′ + 4𝑦 = 𝑡 3 𝑒 2𝑡 , 𝑦(0) = 0,𝑦 ′ (0) = 0

4.4 Differentiation and Integration of Laplace Transformation

Differentiation of Laplace transformation

Suppose that 𝐹(𝑠) = ℒ{𝑓(𝑡)} exists and that it is possible to interchange the order of
differentiation and integration, then
𝑑 𝑑 ∞ −𝑠𝑡 ∞ 𝜕 ∞
𝐹(𝑠) = ∫ 𝑒 𝑓(𝑡)𝑑𝑡 = ∫0 [𝑒 −𝑠𝑡 𝑓(𝑡)]𝑑𝑡 = − ∫0 𝑒 −𝑠𝑡 𝑡𝑓(𝑡)𝑑𝑡 = −ℒ{𝑡𝑓(𝑡)}
𝑑𝑠 𝑑𝑠 0 𝜕𝑠

Badri A, Moges B. and Teklebrhan B. 121 AKU

 APPLIED MATHEMATICS III

𝑑 𝑑
That is ℒ{𝑡𝑓(𝑡)} = − 𝑑𝑠 𝐹(𝑠) = − 𝑑𝑠 ℒ{𝑓(𝑡)}

𝑑 𝑑 𝑑
Similarly by the above resultℒ{𝑡 2 𝑓(𝑡)} = ℒ{𝑡. 𝑡𝑓(𝑡)} = − 𝑑𝑠 ℒ{𝑡𝑓(𝑡)} = − 𝑑𝑠 (− 𝑑𝑠 ℒ{𝑓(𝑡)})

𝑑2
= ℒ{𝑓(𝑡)}
𝑑𝑠2

From this we can generalize to the following theorem

Theorem (Derivatives Transforms)

𝑑𝑛
If 𝐹(𝑠) = ℒ{𝑓(𝑡)} and 𝑛 = 1,2,3, … ., then ℒ{𝑡 𝑛 𝑓(𝑡)} = (−1)𝑛 𝑑𝑠𝑛 ℒ{𝑓(𝑡)}

Example 4.28: Evaluate ℒ{𝑡𝑠𝑖𝑛𝑘𝑡}

𝑘
Solution: Here (𝑡) = 𝑠𝑖𝑛𝑘𝑡 , 𝑛 = 1 and 𝐹(𝑠) = 𝑠2 +𝑘 2 , then by the above theorem

𝑑 𝑑
𝑑 𝑑 𝑘 (𝑠2 +𝑘 2 ) (𝑘)−𝑘 (𝑠2 +𝑘 2 ) (𝑠2 +𝑘 2 )(0)−2𝑘𝑠
ℒ{𝑡𝑠𝑖𝑛𝑘𝑡} = (−1)1 𝑑𝑠 ℒ{𝑠𝑖𝑛𝑘𝑡} = − 𝑑𝑠 (𝑠2 +𝑘 2 ) = −( 𝑑𝑠 𝑑𝑠
)= −( )
(𝑠2 +𝑘 2 )2 (𝑠2 +𝑘 2 )2

(𝑠2 +𝑘 2 )(0)−2𝑘𝑠 −(−2𝑘𝑠) 2𝑘𝑠

=−( )= = (𝑠2 +𝑘 2 )2
(𝑠2 +𝑘 2 )2 (𝑠2 +𝑘 2 )2

2𝑘𝑠
Hence ℒ{𝑡𝑠𝑖𝑛𝑘𝑡} = (𝑠2 +𝑘 2 )2

Example 4.29: Evaluate ℒ{𝑡𝑒 3𝑡 }

1
Solution: 𝑓(𝑡) = 𝑒 3𝑡 , 𝐹(𝑠) = , 𝑠 > 3 and 𝑛 = 1
𝑠−3

𝑑 𝑑
𝑑 𝑑 1 (𝑠−3) (1)−1. (𝑠−3) (𝑠−3)(0)−1 −(−1)
ℒ{𝑡𝑒 3𝑡 } = (−1)1 𝑑𝑠 ℒ{𝑒 3𝑡 } = − 𝑑𝑠 (𝑠−3) = −( 𝑑𝑠 𝑑𝑠
) = −( ) = (𝑠−3)2
(𝑠−3)2 (𝑠−3)2

−1
= (𝑠−3)2

Activity Evaluate each of the following

a) ℒ{𝑡 2 𝑠𝑖𝑛5𝑡} b. {𝑡 3 𝑐𝑜𝑠3𝑡} c. ℒ{𝑡 3 𝑒 3𝑡 } d. ℒ{4𝑡𝑠𝑖𝑛ℎ2𝑡 + 7𝑡 2 𝑒 −5𝑡 }

Integration of Laplace Transformation

Integration of the transform of a function 𝑓(𝑡) corresponds to the division of 𝑓(𝑡) by 𝑡

Theorem (The integral of a transform)

Badri A, Moges B. and Teklebrhan B. 122 AKU

 APPLIED MATHEMATICS III

Let 𝑓(𝑡)⁄𝑡 be a piecewise continuous, defined for t ≥ 0 and such that |𝑓(𝑡)⁄𝑡| ≤ 𝑀𝑒 −𝑘𝑡 for t ≥
0, if ℒ{𝑓(𝑡)⁄𝑡 } = 𝐺(𝑆) for 𝑠 > 𝑘 and ℒ{𝑓(𝑡)} = 𝐹(𝑆) ,
∞ −1
ℒ{𝑓(𝑡)⁄𝑡 } = ∫𝑠 𝐹(𝑢) 𝑑𝑢 and conversely ℒ −1 {𝐺(𝑠)} = ℒ −1 ℒ{𝐺 ′ (𝑠)}
𝑡

∞
Proof: we have 𝐺(𝑆) = ℒ{𝑓(𝑡)⁄𝑡 } = ∫0 𝑒 −𝑠𝑡 𝑓(𝑡)⁄𝑡 𝑑𝑡 for 𝑠 > 𝑘

∞ 𝑓(𝑡) ∞
However 𝐺 ′ (𝑠) = ∫0 𝑒 −𝑠𝑡 (−𝑡) 𝑑𝑡 = − ∫0 𝑒 −𝑠𝑡 𝑓(𝑡)𝑑𝑡 = − 𝐹(𝑆)
𝑡

Now after integration we have

∞ ∞
∫𝑠 𝐹(𝑢) 𝑑𝑢 = − ∫𝑠 𝐺 ′ (𝑢) 𝑑𝑢 = −(𝐺(∞) − 𝐺(𝑠) ) = 𝐺(𝑠) − 𝐺(∞)

To proceed further we now make use of the fact that the condition |𝑓(𝑡)⁄𝑡| ≤ 𝑀𝑒 −𝑘𝑡 implies
∞
that lim 𝐺(𝑠) = 0 , showing that 𝐺(𝑠) = ℒ{𝑓(𝑡)⁄𝑡 } = ∫𝑠 𝐹(𝑢) 𝑑𝑢 , for 𝑠 > 𝑘
𝑠→∞

The converse result follows by taking the inverse Laplace transform and using the fact that
{𝐹(𝑠)} 1
ℒ −1 {𝐺(𝑠)} = 𝑓(𝑡)⁄𝑡 = ℒ −1 = − 𝑡 ℒ −1 {𝐺 ′ (𝑠)} Together with ℒ{𝑓(𝑡) } = 𝐹(𝑠) = −𝐺 ′ (𝑠)
𝑡

𝑒 −2𝑡 −𝑒 −3𝑡
Example 4.30: Evaluate ℒ { }
𝑡

𝑒 −2𝑡 −𝑒 −3𝑡
Solution: The function is defined and finite for all > 0 , where 𝑓(𝑡) = 𝑒 −2𝑡 − 𝑒 −3𝑡
𝑡
1 1
ℒ{𝑓(𝑡)} = ℒ{𝑒 −2𝑡 − 𝑒 −3𝑡 } = 𝐹(𝑠) = 𝑠+2 + 𝑠+3

𝑒 −2𝑡 −𝑒 −3𝑡 ∞ ∞ 1 1
𝐺(𝑠) = ℒ{𝑓(𝑡)⁄𝑡 } = ℒ { } = ∫𝑠 𝐹(𝑢) 𝑑𝑢 = ∫𝑠 (𝑢+2 + 𝑢+3 ) 𝑑𝑢
𝑡

∞ 1 ∞ 1 ∞ ∞
= ∫𝑠 𝑑𝑢 + ∫𝑠 𝑑𝑢 = [ln(𝑢 + 2)] − [ln(𝑢 + 3)]
𝑢+2 𝑢+3 𝑠 𝑠
= ln(∞ + 2) − ln(𝑠 + 2) − ln(∞ + 3) + ln(𝑠 + 3)
𝑠+3
= ln(𝑠 + 3) − ln(𝑠 + 2) = ln (𝑠+2) because ln(∞ + 2) = ln(∞ + 3) = 0 (∴ lim 𝐺(𝑠) = 0 ) ⟹
𝑠→∞
𝑠+3
𝐺(𝑠) = ln (𝑠+2)

𝑒 −2𝑡 −𝑒 −3𝑡 𝑠+3

Hence 𝐺(𝑠) = ℒ { } = ln (𝑠+2)
𝑡

𝑠+3
Example 4.31: Evaluate ℒ −1 {ln (𝑠+2)}

𝑠+3 𝑑 𝑠+3 𝑑
Solution: let𝐺(𝑠) = ln (𝑠+2), then 𝐺 ′ (𝑠) = 𝑑𝑠 ln (𝑠+2) = 𝑑𝑠 (ln(𝑠 + 3) − ln(𝑠 + 2))

Badri A, Moges B. and Teklebrhan B. 123 AKU

 APPLIED MATHEMATICS III

𝑑 𝑑
𝑑 𝑑 (𝑠+3) (𝑠+2) 1 1
𝑑𝑠 𝑑𝑠
= 𝑑𝑠 ln(𝑠 + 3) − 𝑑𝑠 ln(𝑠 + 2) = − = 𝑠+3 − 𝑠+2
𝑠+3 𝑠+2

1 1
⟹ 𝐺 ′ (𝑠) = 𝑠+3 − 𝑠+2

1 1 1 1
Now ℒ −1 {𝐺 ′ (𝑠)} = ℒ −1 {𝑠+3 − 𝑠+2 } = ℒ −1 {𝑠+3} − ℒ −1 {𝑠+2} = e−3t − e−2t

From the second part of the above theorem we have

𝑠+3 −1 −1
ℒ −1 {𝐺(𝑠)} = ℒ −1 {ln ( )} = ℒ −1 ℒ{𝐺 ′ (𝑠)} = (e−3t − e−2t ) = (e−2t − e−3t )⁄𝑡
𝑠+2 𝑡 𝑡

𝑠+3
Hence ℒ −1 {ln (𝑠+2)} = (e−2t − e−3t )⁄𝑡

Exercises 4.4

1. In each of the following use the Derivatives Transforms to evaluate ℒ{𝑓(𝑡)}

(a) 𝑡 𝑐𝑜𝑠𝑤𝑡 (b) 𝑡 2 𝑠𝑖𝑛 3𝑡 (c) 𝑡 2 𝑐𝑜𝑠ℎ 2𝑡 (d) 𝑡𝑒 −𝑘𝑡 𝑠𝑖𝑛 𝑡 (e) 𝑡 𝑛 𝑒 𝑘𝑡

2. Use the integral of a transform to find the Laplace transform of each of the following

𝑠𝑖𝑛 3𝑡 𝑒 −5𝑡 𝑐𝑜𝑠 8𝑡

(a) ℒ { 𝑡
} (b). ℒ { 𝑡
} (c). ℒ { 𝜋𝑡
}

𝑠+𝑎 𝑒 −𝑏𝑡 −𝑒 −𝑎𝑡

3. Show that ℒ −1 {𝑙𝑛 (𝑠+𝑏 ) = 𝑡

Badri A, Moges B. and Teklebrhan B. 124 AKU

 APPLIED MATHEMATICS III

4.5 Convolution and Integral Equations

4.5.1 Convolution
Definition: let the functions 𝑓(𝑡) and 𝑔(𝑡) be defined for 𝑡 ≥ 0. Then the convolution of the
functions 𝑓 and 𝑔 denoted by (𝑓 ∗ 𝑔)(𝑡), and in abbreviated form by 𝑓 ∗ 𝑔 is a function of t
defined as the integral;
𝑡
(𝑓 ∗ 𝑔)(𝑡) = ∫0 𝑓(𝜏)𝑔(𝑡 − 𝜏)𝑑𝜏

Note: From the definition it follows almost immediately the convolution has the properties

Badri A, Moges B. and Teklebrhan B. 125 AKU

 APPLIED MATHEMATICS III

𝑓∗𝑔 = 𝑔∗𝑓 (commutative law)

𝑓 ∗(𝑔1 + 𝑔2 )= 𝑓 ∗ 𝑔1 + 𝑓 ∗ 𝑔2 (distributive law)

(𝑓 ∗ 𝑔) ∗ ℎ = 𝑓 ∗ (𝑔 ∗ ℎ) (associative law)

𝑓∗0= 0∗𝑓

similar to those of multiplication of numbers. However, there are differences of which you
should be aware.

Example 4.32: if 𝑓(𝑡) = 𝑒 𝑡 and 𝑔(𝑡) = 𝑠𝑖𝑛𝑡, then find 𝑓 ∗ 𝑔

𝑡 𝑡
Solution: 𝑓 ∗ 𝑔 = (𝑔 ∗ 𝑓)(𝑡) = ∫0 𝑓(𝜏)𝑔(𝑡 − 𝜏)𝑑𝜏 = ∫0 𝑒 𝜏 sin(𝑡 − 𝜏)𝑑𝜏

Now using integration by parts we have,

let = 𝑒 𝜏 , 𝑑𝑢 = 𝑒 𝜏 𝑑𝜏 and let 𝑑𝑣 = sin(𝑡 − 𝜏) , 𝑣 = cos(𝑡 − 𝜏),

𝑡 𝑡 𝑡
𝑓 ∗ 𝑔 = ∫0 𝑒 𝜏 sin(𝑡 − 𝜏)𝑑𝜏 = [𝑒 𝜏 cos(𝑡 − 𝜏)] − ∫0 𝑒 𝜏 cos(𝑡 − 𝜏)𝑑𝜏 ……………… (1)
0
𝑡
Again using integrating parts on ∫0 𝑒 𝜏 cos(𝑡 − 𝜏)𝑑𝜏 in (1) we have

Let 𝑢 = 𝑒 𝜏 , 𝑑𝑢 = 𝑒 𝜏 𝑑𝜏 and Let 𝑑𝑣 = cos(𝑡 − 𝜏), 𝑣 = −sin(𝑡 − 𝜏) ,then

𝑡 𝑡 𝑡
∫0 𝑒 𝜏 cos(𝑡 − 𝜏)𝑑𝜏 = [−𝑒 𝜏 sin(𝑡 − 𝜏)] + ∫0 𝑒 𝜏 sin(𝑡 − 𝜏)𝑑𝜏 ………………….. (2)
0
Now substituting (2) into (1) we have,
𝑡 𝑡 𝑡 𝑡
∫0 𝑒 𝜏 sin(𝑡 − 𝜏)𝑑𝜏 = [𝑒 𝜏 cos(𝑡 − 𝜏)] − ([−𝑒 𝜏 sin(𝑡 − 𝜏)] + ∫0 𝑒 𝜏 sin(𝑡 − 𝜏)𝑑𝜏]
0 0
𝑡 𝑡 𝑡
= [𝑒 𝜏 cos(𝑡 − 𝜏)] + [𝑒 𝜏 sin(𝑡 − 𝜏)] − ∫0 𝑒 𝜏 sin(𝑡 − 𝜏)𝑑𝜏)
0 0
𝑡 𝑡 𝑡
⟹ 2 ∫0 𝑒 𝜏 sin(𝑡 − 𝜏)𝑑𝜏 = [𝑒 𝜏 cos(𝑡 − 𝜏)] + [𝑒 𝜏 sin(𝑡 − 𝜏)] )
0 0
𝑡 1 𝑡 𝑡
⟹ ∫0 𝑒 𝜏 sin(𝑡 − 𝜏)𝑑𝜏 = 2 [𝑒 𝜏 cos(𝑡 − 𝜏)] + [𝑒 𝜏 sin(𝑡 − 𝜏)] )
0 0
1 1
= 2 (𝑒 𝑡 cos(𝑡 − 𝑡) − 𝑒 0 cos(𝑡 − 0)) + 2 (𝑒 𝑡 𝑠𝑖𝑛(𝑡 − 𝑡)𝑒 0 − 𝑒 0 𝑠𝑖𝑛(𝑡 − 0))

1 1
= 2 (𝑒 𝑡 − cos 𝑡) + 2 (− 𝑠𝑖𝑛 𝑡)

1
= 2 (𝑒 𝑡 − cos 𝑡 − 𝑠𝑖𝑛 𝑡)

Badri A, Moges B. and Teklebrhan B. 126 AKU

 APPLIED MATHEMATICS III

𝑡 𝑡 1
Hence 𝑓 ∗ 𝑔 = (𝑔 ∗ 𝑓)(𝑡) = ∫0 𝑓(𝜏)𝑔(𝑡 − 𝜏)𝑑𝜏 = ∫0 𝑒 𝜏 sin(𝑡 − 𝜏)𝑑𝜏 = 2 (𝑒 𝑡 − cos 𝑡 − 𝑠𝑖𝑛 𝑡)

Theorem (Convolution theorem)

If 𝑓(𝑡) and 𝑔(𝑡) are piecewise continuous on [0, ∞) and of exponential order, then

ℒ{𝑓 ∗ 𝑔} = ℒ{𝑓(𝑡)}. ℒ{𝑔(𝑡)} = 𝐹(𝑠)𝐺(𝑠)

or equivalently
𝑡
ℒ −1 {𝐹(𝑠)𝐺(𝑠)} = 𝑓 ∗ 𝑔 = ∫0 𝑓(𝜏)𝑔(𝑡 − 𝜏)𝑑𝜏

Proof
∞ ∞
Let 𝐹(𝑠) = ∫0 𝑒 −𝑠𝜏 𝑓(𝜏)𝑑𝜏 and 𝐺(𝑠) = ∫0 𝑒 −𝑠𝑝 𝑔(𝑝)𝑑𝑝

We now set 𝑡 = 𝑝 + 𝜏, where 𝜏 is at first constant. Then 𝑝 = 𝑡 − 𝜏, and 𝑡 varies from 𝜏 to

∞,Thus,
∞ ∞
𝐺(𝑠) = ∫𝜏 𝑒 −𝑠(𝑡− 𝜏) 𝑔( 𝑡 − 𝜏)𝑑𝑡 = 𝑒 𝑠𝜏 ∫𝜏 𝑒 −𝑠𝑡 𝑔(𝑡 − 𝜏)𝑑𝑡

𝜏 in 𝐹 and 𝑡 in 𝐺 vary independently. Hence we can insert the G-integral into the F-integral.

Cancellation of 𝑒 −𝑠𝜏 and 𝑒 𝑠𝜏 then gives

∞ ∞ ∞ ∞
𝐹(𝑠)𝐺(𝑠) = ∫0 𝑒 −𝑠𝜏 𝑓(𝜏) 𝑒 𝑠𝜏 ∫𝜏 𝑒 −𝑠𝑡 𝑔(𝑡 − 𝜏)𝑑𝑡𝑑𝜏 = ∫0 𝑓(𝜏) ∫𝜏 𝑒 −𝑠𝑡 𝑔(𝑡 − 𝜏)𝑑𝑡𝑑𝜏

Here we integrate for fixed 𝜏 over 𝑡 from 𝜏 to ∞.this is the shaded region in Fig 4.4. Under the
assumption on f and g the order of integration can be reversed. We then integrate first over 𝜏

From 𝑜 to 𝑡 and then over 𝑡 from 0 to ∞, that is,

∞ ∞ ∞
𝐹(𝑠)𝐺(𝑠) = ∫ 𝑒 −𝑠𝜏 ∫ 𝑓(𝜏)𝑔(𝑡 − 𝜏)𝑑𝜏𝑑𝑡 = ∫ 𝑒 −𝑠𝜏 ℎ(𝑡)𝑑𝑡 = ℒ{ℎ} = 𝐻(𝑠
0 𝜏 0

Figure 4.4, region of integration in the 𝑡𝜏-plane.

Badri A, Moges B. and Teklebrhan B. 127 AKU

 APPLIED MATHEMATICS III

Example 4.33: if 𝑓(𝑡) = 𝑒 𝑡 and 𝑔(𝑡) = 𝑠𝑖𝑛𝑡 , then the convolution theorem states that the
Laplace transform of the convolution of 𝑓 and 𝑔 is the product of their Laplace transforms.
𝑡
That is ℒ{𝑓 ∗ 𝑔} = ℒ {∫0 𝑒 𝜏 sin(𝑡 − 𝜏) 𝑑𝜏} = ℒ{𝑓(𝑡)}. ℒ{𝑔(𝑡)} = ℒ{𝑒 𝑡 }. ℒ{𝑠𝑖𝑛𝑡}

1 1 1
= = (𝑠−1)(𝑠2 +1)
𝑠−1 𝑠2 +1

𝒔
Example 4.34: Evaluate a) ℒ{𝑡 2 ∗ 𝑐𝑜𝑠𝑡} b) 𝓛−𝟏 {(𝒔𝟐 +𝒂𝟐 )𝟐 }

2
Solution: a) Here 𝑓(𝑡) = 𝑡 2 and ℒ{𝑓(𝑡)} = ℒ{𝑡 2 } = 𝑠3 and 𝑔(𝑡) = 𝑐𝑜𝑠𝑡 , ℒ{𝑔(𝑡)} = ℒ{𝑐𝑜𝑠𝑡}

𝑠
= , then by the convolution theorem
𝑠2 +1

2 𝑠 2𝑠 2
ℒ{𝑡 2 ∗ 𝑐𝑜𝑠𝑡} = ℒ{𝑡 2 }ℒ{𝑐𝑜𝑠𝑡} = = 𝑠3 (𝑠2 +1) =
𝑠3 𝑠2 +1 𝑠2 (𝑠2 +1)

𝒔 𝟏 𝒔 𝟏 𝒔
b) Writing 𝟐 = 𝒔𝟐 +𝒂𝟐 and letting 𝐹(𝑠) = 𝒔𝟐 +𝒂𝟐 and 𝐺(𝑠) = 𝒔𝟐 +𝒂𝟐 , we have
(𝒔𝟐 +𝒂𝟐 ) 𝒔𝟐 +𝒂𝟐

𝟏 𝒂 1 𝒂 1
𝓛−𝟏 {𝐹(𝑠)} = 𝓛−𝟏 {𝒔𝟐 +𝒂𝟐 } = 𝓛−𝟏 {𝒂(𝒔𝟐 +𝒂𝟐)} = 𝑎 𝓛−𝟏 {𝒔𝟐 +𝒂𝟐 } = 𝑠𝑖𝑛𝑎𝑡
𝑎

𝒔
Similarly,𝓛−𝟏 {𝐺(𝑠)} = 𝓛−𝟏 {𝒔𝟐 +𝒂𝟐 } = 𝑐𝑜𝑠𝑎𝑡, then it follows from the convolution theorem that

𝑠 1
ℒ −1 {(𝑠2 +𝑎2 )2 } = ℒ −1 {𝐹(𝑠)𝐺(𝑠)} = (1⁄𝑎 )(𝑠𝑖𝑛𝑎𝑡 ∗ 𝑐𝑜𝑠𝑎𝑡) = 2𝑎 𝑡 𝑠𝑖𝑛𝑎𝑡

𝑡
Note: when evaluating convolution integrals of ∫0 𝑠𝑖𝑛𝑎𝜏𝑐𝑜𝑠𝑎(𝑡 − 𝜏)𝑑𝜏 , instead of expanding a
term such as 𝑐𝑜𝑠𝑎(𝑡 − 𝜏) and 𝑠𝑖𝑛𝑎(𝑡 − 𝜏) using integration by parts, it is often quicker to
𝑒 𝑖𝑎𝑡 +𝑒 −𝑖𝑎𝑡 𝑒 −𝑖𝑎(𝑡−𝜏)𝑡 +𝑒 −𝑖(𝑡−𝜏)𝑎𝑡
replace 𝑠𝑖𝑛𝑎𝑡 and 𝑐𝑜𝑠𝑎(𝑡 − 𝜏) by 𝑠𝑖𝑛𝑎𝑡 = , 𝑐𝑜𝑠𝑎(𝑡 − 𝜏) =
2𝑖 2

Before performing the integrations, and again using these identities to interpret the result in
terms of trigonometric functions

4.5.2 Integral Equations

Convolution helps in solving certain integral equations, that is, equations in which the unknown
function 𝑦(𝑡) appears in an integral (and perhaps also outside of it).This concerns equations with
an integral of the form of a convolution.

Volterra Integral Equation

Badri A, Moges B. and Teklebrhan B. 128 AKU

 APPLIED MATHEMATICS III

Definition (Volterra Integral Equation): The integral equation of the form

𝑡
𝑓(𝑡) = 𝑔(𝑡) + ∫0 𝑓(𝜏)ℎ(𝑡 − 𝜏)𝑑𝜏,

Where 𝑔(𝑡) and ℎ(𝑡) are known functions is called Volterra Integral Equation for𝑓(𝑡).

Note: Volterra Integral Equation has the convolution form with the symbol ℎ playing the part of
𝑔 in convolution.
𝑡
Example 4.35: solve 𝑓(𝑡) = 3𝑡 2 − 𝑒 −𝑡 − ∫0 𝑓(𝜏)𝑒 𝑡−𝜏 𝑑𝜏 for 𝑓(𝑡)

Solution: Here 𝒉(𝒕 − 𝝉) = 𝑒 𝑡−𝜏 so that 𝒉(𝒕) = 𝑒 𝑡 and 𝒈(𝒕) = 3𝑡 2 − 𝑒 −𝑡

Taking the Laplace transform of each term, we have

𝑡 𝑡
ℒ{𝑓(𝑡)} = ℒ{3𝑡 2 − 𝑒 −𝑡 } − ℒ {∫0 𝑓(𝜏)𝑒 𝑡−𝜏 𝑑𝜏} = ℒ{3𝑡 2 } − ℒ{𝑒 −𝑡 } − ℒ {∫0 𝑓(𝜏)𝑒 𝑡−𝜏 𝑑𝜏}

2 1 2 1
⟹ 𝐹(𝑠) =3.𝑠3 − 𝑠+1 − ℒ{𝑓(𝑡)}ℒ{ℎ(𝑡)} =3.𝑠3 − 𝑠+1 − ℒ{𝑓(𝑡)}ℒ{𝑒 𝑡 }

2 1 6 1 1
= 3.𝑠3 − 𝑠+1 − ℒ{𝑓(𝑡)}ℒ{𝑒 −𝑡 } = 𝑠3 − 𝑠+1 − 𝐹(𝑠) 𝑠−1

1 6 1 𝑠−1+1 6 1 𝑠 6 1
⟹ (1 + 𝑠−1) 𝐹(𝑠) = 𝑠3 − 𝑠+1 ⟹ ( ) 𝐹(𝑠) = 𝑠3 − 𝑠+1 ⟹ 𝐹(𝑠) = 𝑠3 − 𝑠+1
𝑠−1 𝑠−1

𝑠−1 6 1 6(𝑠−1) 𝑠−1 6 6 𝑠−1

⟹ 𝐹(𝑠) = (𝑠3 − 𝑠+1) = − 𝑠(𝑠+1) = − 𝑠4 − 𝑠(𝑠+1) …………………….(1)
𝑠 𝑠4 𝑠3

𝑠−1
Decomposing 𝑠(𝑠+1) in to partial fractions we have

𝑠−1 𝐴 𝐵 𝐴(𝑠+1)+𝐵𝑠 (𝐴+𝐵)𝑠+𝐴 𝐴+𝐵 =1

= 𝑠 + 𝑠+1 = = ⟹ (𝐴 + 𝐵)𝑠 + 𝐴 = 𝑠 − 1 { ⟹ 𝐴 = −1
𝑠(𝑠+1) 𝑠(𝑠+1) 𝑠(𝑠+1) 𝐴 = −1
and 𝐵 = 2
𝑠−1 −1 2
Now = + 𝑠+1 from this (1) becomes
𝑠(𝑠+1) 𝑠

6 6 𝑠−1 6 6 1 2
𝐹(𝑠) = 𝑠3 − 𝑠4 − 𝑠(𝑠+1) = 𝑠3 − 𝑠4 + 𝑠 − 𝑠+1 taking the inverse Laplace transform of each term
we have
6 6 𝑠−1 6 6 1 2 6 6 1
ℒ −1 {𝐹(𝑠)} = ℒ −1 {𝑠3 − 𝑠4 − 𝑠(𝑠+1) = 𝑠3 − 𝑠4 + 𝑠 − 𝑠+1} = ℒ −1 {𝑠3 } − ℒ −1 {𝑠4 } + ℒ −1 {𝑠 } −
2
ℒ −1 {𝑠+1}

2! 3! 1 1
⟹ 𝑓(𝑡) = 3ℒ −1 {𝑠3 } − ℒ −1 {𝑠4 } + ℒ −1 {𝑠 } − 2 {𝑠+1} = 3t 2 − t 3 + 1 − 2e−t

Badri A, Moges B. and Teklebrhan B. 129 AKU

 APPLIED MATHEMATICS III

Hence 𝑓(𝑡) = 3t 2 − t 3 + 1 − 2e−t

Integro-differential equations
We now consider a differential equation of an unusual type, these equations occur in many
applications of mathematics, one of which arises in the continuum mechanics of polymers, where
the dynamical response 𝑦(𝑡) of certain types of material at time 𝑡 depends on a derivative of
𝑦(𝑡) and the time-weighed cumulative effect of what has happened to the material prior to time
𝑡.for obvious reasons materials of this type are called materials with memory.

Definition: Differential equations in which the function 𝑦(𝑡) occurs not only as the dependent
variable in the differential equation, but also inside a convolution integral that forms the
Nonhomogeneous term are called Integro-differential equations.

In other words, equations that involve both the integral of an unknown function and its
derivatives are called Integro-differential equations.

Example 4.36: Solve the equation

𝑡
𝑦 ′′ + 𝑦 = ∫0 sin 𝜏 𝑦(𝑡 − 𝜏)𝑑𝜏,

Subject to the initial conditions 𝑦(0) = 1 and 𝑦′(0) = 0

Solution taking the Laplace transform in the usual way gives

𝑡
𝑠 2 𝑌(𝑠) − 𝑠 + 𝑌(𝑠) = ℒ {∫ sin 𝜏 𝑦(𝑡 − 𝜏)𝑑𝜏}
0

Here the last term is the Laplace of a convolution integral, so from the convolution theorem it
follows that
𝑡 𝑡 𝑡 𝑌(𝑠)
ℒ {∫0 sin 𝜏 𝑦(𝑡 − 𝜏)𝑑𝜏} = ℒ {∫0 sin 𝑡} ℒ {∫0 𝑦(𝑡)} = 𝑠2 +1

Using this result in the transformed equation, solving for 𝑌(𝑠), and expanding the result using
partial fractions gives
𝑌(𝑠) 𝑌(𝑠)
𝑠 2 𝑌(𝑠) − 𝑠 + 𝑌(𝑠) = 𝑠2 +1 or (𝑠 2 + 1) 𝑌(𝑠) = 𝑠2 +1 + 𝑠

𝑠2 +1 11 1 𝑠
i.e. 𝑌(𝑠) = 𝑠(𝑠2 +2) = 2 𝑠 + 2 (𝑠2 +2)

After the inverse Laplace transform is taken, the solution becomes

1
𝑦(𝑡) = (1 + 𝑐𝑜𝑠√2𝑡) for 𝑡 > 0
2

Badri A, Moges B. and Teklebrhan B. 130 AKU

 APPLIED MATHEMATICS III

Exercises 4.5

1. Define each of the following

(a) Convolution (b) Volterra Integral Equation (c) Integro-differential equations

2. If 𝑓(𝑡) = 𝑡 2 and 𝑔(𝑡) = 𝑐𝑜𝑠𝑡 , then show that (𝑓 ∗ 𝑔)(𝑡) = 2(𝑡 − 𝑠𝑖𝑛𝑡) = (𝑔 ∗ 𝑓)(𝑡)

3. Use the convolution theorem to evaluate each of the following

(a) ℒ{1 ∗ 𝑡 3 } (b). ℒ{𝑡 2 ∗ 𝑡𝑒 𝑡 } (c). ℒ{𝑒 −𝑡 ∗ 𝑒 𝑡 𝑐𝑜𝑠𝑡} (d). ℒ{𝑒 2𝑡 ∗ 𝑠𝑖𝑛 𝑡}

4. In each of the following use the Laplace transform to solve the given integral equation or Integro-
differential equation.
𝑡 𝑡
(a) 𝑓(𝑡) + ∫0 (𝑡 − 𝜏)𝑓(𝜏)𝑑𝜏 = 𝑡 (b) 𝑓(𝑡) = 2𝑡 − 4 ∫0 𝑠𝑖𝑛 𝜏 𝑓(𝑡 − 𝜏)𝑑𝜏

𝑡
(c) 𝑦 ′ (𝑡) = 1 − 𝑠𝑖𝑛𝑡 − ∫0 𝑦(𝜏)𝑑𝜏 , 𝑦(0) = 0

𝑑𝑦 𝑡
(d) 𝑑𝑡
+ 6𝑦(𝑡) + 9 ∫0 𝑦(𝜏)𝑑𝜏 = 1, 𝑦(0) = 0

Unit Summary:

 The Fourier Cosine and Sine Transforms can be considered as a special cases of the
Fourier transform 𝑜𝑓 𝑓(𝑥) when 𝑓(𝑥) is even or odd function over the real axis
 Fourier cosine and Fourier sine transforms of f(x) denoted by
𝐹𝐶 (𝑤) 𝑜𝑟𝑓𝑐 ^ 𝑎𝑛𝑑 𝐹𝑠 (𝑤) 𝑜𝑟 𝑓𝑠 ^ respectively is defined as;

2 ∞ 2 ∞
𝐹𝑐 (𝑤) = √𝜋 ∫0 𝑓(𝑥) 𝑐𝑜𝑠𝑤𝑥𝑑𝑥 𝐹𝑠 (𝑤) = √𝜋 ∫0 𝑓(𝑥) 𝑠𝑖𝑛𝑤𝑥𝑑𝑥

 The Fourier cosine and sine transform are linear transforms i.e. for any two functions
𝑓(𝑥) and 𝑔(𝑥) whose Fourier cosine and sine transform exist and for any constants a and
b, then

(a) 𝑓𝑐 ^ [𝑎𝑓(𝑥) + 𝑏𝑔(𝑥)] = 𝑎𝑓𝑐 ^ [𝑓(𝑥)] + 𝑏𝑓𝑐 ^ [𝑔(𝑥)] and

Badri A, Moges B. and Teklebrhan B. 131 AKU

 APPLIED MATHEMATICS III

(𝑏) 𝑓𝑠 ^ [𝑎𝑓(𝑥) + 𝑏𝑔(𝑥)] = 𝑎𝑓𝑠 ^ [𝑓(𝑥)] + 𝑏𝑓𝑠 ^ [𝑔(𝑥)]

 Let 𝑓(𝑥) and 𝑓′(𝑥) be continuous and absolutely integrable on the interval[0, ∞) and
𝑓′′(𝑥) be piecewise continuous on every subinterval[0, 𝑙), then the Fourier cosine and
sine transforms of derivatives are

2
a) 𝑓𝑐 ^ [𝑓′(𝑥)] = 𝑤𝐹𝑠 (𝑤) − √𝜋 𝑓(0)

b) 𝑓𝑠 ^ [𝑓′(𝑥)] = −𝑤𝐹𝑐 (𝑤)

2
c) 𝑓𝑐 ^ [𝑓′′(𝑥)] = −𝑤 2 𝐹𝑐 (𝑤) − √𝜋 𝑓′(0)

2
d) 𝑓𝑠 ^ [𝑓′′(𝑥)] = −𝑤 2 𝐹𝑠 (𝑤) + 𝑤√𝜋 𝑓(0)

 Fourier transforms of a function f (𝑥) can be derived from the complex Fourier integral
representation of 𝑓(𝑥) on the real line.
 The Fourier transform denoted by 𝐹(𝑤) 𝑜𝑟 ℱ(𝑓(𝑥))of a function 𝑓(𝑥) is defined as
1 ∞
𝐹(𝑤) = ∫ 𝑓(𝑥) 𝑒 −𝑖𝑤𝑥 𝑑𝑥
√2𝜋 −∞

 Fourier transform is linear, that is, for any functions 𝑓(𝑥) and 𝑔(𝑥)Whose Fourier
Transform exist and for any constants 𝑎, 𝑏

ℱ[𝑎𝑓(𝑥) + 𝑏𝑔(𝑥)] = 𝑎ℱ(𝑓(𝑥)) + 𝑏ℱ(𝑔(𝑥))

 Then the integral

∞
ℒ{𝑓(𝑡)} = 𝐹(𝑠) = ∫0 𝑒 −𝑠𝑡 𝑓(𝑡)𝑑𝑡
Where f is a function defined for t ≥ 0 is called the Laplace transform of f, provided that

the integral converges

 Laplace transform is a linear transform i.e. for a linear combination of functions we can
write
∞ ∞ ∞
∫0 𝑒 −𝑠𝑡 [𝛼𝑓(𝑡) + 𝛽𝑔(𝑡)]𝑑𝑡 = 𝛼 ∫0 𝑒 −𝑠𝑡 𝑓(𝑡)𝑑𝑡 + 𝛽 ∫0 𝑒 −𝑠𝑡 𝑔(𝑡)𝑑𝑡

Or ℒ{ 𝛼𝑓(𝑡) + 𝛽𝑔(𝑡)} = 𝛼ℒ{𝑓(𝑡)} + 𝛽ℒ{𝑔(𝑡)} = 𝛼𝐹(𝑠) + 𝛽𝐺(𝑠)

When ever both integrals converge for 𝑠 > 𝑐

 A function 𝑓 is said to be of exponential order 𝑐 if there exists constants 𝑐, 𝑀 >

0 𝑎𝑛𝑑 𝑇 > 0 such that |𝑓(𝑡)| ≤ 𝑀𝑒 𝑐𝑡 for all 𝑡 > 𝑇

Badri A, Moges B. and Teklebrhan B. 132 AKU

 APPLIED MATHEMATICS III

 If f is piecewise continuous on [0, ∞) and exponential order 𝑐,then ℒ{𝑓(𝑡)} exists for𝑠 >
𝑐.
 If ℒ{𝑓(𝑡)} = 𝐹(𝑠) we then say 𝑓(𝑡) is the inverse Laplace transform of 𝐹(𝑠) and write
𝑓(𝑡) = ℒ −1 {𝐹(𝑠)}.
 The inverse Laplace transform is also a linear transform, that is for constants 𝛼 and 𝛽 and
for some functions 𝐹 and 𝐺 that are transforms of 𝑓 and 𝑔 respectively, then

ℒ −1 {𝛼𝐹(𝑠) + 𝛽𝐺(𝑆)} = 𝛼ℒ −1 {𝐹(𝑠)} + 𝛽ℒ −1 {𝐺(𝑠)}

 Laplace transform is used to solve differential equations

 If 𝐹(𝑠) = ℒ{𝑓(𝑡)} and 𝑛 = 1,2,3, … ., then derivative of transforms is given as
𝑑𝑛
ℒ{𝑡 𝑛 𝑓(𝑡)} = (−1)𝑛 𝑑𝑠𝑛 ℒ{𝑓(𝑡)}

 Integration of the transform of a function 𝑓(𝑡) corresponds to the division of 𝑓(𝑡) by 𝑡

 Let 𝑓(𝑡)⁄𝑡 be a piecewise continuous, defined for t ≥ 0 and such that |𝑓(𝑡)⁄𝑡| ≤ 𝑀𝑒 −𝑘𝑡
for t ≥ 0, if ℒ{𝑓(𝑡)⁄𝑡 } = 𝐺(𝑆) for 𝑠 > 𝑘 and ℒ{𝑓(𝑡)} = 𝐹(𝑆) , then the integration of
Laplace transform is given as
∞
ℒ{𝑓(𝑡)⁄𝑡 } = ∫𝑠 𝐹(𝑢) 𝑑𝑢 and conversely
−1
ℒ −1 {𝐺(𝑠)} = ℒ −1 ℒ{𝐺 ′ (𝑠)}
𝑡
 let the functions 𝑓(𝑡) and 𝑔(𝑡) be defined for 𝑡 ≥ 0. Then the convolution of the
functions 𝑓 and 𝑔 denoted by (𝑓 ∗ 𝑔)(𝑡), and in abbreviated form by 𝑓 ∗ 𝑔 is a function
of t defined as the integral;
𝑡
(𝑓 ∗ 𝑔)(𝑡) = ∫0 𝑓(𝜏)𝑔(𝑡 − 𝜏)𝑑𝜏

 The integral equation of the form

𝑡
𝑓(𝑡) = 𝑔(𝑡) + ∫0 𝑓(𝜏)ℎ(𝑡 − 𝜏)𝑑𝜏,

Where 𝑔(𝑡) and ℎ(𝑡) are known functions is called Volterra Integral Equation for𝑓(𝑡).

 Equations that involve both the integral of an unknown function and its derivatives are
called Integro-differential equations.

Badri A, Moges B. and Teklebrhan B. 133 AKU

 APPLIED MATHEMATICS III

Miscellaneous Exercises

1. Find the Fourier cosine and Fourier sine transform of each of the stated function

𝑥, 0≤𝑥≤1
𝑠𝑖𝑛 𝑥, 0 ≤ 𝑥 ≤ 𝜋 𝑐𝑜𝑠 𝑥, 0 ≤ 𝑥 ≤ 𝜋
(a) 𝑓(𝑥) = { (b). 𝑓(𝑥) = { (c). 𝑓(𝑥) = { 2 − 𝑥, 1 ≤ 𝑥 ≤ 2
0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

1 − 𝑥2, 0 ≤ 𝑥 < 1
(d) 𝑓(𝑥) = {
0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

2. Find the Fourier sine transform of 𝑓(𝑥) = 𝑒 −𝑎𝑥 , 𝑎 > 0 and prove that
∞ 𝑥𝑠𝑖𝑛𝛼𝑥 𝜋
∫0 𝑑𝑥 = 𝑒 −𝑎𝛼 , 𝛼 > 0
𝑎 2 +𝑥 2 2

3. Find the Fourier cosine and Fourier sine transform s of each of the following

Badri A, Moges B. and Teklebrhan B. 134 AKU

 APPLIED MATHEMATICS III

(a) 𝑓(𝑥) = 𝑥 𝛼−1 , 0 < 𝛼 < 1 (b.) 𝑓(𝑥) = 𝑥𝑒 −𝑎𝑥 (c). f (𝑥) = 𝑒 −𝑥 𝑐𝑜𝑠𝑥 , 𝑥 > 0

4. Find the Fourier transform of each of the following functions

𝑖𝑎𝑥 𝑥, 0<𝑥<𝑎 𝑒 𝑥, |𝑥| < 𝑎
(a) 𝑓(𝑥) = { 𝑎 ,0 < 𝑥 < 1 (b). 𝑓(𝑥) = { (c). 𝑓(𝑥) = {
0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒 0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒
−𝑥2
𝑠𝑖𝑛𝑎𝑥 𝑥2, |𝑥| < 𝑥0
(d) 𝑓(𝑥) = 𝑒 2 (e) 𝑓(𝑥) = ,𝑎 > 0 (f) 𝑓(𝑥) = {
𝑥 0, 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

𝑥 𝑎 𝑒 −𝑥 , 𝑥>0
(g) ) 𝑓(𝑥) = {
0, 𝑥≤0

1 − 𝑥 2 , |𝑥| < 1
5. Find the Fourier transform of 𝑓(𝑥) = { and hence show that
0, |𝑥| > 1
∞ 𝑥𝑐𝑜𝑠𝑥−𝑠𝑖𝑛𝑥 𝑥 −3𝜋
∫0 𝑥3
𝑐𝑜𝑠 2 𝑑𝑥 = 16

6. In each of the following find ℒ{𝑓(𝑡)}

𝜋
(a) 𝑓(𝑡) = 𝑠𝑖𝑛2𝑡𝑐𝑜𝑠2𝑡 (b). 𝑓(𝑡) = 10𝑐𝑜𝑠(𝑡 − ) (c). 𝑓(𝑡) = (1 + 𝑒 2𝑡 )2
6

(d) 𝑓(𝑡) = (𝑒 𝑡 − 𝑒 −𝑡 )2 (e) 𝑓(𝑡) = 𝑒 𝑡 𝑠𝑖𝑛 5𝑡 (f) 𝑓(𝑡) = 𝑒 2𝑡 (𝑡 − 1)2 (g) 𝑓(𝑡) = 𝑡10 𝑒 −7𝑡
𝑡
(h) 𝑓(𝑡) = (1 − 𝑒 𝑡 + 3𝑒 −4𝑡 )𝑐𝑜𝑠 5𝑡 (i) 𝑓(𝑡) = 𝑒 3𝑡 (9 − 4𝑡 + 10 𝑠𝑖𝑛 2)

7. Evaluate each of the following

1 1 𝑠−3 6𝑠+3
(a) ℒ −1 {(𝑠2 } (b). ℒ −1 { } (c ) ℒ −1 { } (d) ℒ −1 { }
+1)(𝑠2 +4) 𝑠4 −9 (𝑠−√3)(𝑠+√3) 𝑠4 +5𝑠2 +4

1 1 2𝑠−1 (𝑠+1)2
(e) ℒ −1 { } (f) ℒ −1 { } (g) ℒ −1 { } (g) ℒ −1 { }
(𝑠+2)3 𝑠2 +2𝑠+5 𝑠2 (𝑠+1)3 (𝑠+2)4

8. Use the Laplace transforms the given initial-value and boundary problem

(a) 𝑦 ′′ + 20𝑦 ′ + 5𝑦 = 0, 𝑦(0) = 2 , 𝑦′(0) = 0

(b) 𝑦 ′′ − 𝑦 ′ = 𝑒 𝑡 𝑐𝑜𝑠 𝑡 , 𝑦(0) = 0 , 𝑦′(0) = 0

(c) 𝑦 ′′ − 2𝑦 ′ = 1 + 𝑡 , 𝑦(0) = 0 , 𝑦′(0) = 2

(d) 𝑦 ′′ − 4𝑦 ′ + 4𝑦 = 𝑡 3 𝑒 2𝑡 , 𝑦(0) = 0 , 𝑦′(0) = 0

(e) 𝑦 ′′ + 6𝑦 ′ + 𝑦 = 0 , 𝑦(0) = 4 , 𝑦 ′(0) = −3

(f) 𝑦 ′′ + 10𝑦 ′ + 2𝑦 = 0 , 𝑦(0) = 4 , 𝑦′(𝜋) = 0

9. In each of the following use the convolution theorem to find the Laplace transform

Badri A, Moges B. and Teklebrhan B. 135 AKU

 APPLIED MATHEMATICS III

𝑡 𝑡 𝑡 𝑡
(a) ℒ {∫0 𝑒 𝜏 𝑑𝜏} (b). ℒ {∫0 𝑒 −𝜏 𝑐𝑜𝑠 𝜏 𝑑𝜏} (c). ℒ {∫0 𝜏𝑠𝑖𝑛 𝜏 𝑑𝜏} (d) ℒ {∫0 𝜏 𝑒 𝑡−𝜏 𝑑𝜏}

10. In each of the following use the Laplace transform to solve the given integral equation or Integro-
differential equation.
𝑡 𝑡
(a) 𝑓(𝑡) = 𝑡𝑒 𝑡 + ∫0 𝜏𝑓(𝑡 − 𝜏)𝑑𝜏 (b) 𝑓(𝑡) + 2 ∫0 𝑓(𝜏)𝑐𝑜𝑠(𝑡 − 𝜏)𝑑𝜏 = 4𝑒 −𝑡 + 𝑠𝑖𝑛 𝑡

𝑡 𝑡
(c ) 𝑓(𝑡) + ∫0 𝑓(𝜏)𝑑𝜏 = 1 (d) ) 𝑓(𝑡) = 𝑐𝑜𝑠 𝑡 + ∫0 𝑒 −𝜏 𝑓(𝑡 − 𝜏)𝑑𝜏

𝑡
(e) 𝑡 − 2𝑓(𝑡) = ∫0 (𝑒 𝜏 − 𝑒 −𝜏 )𝑓(𝑡 − 𝜏)𝑑𝜏

𝑡 𝑡
(f) 𝑦 ′ + 4𝑦 = 4 ∫0 𝑠𝑖𝑛 𝜏 𝑦(𝑡 − 𝜏)𝑑𝜏, with 𝑦(0)=1 (g) 𝑦 ′ + 𝑦 = 4 ∫0 𝑒 −2𝜏 𝑦(𝑡 − 𝜏)𝑑𝜏, with 𝑦(0)=3

𝑡 𝑡
(h) 𝑦 ′′ − 𝑦 = ∫0 𝑠𝑖𝑛ℎ 𝜏 𝑦(𝑡 − 𝜏)𝑑𝜏, with 𝑦(0)=4 (k) 𝑦 ′′ − 4𝑦 = 2 ∫0 𝑠𝑖𝑛ℎ 2𝜏 𝑦(𝑡 − 𝜏)𝑑𝜏 , with 𝑦(0)=1

References

Antimirov, M. Ya., Applied Integral Transforms. Providence, RI: American Mathematical

Society, 1993.
ANdreid.POlyanim, Alexander V.Manzhirow,Hand book of mathematics for engineers and
scientists CRC Press,2006
Bracewell, R., The Fourier Transform and Its Applications. 3rd ed. New York: McGraw-
Hill, 2000.
Dennis G Zill, A first course in differential equations, 10th ed, Brooks/Cole, Cengage
Learning,2005
D. Zwillinger, Handbook of Differential Equations, 2nd ed., Academic Press, Boston, 1992
Erwin Kreyzing, Advanced engineering mathematics, 10th ed, wiley, 2000

Badri A, Moges B. and Teklebrhan B. 136 AKU

 APPLIED MATHEMATICS III

H.Dymand and H.p.Mikean ,Fourier Series and integral, Academic Press,Newyork,1972

I.N. Sneddon, Fourier Transforms, McGraw-Hill, New York, 1951
I.N. Sneddon, The Use of Integral Transforms, McGraw-Hill, New York,1972
J. L. Schiff, The Laplace Transform: Theory and Applications, Springer-Verlag, New York, 1999
J. L. Troutman, Boundary Value Problems of Applied Mathematics, PWS, Boston, 1994
J. P. Keener, Principle of Applied Mathematics: Transformation and Approximation,
Addison-Wesley, Reading, MA, 1988
J. D. Logan, Applied Mathematics: A Contemporary Approach, Wiley-Interscience,New
York, 1981
PPGDYke, An introduction to Laplace Transforms and Fourier Series, Springer, 1999
R.J Beerends, Fourier and Laplace Transforms, Cambridge University,2003

UNIT FIVE
VECTOR CALCULUS
Introduction

Vector calculus deals with the application of calculus operations on vectors (vector fields) .We
will often need to evaluate integrals, derivatives, and other operations that use integrals and
derivatives. The rules needed for these evaluations constitute vector calculus. In particular, line,
volume, and surface integration are important, as are directional derivatives. The relations
defined here are very useful in the context of electromagnetic but, even without reference to

Badri A, Moges B. and Teklebrhan B. 137 AKU

 APPLIED MATHEMATICS III

electromagnetics, we will show that the definitions given here are simple extensions to familiar
concepts and they simplify a number of important aspects of calculation.

We will discuss in particular the ideas of line, surface, and volume integration, and the general
ideas of gradient, divergence, and curl, as well as the divergence and Stokes theorems. These
notions are of fundamental importance for the understanding of electromagnetic fields.

More over, Vector fields have many important applications, as they can be used to represent
many physical quantities: the vector at a point may represent the strength of some force (gravity,
electricity, and magnetism) or a velocity (wind speed or the velocity of some other fluid).

Unit Objectives:

At the end of this unit each student should able to:

 Understand about scalar fields and vector fields in vector calculus;

 Learn about graphical representation of vector fields in 𝑅 2 (𝑜𝑟𝑅 3 );
 Identify and understand about Curves, Arc Length and Tangent to a Curve;
 Understand the idea of Gradient of a Scalar Field; Divergence and Curl of a Vector Field;
 Develop and learn about line integrals and Green’s Theorem;
 Understand about surface integral; Divergence theorem of Gauss; Applications;
 Realize Stock’s Theorem and its application;
 Know and learn about line integral independent of path.

5.1 Scalar Fields and Vector Fields

Overview:

In this section, we are going to deal with the definition of scalar fields and vector fields by
considering various examples.

Section Objectives:

At the end of this subtopic, students will be able to:

 Define scalar fields and vector fields;

 Represent vector fields graphically;

Badri A, Moges B. and Teklebrhan B. 138 AKU

 APPLIED MATHEMATICS III

A two-dimensional vector field is a function f that maps each point (x, y) in 𝑅 2 to a two
dimensional Vector 〈𝑢, 𝑣〉, and similarly a three-dimensional vector field maps (𝑥, 𝑦, 𝑧) to
〈𝑢, 𝑣, 𝑤〉. Since a vector has no position, we typically indicate a vector field in graphical form by
placing the vector f(x, y) with its tail at (x, y). For such a graph to be readable, the vectors must
be fairly short, which is accomplished by using a different scale for the vectors than for the axes.
Such graphs are thus useful for understanding the sizes of the vectors relative to each other but
not their absolute size.
Definition: If to each point 𝑃 of a set 𝐷 ⊆ 𝑅 3 (𝑜𝑟𝑅 2 ) is assigned a scalar𝑓(𝑃), then a scalar
field is said to be defined in 𝐷 and the function 𝑓: 𝐷 ⟶ 𝑅 is called a scalar function (or a
scalar field). Likewise, if to each point 𝑃 in 𝐷 is assigned a vector 𝑭(𝑃) ∈ 𝑅 3 (𝑜𝑟𝑅 2 ) then a
vector field is said to be defined in 𝐷 and the vector-valued function 𝑭: 𝐷 ⟶ 𝑅 3 (𝑜𝑟𝑅 2 ) is called
a vector function (or a vector field).

If we introduce Cartesian coordinates 𝑥, 𝑦, 𝑧 then instead of 𝑓(𝑃) we can write 𝑓(𝑥, 𝑦, 𝑧) and

𝑭(𝑥, 𝑦, 𝑧) = (𝑭1 (𝑥, 𝑦, 𝑧 ), 𝑭2 (𝑥, 𝑦, 𝑧 ), 𝑭3 (𝑥, 𝑦, 𝑧 )) = 𝑭1 (𝑥, 𝑦, 𝑧 )𝑖 + 𝑭2 (𝑥, 𝑦, 𝑧 )𝑗+𝑭3 (𝑥, 𝑦, 𝑧 )𝑘

where 𝑭1 , 𝑭2 , 𝑭3 are the components of 𝑭 or 𝑭 = 𝑭1 + 𝑭2 +𝑭3 .

Remarks (i) more generally, if 𝐷 is a subset of 𝑅 𝑛 , then a scalar field in 𝐷 is a function

𝑓: 𝐷 ⟶ 𝑅 and a vector field in 𝐷 is a function 𝑭: 𝐷 ⟶ 𝑅 𝑛 . in the latter case

𝑭(𝑥1 , 𝑥2 , … , 𝑥𝑛 ) = (𝑭1 (𝑥1 , 𝑥2 , … , 𝑥𝑛 ), 𝑭2 (𝑥1 , 𝑥2 , … , 𝑥𝑛 ), … , 𝑭𝑛 (𝑥1 , 𝑥2 , … , 𝑥𝑛 )), where

𝑭1 , 𝑭2 , … , 𝑭𝑛 are the components of 𝑭. If 𝑛 = 2, 𝑓 (resp. 𝑭) is called a scalar (resp. vector) field
in the plane. If 𝑛 = 3, 𝑓 (resp. 𝑭) is called a scalar (resp. vector) field in space.

(ii). A scalar field or a vector field arising from geometric or physical considerations must
depend only on the points P where it is defined and not on the particular choice of Cartesian
coordinates.

Example 5.1: The scalar function of position 𝐹(𝑥, 𝑦, 𝑧) = 𝑥𝑦𝑧 2 for (𝑥, 𝑦, 𝑧) inside the unit
sphere 𝑥 2 + 𝑦 2 + 𝑧 2 = 1 defines a scalar field throughout the unit sphere

Example 5.2:(Euclidean distance) Let 𝐷 = 𝑅 3 and 𝑓(𝑃) = ‖𝑃𝑃0 ‖ the distance of point 𝑃 from
a fixed point 𝑃0 in space. 𝑓(𝑃) defines a scalar field in space. if we introduce a Cartesian
coordinate system in which 𝑃0 : (𝑥0 , 𝑦0 , 𝑧0 ) then

𝑓(𝑃) = 𝑓(𝑥, 𝑦, 𝑧) = ‖𝑥 − 𝑥0 , 𝑦 − 𝑦0 , 𝑧 − 𝑧0 ‖ = √(𝑥 − 𝑥0 )2 + (𝑦 − 𝑦0 )2 + (𝑧 − 𝑧0 )2

Note that the value of 𝑓(𝑃) does not depend on the particular choice of Cartesian coordinate
system.

The best way to picture a vector field is to draw the arrow representing the vector F(x, y) starting
at the point(𝑥, 𝑦)of course, it’s impossible to do this for all points (𝑥, 𝑦) ,but we can gain a

Badri A, Moges B. and Teklebrhan B. 139 AKU

 APPLIED MATHEMATICS III

reasonable impression of 𝑭 by doing it for a few representative points in 𝐷 as shown in the figure
below. Since F(x, y) is a two-dimensional vector, we can write it in terms of its component
functions.

Fig.5.1 Vector field in 𝑅 2 Fig.5.2. Vector field in 𝑅 3

Example 5.3: A vector field on 𝑅 2 is defined by 𝑭(𝑥, 𝑦) = −𝑦𝑖 + 𝑥𝑗. Describe 𝑭 by sketching
some of the vectors 𝑭(𝑥, 𝑦).

Solution: Since(1, 0) = 𝑗, we draw the vector 𝑗 = 〈0,1〉 starting at the point (1, 0).since
𝑭(0, 1) = −𝑖, we draw the vector 〈−1,0〉 with starting point (0, 1).continuing in this way, we
calculate several other representative values of 𝑭(𝑥, 𝑦) in the table and draw the corresponding
vectors.

(𝑥, 𝑦) 𝑭(𝑥, 𝑦) (𝑥, 𝑦) 𝑭(𝑥, 𝑦)

(1, 0) 〈0,1〉 (−1, 0) 〈0, −1〉
(2,2) 〈−2,2〉 (−2, −2) 〈2, −2〉
(3,0) 〈0,3〉 (−3,0) 〈0, −3〉
(0,1) 〈−1,0〉 (0, −1) 〈1,0〉
(−2,2) 〈−2, −2〉 (2, −2) 〈2,2〉
(0,3) 〈−3,0〉 (0, −3) 〈3,0〉

Fig.5.3: 𝑭(𝑥, 𝑦) = −𝑦𝑖 + 𝑥𝑗

It appears from Figure 5 that each arrow is tangent to a circle with center the origin. To confirm this, we
take the dot product of the position vector 𝒙 = 𝑥𝑖 + 𝑦𝑗 with the vector with the vector𝑭(𝒙) = 𝑭(𝑥, 𝑦):

Badri A, Moges B. and Teklebrhan B. 140 AKU

 APPLIED MATHEMATICS III

𝒙. 𝑭(𝒙) = (𝑥𝑖 + 𝑦𝑗).(−𝑦𝑖 + 𝑥𝑗) = −𝑥𝑦 + 𝑦𝑥 = 0

This shows that 𝑭(𝑥, 𝑦) is perpendicular to the position vector 〈𝑥, 𝑦〉 and is therefore tangent to
the circle with center the origin and radius |𝒙| = √𝑥 2 + 𝑦 2 . Notice also that

|𝑭(𝑥, 𝑦)| = √(−𝑦)2 + 𝑥 2 = |𝒙|

So the magnitude of the vector 𝑭(𝑥, 𝑦) is equal to the radius of the circle

Some computer algebra systems are capable of plotting vector fields in two or three dimensions.
They give a better impression of the vector field than is possible by hand because the computer
can plot a large number of representative vectors. Figure 5.4 shows a computer plot of the vector
field in Example 1; Figures 5.5 and 5.6 show two other vector fields. Notice that the computer
scales the lengths of the vectors so they are not too long and yet are proportional to their true
lengths.

Fig.5.4 𝑭(𝑥, 𝑦) = 〈−𝑦, 𝑥〉 Fig.5.5 𝑭(𝑥, 𝑦) = 〈𝑦, 𝑠𝑖𝑛𝑥〉 Fig.5.6 𝑭(𝑥, 𝑦) =

〈ln(1 + 𝑦 2 ) , ln(1 + 𝑥 2 )〉

Example 5.4: Sketch the vector field on 𝑅 3 given by 𝑭(𝑥, 𝑦) = 𝑧𝒌

Solution The sketch is shown in the figure below. Notice that all vectors are vertical and upward
above the 𝑥𝑦-plane or downward below it. The magnitude increases with the distance from the
𝑥𝑦-plane

Badri A, Moges B. and Teklebrhan B. 141 AKU

 APPLIED MATHEMATICS III

Fig 5.7 𝑭(𝑥, 𝑦) = 𝑧𝒌

We were able to draw the vector field in Example 2 by hand because of its particularly simple
formula. Most three-dimensional vector fields, however, are virtually impossible to sketch by
hand and so we need to resort to a computer algebra system. Examples are shown in Figures
Fig.5.8 and Fig5.9.If the vector field in Figure Fig5.9 represents a velocity field, then a particle
would be swept upward and would spiral around the -axis in the clockwise direction as viewed
from above.

𝑦 𝑥 𝑧
Fig.5.8 𝑭(𝑥, 𝑦, 𝑧) = 𝑦𝑖 + 𝑧𝑗 + 𝑥𝑘 Fig5.9: 𝑭(𝑥, 𝑦, 𝑧) = 𝑧 𝑖 + 𝑧 𝑗 + 4 𝑘

Example 5.5: Newton’s Law of Gravitation states that the magnitude of the gravitational force
between two objects with masses 𝑚 and 𝑀 is
𝑚𝑀𝐺
|𝑭| =
𝑟2

Where r is the distance between the objects and 𝐺 is the gravitational constant.(this is an example
of an inverse square law .) let us assume that the object with mass 𝑀 is located at the origin in
𝑅 3 .(For instance, 𝑀 could be the mass of the earth and the origin would be at its center.)Let the
position vector of the object with mass 𝑚 be 𝒙 = 〈𝑥, 𝑦, 𝑧〉. Then 𝑟 = |𝒙|, so 𝑟 2 = |𝒙|2. The
gravitational force exerted on this second object acts toward the origin, and the unit vector in this
direction is
𝒙
− |𝒙|

Therefore the gravitational force acting on the object at 𝒙 = 〈𝑥, 𝑦, 𝑧〉 is

𝑚𝑀𝐺
𝑭(𝒙) = − |𝒙|3
𝒙 ∗

[Physicists often use the notation 𝒓 instead of 𝒙 for the position vector, so you may see formula
(∗) written in the form 𝑭 = − (𝑚𝑀𝐺 ⁄𝒓3 )𝒓.] The function given by equation ∗ is an example of

Badri A, Moges B. and Teklebrhan B. 142 AKU

 APPLIED MATHEMATICS III

a vector field, called the gravitational field, because it associates a vector [the force 𝑭(𝒙)] with
every point 𝒙 in the space.

Formula ∗ is a compact way of writing the gravitational field, but we can also write it in terms
of its component functions by using the fact that 𝒙 = 𝒙𝒊 + 𝒚𝒋 + 𝒛𝒌 and

|𝒙| = √𝑥 2 + 𝑦 2 + 𝑧 2 :
−𝑚𝑀𝐺𝑥 −𝑚𝑀𝐺𝑦 −𝑚𝑀𝐺𝑧
𝑭(𝑥, 𝑦, 𝑧) = 3 𝑖+ 3 𝑗+ 3 𝑘
(𝑥 2 +𝑦 2 +𝑧 2 ) ⁄2 (𝑥 2 +𝑦 2 +𝑧 2 ) ⁄2 (𝑥 2 +𝑦 2 +𝑧 2 ) ⁄2

Fig5.10 gravitational force field

Exercise 5.1

1. In each of the following sketch the vector field 𝑭

1
a. 𝑭(𝑥, 𝑦) = 𝑦𝑖 + 2 𝑗 b. 𝑭(𝑥, 𝑦) = 𝑖 + 𝑥𝑗 c. 𝑭(𝑥, 𝑦) = (𝑥 − 𝑦)𝑖 + 𝑥𝑗

5.2 VECTOR CULCULAS

Overview:
In this section, we are going to deal with the calculus of vector fields by considering various
examples

Section Objectives:

At the end of this subtopic, students will be able to:

 Define and identify the calculus of vector fields(vector valued function);

 Understand and know about the limits derivative and integration of vector fields.

Badri A, Moges B. and Teklebrhan B. 143 AKU

 APPLIED MATHEMATICS III

In general, a function is a rule that assigns to each element in the domain an element in the range.
A vector-valued function, or vector function, is simply a function whose domain is a set of real
numbers and whose range is a set of vectors. We are most interested in vector functions 𝒓 whose
values are three-dimensional vectors. This means that for every number 𝑡 in the domain of 𝒓
there is a unique vector𝑉3 in denoted by 𝒓(𝒕).If 𝑓(𝑡), 𝑔(𝑡)and ℎ(𝑡) are the components of the
vector 𝒓(𝒕).then 𝑓, 𝑔and ℎ real-valued functions called the component functions of 𝒓 and we
can write

𝒓(𝒕) = 〈𝑓(𝑡), 𝑔(𝑡), ℎ(𝑡)〉 = 𝑓(𝑡)𝐢 + 𝑔(𝑡)𝒋 + ℎ(𝑡)𝐤

The limit of a vector function 𝒓 is defined by taking the limits of its component functions as
follows.
.
Definition: if 𝒓(𝒕) = 〈𝑓(𝑡), 𝑔(𝑡), ℎ(𝑡)〉, then
lim 𝒓(𝒕) = 〈lim 𝑓(𝑡), lim 𝑔(𝑡) , lim ℎ(𝑡)〉
𝑡→𝑎 𝑡→𝑎 𝑡→𝑎 𝑡→𝑎

Provided the limits of the component function exist

𝑠𝑖𝑛 𝑡
Example 5.6: findlim 𝒓(𝒕), where 𝒓(𝒕) = (1 + 𝑡 3 )𝒊 + 𝑡𝑒 −𝑡 𝒋 + 𝒌
𝑡→0 𝑡

Solution According to the above definition, the limit of 𝒓 is the vector whose components are
the limits of the component functions of 𝒓:

𝑠𝑖𝑛 𝑡
lim 𝒓(𝒕) = [ lim(1 + 𝑡 3 )]𝒊 + [lim 𝑡𝑒 −𝑡 ]𝒋 + [lim ]𝒌 = 𝒊 + 𝒌
𝑡→0 𝑡→0 𝑡→0 𝑡→0 𝑡

Note limits of vector functions obey the same rules as limits of real-valued functions

A vector function 𝒓 is continuous at 𝑎 if

lim 𝒓(𝒕) = 𝒓(𝑎)

𝑡→𝑎

In view of the above definition of limit, we see that 𝒓 is continuous at 𝑎 if and only if its
component function 𝑓, 𝑔 and ℎ are continuous at 𝑎.

Definition: The derivative 𝒓′ of a vector function 𝒓 is defined in much the same way as for real –
valuedBadri
functions:
A, Moges B. and Teklebrhan B. 144 AKU
𝑑𝒓 𝒓(𝑡+ℎ)+𝒓(𝑡)
𝑑𝑡
= 𝒓′(t) = lim
ℎ→0 ℎ
 APPLIED MATHEMATICS III

The following theorem gives us a convenient method for computing the derivative of a vector
function 𝒓: just differentiate each component of 𝒓

Theorem: 𝒓(𝒕) = 〈𝑓(𝑡), 𝑔(𝑡), ℎ(𝑡)〉 = 𝑓(𝑡)𝐢 + 𝑔(𝑡)𝒋 + ℎ(𝑡)𝐤, where 𝑓, 𝑔 and ℎ are
differentiable functions, then

𝒓′(t) = 〈𝑓′(𝑡), 𝑔′(𝑡), ℎ′(𝑡)〉 = 𝑓′(𝑡)𝐢 + 𝑔′(𝑡)𝒋 + ℎ′(𝑡)𝐤

Example 5.7: Find the derivative of 𝒓(𝑡) = (1 + 𝑡 3 )𝒊 + 𝑡𝑒 −𝑡 𝒋 + 𝑠𝑖𝑛 2𝑡𝒌.

Solution
According o the above theorem, we differentiate each component of 𝒓:

𝒓′(t) = 3𝑡 2 𝒊 + (1 − 𝑡)𝑒 −𝑡 𝒋 + 2𝑐𝑜𝑠 2𝑡𝒌

Differentiation Rules
The next theorem shows the differentiation formulas for real-valued functions have their
counterparts for vector-valued functions.
Theorem: suppose 𝒖 and 𝒗 are differentiable vector functions, 𝑐 is a scalar, and 𝑓 is a real-
valued function, then
𝒅 𝒅 𝒅
1. [𝒖(𝑡) + 𝒗(𝑡)] = [𝒖(𝑡)] + [𝒗(𝑡)]
𝒅𝒕 𝒅𝒕 𝒅𝒕
𝒅 𝒅
2. [𝑐𝒖(𝑡)] = 𝑐 [𝒖(𝑡)]
𝒅𝒕 𝒅𝒕
𝒅
3. [𝑓(𝑡)𝒖(𝑡)] = 𝑓 ′ (𝑡)𝒖(𝑡) + 𝑓(𝑡)𝒖′(𝑡)
𝒅𝒕
𝒅
4. [𝒖(𝑡). 𝒗(𝑡)] = 𝒖′ (𝑡). 𝒗(𝑡) + 𝒖(𝑡). 𝒗′(𝑡)
𝒅𝒕
𝒅
5. [𝒖(𝑡) × 𝒗(𝑡)] = 𝒖′ (𝑡) × 𝒗(𝑡) + 𝒖(𝑡) × 𝒗′(𝑡)
𝒅𝒕
𝒅
6. [𝒖(𝑓(𝑡)] = 𝑓′(𝑡) 𝒖′(𝑓(𝑡)
𝒅𝒕

Badri A, Moges B. and Teklebrhan B. 145 AKU

 APPLIED MATHEMATICS III

Proof (Exercise)

Integrals
The definite integral of a continuous vector function 𝒓(𝑡) can be defined in much the same way
as for real-valued functions except that the integral is a vector. But we can express the integral
of 𝒓 in terms of the integrals of its component functions 𝑓, 𝑔 and ℎ as follows.

𝑏 𝑛

∫ 𝒓(𝑡)𝑑𝑡 = lim ∑ 𝒓(𝑡 ∗ 𝑖 ) ∆𝑡

𝑎 𝑛→∞
𝑖=1
𝑛 𝑛 𝑛

= 𝑙𝑖𝑚 [(∑ 𝑓(𝑡∗ 𝑖 )∆𝑡) 𝒊 + (∑ 𝑔(𝑡∗ 𝑖 )∆𝑡) 𝒋 + (∑ ℎ(𝑡∗ 𝑖 )∆𝑡) 𝒌]

𝑛→∞
𝑖=1 𝑖=1 𝑖=1

And so

𝑏 𝑏 𝑏 𝑏
∫ 𝒓(𝑡)𝑑𝑡 = (∫ 𝑓(𝑡)𝑑𝑡) 𝒊 + (∫ 𝑔(𝑡)𝑑𝑡) 𝒋 + (∫ ℎ(𝑡)𝑑𝑡) 𝒌
𝑎 𝑎 𝑎 𝑎

This means that we can evaluate integral of a vector function by integrating each component
function.

Note: 1. We can extend the fundamental theorem of calculus to continuous vector functions as
follows:
𝑏
𝑏
∫ 𝒓(𝑡)𝑑𝑡 = 𝑹(𝑡)] = 𝑹(𝑏) − 𝑹(𝑎)
𝑎 𝑎
Where 𝑹 is an antiderivative of 𝒓, that is
𝑹′ (𝑡) = 𝒓(𝑡)
𝟐.We use the notation ∫ 𝒓(𝑡)𝑑𝑡 for indefinite integrals (antiderivatives)
Example 5.8: If 𝒓(𝑡) = 2 cos 𝑡 𝒊 + sin 𝑡 𝒋 + 2𝑡 𝒌, then
𝑏 𝑏
∫ 𝒓(𝑡)𝑑𝑡 = (2 cos 𝑡)𝒊 + (∫ sin 𝑡 𝑑𝑡) 𝒋 + (∫ 2𝑡𝑑𝑡 ) 𝒌
𝑎 𝑎

= 2𝑠𝑖𝑛 𝑡 𝒊 − 𝑐𝑜𝑠 𝑡 𝒋 + 𝑡2 𝒌 + 𝑪

Badri A, Moges B. and Teklebrhan B. 146 AKU

 APPLIED MATHEMATICS III

where 𝑪 is a vector constant integration, and

𝜋 𝜋
2
2 𝜋22
∫ 𝒓(𝑡)𝑑𝑡 = [2𝑠𝑖𝑛 𝑡 𝒊 − 𝑐𝑜𝑠 𝑡 𝒋 + 𝑡 𝒌] = 2𝒊 + 𝒋 + 𝒌
0 0 4

Exercises

1.Evaluate each of the following limits

lim 〈cos t, sin t, tlnt〉
(a)
t → 0+
lnt
(b) lim + 〈arctan t , e−2t , 〉
t→0 t
2.Find the derivative of each of the following vector functions
(a) 𝐫(t) = 〈t 2 , 1 − t, √t〉
(b) 𝐫(t) = 〈cos 3t , t, sin 3t〉
3. Evaluate the following integrals
1
(a) ∫ (16t 3 𝐢 − 9t 2 𝐣 + 25t 4 𝐤)dt
0
4
1
(b) ∫ (√t𝐢 + te−t 𝐣 + 𝐤)dt
1 t2

5.3 Curves, Arc length and Tangent

Overview:

In this section, we are going to introduce Curves, Arc length and Tangent by considering various
examples.

Section Objectives:

At the end of this subtopic, students will be able to:

 Define curve and find the length of curve segment(arc length)

 Identify the tangent of a curve;

Curves
Vector calculus has important applications to curves and surfaces in physics and geometry. The
application of vector calculus to geometry is a field known as differential geometry.

Badri A, Moges B. and Teklebrhan B. 147 AKU

 APPLIED MATHEMATICS III

Differential geometric methods are applied to problems in mechanics, computer-aided as well as

traditional engineering design, geodesy, geography, space travel, and relativity theory.

Bodies that move in space form paths that may be represented by curves C. This and other
applications show the need for parametric representations of C with parameter t, which may
denote time or something else .A typical parametric representation is given by.

𝒓(𝑡) = [𝑥(𝑡), 𝑦(𝑡), 𝑧(𝑡)] = 𝑥(𝑡)𝒊 + 𝑦(𝑡)𝒋 + 𝑧(𝑡)𝒌

Fig 5.11 Parametric representation of a curve

Here 𝑡 is a parameter and 𝑥, 𝑦, 𝑧 are Cartesian coordinates that is the usual rectangular
coordinates. To each value 𝑡 = 𝑡0 there corresponds a point of 𝐶 with position vector 𝒓(𝑡0 )
whose coordinates are 𝑥(𝑡0 ), 𝑦(𝑡0 ), 𝑧(𝑡0 ).

Definition: A curve in 𝑅 2 (or𝑅 3 ) is a differentiable function 𝒓: [𝑎, 𝑏] ⟶ 𝑅 2 (or𝑅 3 ). The initial

point is 𝒓(𝒂) and the final point is 𝒓(𝑏).the domain of the curve is the interval [𝑎, 𝑏]. A portion
of 𝒓 defined on an interval [𝑐, 𝑑] ⊆ [𝑎, 𝑏] is called a curve segment.

Example 5.9: The line is the simplest curve in the plane as its coordinate functions are linear
.Explicitly the curve

𝒓(𝑡) = 𝒑 + 𝑡𝒗 = (𝑥0 + 𝑡𝑢, 𝑦0 + 𝑡𝑣), where 𝒗 ≠ 0

Is a straight line through the reference point 𝒑 = 𝒓(0) = (𝑥0 , 𝑦0 ) in the direction 𝒗 = (𝑢, 𝑣)

Here, 𝑡 is the signed distance from point 𝒓(𝑡) on the line to 𝒑 as scaled by‖𝒗‖.

Badri A, Moges B. and Teklebrhan B. 148 AKU

 APPLIED MATHEMATICS III

Fig 5.12: the straight line curve

As shown on the above figure, the vector 𝒑 to a point (𝑥, 𝑦) on the line must be either in the
direction of (𝑢, 𝑣) or in the opposite direction. Hence, the cross product of the two vectors must
be zero, that is,

(𝑥 − 𝑥0 , 𝑦 − 𝑦0 ) × (𝑢, 𝑣) = 0

Expansion of the above cross product yields an implicit equation of the line that relates the 𝑥 and
𝑦 coordinates of every incident point:

𝑣𝑥 − 𝑢𝑦 − 𝑣𝑥0 + 𝑢𝑦0 = 0

Example 5.10: sketch and identify the curve defined by the parametric equations

𝑥 = 𝑡 2 − 2𝑡 𝑦 =𝑡+1

Solution: Here 𝒓(𝑡) = [𝑥(𝑡), 𝑦(𝑡)] = [𝑡 2 − 2𝑡, 𝑡 + 1 ].each value of t a point on the curve, as
shown in the table. For instance, if 𝑡 = 0, then 𝑥 = 0 , 𝑦 = 1 and so the corresponding point is
(0,1).in Fig5.13 we plot the points (𝑥, 𝑦) Determined by several values of the parameter and we
join them to produce a curve

𝒕 𝒙 𝒚
-2 8 -1
-1 3 0
0 0 1
1 -1 2
2 0 3
3 3 4
4 8 5
Fig5.14 The graph of the curve 𝑥 = 𝑡 2 − 2𝑡 𝑦 =𝑡+1
Fig5.13 Tabular values of the curve 𝑥 = 𝑡 2 − 2𝑡 𝑦 = 𝑡 + 1

Badri A, Moges B. and Teklebrhan B. 149 AKU

 APPLIED MATHEMATICS III

A particle whose position is given by the parametric equations moves along the curve in the
direction of the arrows as 𝑡 increases. Notice that the consecutive points marked on the curve
appear at equal time intervals but not at equal distances. That is because the particle slows down
and then speeds up as increases.
It appears from Fig5.14 that the curve traced out by the particle may be a parabola. This can be
confirmed by eliminating the parameter 𝑡 as follows. We obtain 𝑡 = 𝑦 − 1 from the second
equation and substitute into the first equation. This gives

𝑥 = 𝑡 2 − 2𝑡 = (𝑦 − 1)2 − 2(𝑦 − 1) = 𝑦 2 − 4𝑦 + 3
And so the curve represented by the given parametric equation is the parabola

𝑥 = 𝑦 2 − 4𝑦 + 3.

Note: in e the example above no restriction was placed on the parameter 𝑡, so we assumed that
𝑡 could be any real number.but sometimes we restrict 𝑡 to lie in finite interval.For instance,the
parametric curve
𝑥 = 𝑡 2 − 2𝑡 𝑦 =𝑡+1 0≤𝑡 ≤4

Shown in Fig 5.15 is the part of the parabola in the above example that starts at the point (0,1)
and ends at the point (8,5). The arrowhead indicates the direction in which the curve is traced
as 𝑡 increases from 0 to 4.

Fig 5.15 𝑥 = 𝑡 2 − 2𝑡 𝑦 =𝑡+1 0≤𝑡 ≤4

In general, the curve with parametric equations 𝑥 = 𝑓(𝑡) 𝑦 = 𝑔(𝑡) 𝑎≤𝑡≤𝑏

Has initial point (𝑓(𝑎), 𝑔(𝑎)) and terminal point (𝑓(𝑏), 𝑔(𝑏))

Example 5.11: what curve is represented by the following parametric equations?

𝑥 = 𝑐𝑜𝑠𝑡 𝑦 = 𝑠𝑖𝑛𝑡 0 ≤ 𝑡 ≤ 2𝜋

Badri A, Moges B. and Teklebrhan B. 150 AKU

 APPLIED MATHEMATICS III

Solution: if we plot points, it appears the curve is a circle. We can confirm this impression by
eliminating𝑡. Observe that

𝑥 2 + 𝑦 2 = 𝑐𝑜𝑠 2 𝑡 + 𝑠𝑖𝑛2 𝑡 = 1

Thus the point (𝑥, 𝑦) moves on the unit circle 𝑥 2 + 𝑦 2 = 1.notice that in this example the
parameter 𝑡 can be interpreted as the angle (in radians) shown in figure.as 𝑡 increases from 0 to
2𝜋, the point (𝑥, 𝑦) = (𝑐𝑜𝑠𝑡, 𝑠𝑖𝑛𝑡) moves once around the circle in the counter clockwise
direction starting from the point (1,0).

Fig.5.16 The curve 𝑥 = 𝑐𝑜𝑠𝑡 𝑦 = 𝑠𝑖𝑛𝑡 0 ≤ 𝑡 ≤ 2𝜋

Example 5.12:The vector function

𝒓(𝑡) = [𝑎𝑐𝑜𝑠𝑡, 𝑏𝑠𝑖𝑛𝑡, 0] = 𝑎𝑐𝑜𝑠𝑡𝑖 + 𝑏𝑠𝑖𝑛𝑡𝑗

Represents an ellipse in the 𝑥𝑦-plane with center at the origin and principal axes in the direction
of the 𝑥- and 𝑦- axis. In fact, since 𝑐𝑜𝑠 2 𝑡 + 𝑠𝑖𝑛2 𝑡 = 1, we obtain from the above equation

𝑥2 𝑦2
+ = 1, 𝑧=0
𝑎2 𝑏 2
If 𝑏 = 𝑎,then

𝑟(𝑡) = [𝑎𝑐𝑜𝑠𝑡, 𝑏𝑠𝑖𝑛𝑡, 0] = 𝑎𝑐𝑜𝑠𝑡𝑖 + 𝑏𝑠𝑖𝑛𝑡𝑗 = [𝑎𝑐𝑜𝑠𝑡, 𝑎𝑠𝑖𝑛𝑡, 0] = 𝑎𝑐𝑜𝑠𝑡𝑖 + 𝑎𝑠𝑖𝑛𝑡𝑗

represents a circle of radius a

Badri A, Moges B. and Teklebrhan B. 151 AKU

 APPLIED MATHEMATICS III

Fig. 5.17 circle of the above example Fig5.18 ellipse of the above example

Example 5.13: the twisted curve c represented by the vector function

𝑟(𝑡) = [𝑎𝑐𝑜𝑠𝑡, 𝑎𝑠𝑖𝑛𝑡, 𝑐𝑡] = 𝑎𝑐𝑜𝑠𝑡𝒊 + 𝑏𝑠𝑖𝑛𝑡𝒋 + 𝑐𝑡𝒌 (𝑐 ≠ 0)

Is called a circular helix, it lies on the cylinder 𝑥 2 + 𝑦 2 = 𝑎2 . If 𝑐 > 0,the helix is shaped like
right-handed screw(fig.). if 𝑐 < 0, it looks like a left handed screw(fig.).if 𝑐 = 0,then it is a circle

Fig 5.19 right-handed circular Helix Fig 5.20 Left-handed circular helix

A simple curve is a curve without multiple points, that is, without points at which the curve
intersects or touches itself. Circle and helix are simple curves. Fig 5.20 shows curves that are not
simple. An example is[𝑠𝑖𝑛2𝑡, 𝑐𝑜𝑠2𝑡, 0].Can you sketch it?
An arc of a curve is the portion between any two points of the curve. For simplicity, we say
“curve” for curves as well as for arcs.

Fig 5.20 Curves with multiple points

Badri A, Moges B. and Teklebrhan B. 152 AKU

 APPLIED MATHEMATICS III

Arc length
Recall from the application of integration that the length 𝐿 of a curve 𝐶 given in the form 𝑦 =
𝐹(𝑥), 𝑎 ≤ 𝑥 ≤ 𝑏, 𝐹 being continuous is given by

𝑏 𝑑𝑦 2
𝐿 = ∫𝑎 √1 + (𝑑𝑥 ) 𝑑𝑥

Suppose that 𝐶 can also be described by the parametric equation 𝑥 = 𝑓(𝑡) and 𝑦 = 𝑔(𝑡),

𝛼 ≤ 𝑡 ≤ 𝛽, where 𝑑𝑥 ⁄𝑑𝑡 = 𝑓 ′ (𝑡) > 0.this means that 𝐶 is traversed once, from the left to right,
𝑑𝑦 𝑑𝑦⁄𝑑𝑡
as 𝑡 increases from 𝛼 𝑡𝑜 𝛽, and 𝑓(𝛼) = 𝑎, 𝑓(𝛽) = 𝑏. Putting 𝑑𝑥 = 𝑑𝑥⁄𝑑𝑡 in to the above formula
and using the substitution Rule , we obtain

𝑏 𝑑𝑦 2 𝛽 𝑑𝑦⁄𝑑𝑡 2 𝑑𝑥
𝐿 = ∫𝑎 √1 + (𝑑𝑥 ) 𝑑𝑥 = ∫𝛼 √1 + (𝑑𝑥⁄𝑑𝑡 ) 𝑑𝑡
𝑑𝑡

Since 𝑑𝑥⁄𝑑𝑡 = 𝑓 ′ (𝑡) > 0, we have

𝛽 𝑑𝑥 𝑑𝑦 2 2
𝐿 = ∫𝛼 √( ) + ( ) 𝑑𝑡
𝑑𝑡 𝑑𝑡

If the curve is in the space, that is if 𝒓(𝑡) = (𝑥(𝑡), 𝑦(𝑡), ℎ(𝑡)) where 𝑥 = 𝑓(𝑡), 𝑦 = 𝑔(𝑡) and
𝑧 = ℎ(𝑡), then the arc length is

𝑏 𝑑𝑥 2 𝑑𝑦 2
𝑑𝑍 2
𝐿 = ∫𝑎 √( 𝑑𝑡 ) + ( 𝑑𝑡 ) + ( 𝑑𝑡 ) 𝑑𝑡

Thus, using Leibniz notation, we have the following result, which has the same form as the last
two formulas.

Theorem: If a curve C is described by the parametric equation 𝑥 = 𝑓(𝑡), 𝑦 = 𝑔(𝑡) and z =

ℎ(𝑡), 𝑎 ≤ 𝑡 ≤ 𝑏, where 𝑓 ′ , 𝑔′ and ℎ′ are continuous on [𝑎, 𝑏] and C is traversed exactly once as
𝑡 increases from 𝑎 to 𝑏, then the length 𝐿 of C is.

𝑏 𝑑𝑥 2 𝑑𝑦 2 𝑏 𝑑𝑥 𝑑𝑦 𝑑𝑧2 2 2
𝐿 = ∫𝑎 √( 𝑑𝑡 ) + ( 𝑑𝑡 ) 𝑑𝑡 or 𝐿 = ∫𝑎 √( 𝑑𝑡 ) + ( 𝑑𝑡 ) + ( 𝑑𝑡 ) 𝑑𝑡

Example 5.13: Find the arc length of the curve traced out by the end points of the vector
function

𝒓(𝑡) = (2𝑡, 𝑙𝑛𝑡, 𝑡 2 ) for 1 ≤ 𝑡 ≤ 𝑒

Badri A, Moges B. and Teklebrhan B. 153 AKU

 APPLIED MATHEMATICS III

Solution here 𝑥 = 𝑓(𝑡) = 2𝑡, 𝑦 = 𝑔(𝑡) = 𝑙𝑛𝑡 and z = ℎ(𝑡) = 𝑡 2 then by the above theorem

𝑏 𝑑𝑥 2
𝑑𝑦 2𝑑𝑧 2 𝑑𝑥 𝑑𝑦 1 𝑑𝑧
𝐿 = ∫𝑎 √( 𝑑𝑡 ) + ( 𝑑𝑡 ) + (𝑑𝑡 ) 𝑑𝑡 where = 2, = and 𝑑𝑡 = 2𝑡
𝑑𝑡 𝑑𝑡 𝑡

𝑒 1 2 𝑒 1 𝑒 4𝑡 4 +4𝑡 2 +1
⟹ 𝐿 = ∫1 √(2)2 + ( 𝑡 ) + (2𝑡)2 𝑑𝑡 = ∫1 √4 + 𝑡 2 + 4𝑡 2 𝑑𝑡 = ∫1 √ 𝑑𝑡
𝑡2

𝑒 2𝑡 +1 2 2
𝑒 2𝑡 +1 𝑒 2 1 𝑒 𝑒1
= ∫1 √( 𝑡 ) 𝑑𝑡 = ∫1 𝑡 𝑑𝑡 = ∫1 (2𝑡 + 𝑡 )𝑑𝑡 = ∫1 2𝑡 𝑑𝑡 + ∫1 𝑡 𝑑𝑡 = [𝑡 2 ] 𝑒1 + [𝑙𝑛𝑡] 𝑒1

⟹ 𝐿 = (𝑒 2 − 1) + 𝑙𝑛𝑒 − 𝑙𝑛1 = 𝑒 2 − 1 + 1 − 0 = 𝑒 2

Hence 𝐿 = 𝑒 2

Example 5.13: find the length of one arch of the cycloid 𝑥 = 𝑟(𝜃 − 𝑠𝑖𝑛𝜃), 𝑦 = 𝑟(1 − 𝑐𝑜𝑠𝜃).
Solution: one arch of cycloid is described by the parameter interval, 0 ≤ 𝜃 ≤ 2𝜋.
𝑑𝑥 𝑑𝑦
Since 𝑑𝜃 = 𝑟(1 − 𝑐𝑜𝑠𝜃) and = 𝑟𝑠𝑖𝑛𝜃. We have
𝑑𝜃

2𝜋 𝑑𝑥 𝑑𝑦2 2
2𝜋
𝐿 = ∫0 √(𝑑𝜃) + (𝑑𝜃) 𝑑𝜃 = ∫0 √(𝑟(1 − 𝑐𝑜𝑠𝜃))2 + (𝑟𝑠𝑖𝑛𝜃)2 𝑑𝜃

2𝜋 2𝜋
= ∫0 √𝑟 2 (1 − 𝑐𝑜𝑠𝜃)2 + 𝑟 2 𝑠𝑖𝑛2 𝜃 𝑑𝜃 = ∫0 √𝑟 2 (1 − 2𝑐𝑜𝑠𝜃 + 𝑐𝑜𝑠 2 𝜃) + 𝑟 2 𝑠𝑖𝑛2 𝜃 𝑑𝜃

2𝜋 2𝜋
= ∫0 √𝑟 2 (1 − 2𝑐𝑜𝑠𝜃 + 𝑐𝑜𝑠 2 𝜃 + 𝑠𝑖𝑛2 𝜃) 𝑑𝜃 = ∫0 √𝑟 2 (2 − 2𝑐𝑜𝑠𝜃) 𝑑𝜃

2𝜋
= ∫0 √2(1 − 𝑐𝑜𝑠𝜃) 𝑑𝜃
1
To evaluate this integral we use the identity 𝑠𝑖𝑛2 𝑥 = (1 − 𝑐𝑜𝑠2𝑥) with 𝜃 = 2𝑥, which
2
gives 1 − 𝑐𝑜𝑠𝜃 = 2𝑠𝑖𝑛2 (𝜃⁄2)𝜃. Since 0 ≤ 𝜃 ≤ 2𝜋.we have 0 ≤ 𝜃⁄2 ≤ 2𝜋 and so sin(𝜃⁄2) ≥
0.therefore

√2(1 − 𝑐𝑜𝑠𝜃) = √4𝑠𝑖𝑛2 (𝜃⁄2) = 2|sin(𝜃⁄2)| = 2𝑠𝑖𝑛(𝜃⁄2)

2𝜋 2𝜋 2𝜋
And so 𝐿 = 2𝑟 ∫0 sin(𝜃⁄2) 𝑑𝜃 = 2𝑟[−2𝑐𝑜𝑠(𝜃⁄2)] = 2𝑟[−2 cos(𝜃⁄2)] = 2𝑟[2 + 2]
0 0
= 8𝑟

Badri A, Moges B. and Teklebrhan B. 154 AKU

 APPLIED MATHEMATICS III

Tangents

In the preceding section we saw that some curves defined by parametric equation 𝑥 = 𝑓(𝑡) and
𝑦 = 𝑔(𝑡) can also be expressed, by eliminating the parameter, in the form 𝑦 = 𝐹(𝑥).that is, if 𝑓′
is continuous and 𝑓 ′ (𝑡) ≠ 0 for 𝑎 ≤ 𝑡 ≤ 𝑏, then the parametric curve 𝑥 = 𝑓(𝑡), 𝑦 = 𝑔(𝑡),
𝑎 ≤ 𝑡 ≤ 𝑏, can be put in the form 𝑦 = 𝐹(𝑥). If we substitute 𝑥 = 𝑓(𝑡) and 𝑦 = 𝑔(𝑡) in the
equation 𝑦 = 𝐹(𝑥), we get

𝑔(𝑡) = 𝐹(𝑓(𝑡))

And so, if 𝑔, 𝐹, and 𝑓 are differentiable, the Chain Rule gives

𝑔′ (𝑡) = 𝐹 ′ (𝑓(𝑡))𝑓 ′ (𝑡) = 𝐹 ′ (𝑥)𝑓 ′ (𝑡)

If 𝑓 ′ (𝑡) ≠ 0, we can solve for 𝐹 ′ (𝑥):

𝑔′ (𝑡)
𝐹 ′ (𝑥) = ……………………………….(1)
𝑓 ′ (𝑡)

Since the slope of the tangent to the curve 𝑦 = 𝐹(𝑥) at (𝑥, 𝐹(𝑥)) is 𝐹 ′ (𝑥), equation 1 enables us
to find tangents to parametric curves without having to eliminate the parameter. Using Leibniz
notation, we can, we can rewrite equation 1 in an easily remembered form.

𝑑𝑦
𝑑𝑦 𝑑𝑥
= 𝑑𝑡
𝑑𝑥 if ≠ 0 …………………………..(2)
𝑑𝑥 𝑑𝑡
𝑑𝑡

𝑑𝑦
It can be seen from Equation 2 that the curve has horizontal tangent when = 0 (provided that
𝑑𝑡
𝑑𝑥 𝑑𝑥 𝑑𝑦
≠ 0) and it has a vertical tangent when = 0(provided that 𝑑𝑡 ≠ 0).This information is
𝑑𝑡 𝑑𝑡
useful for sketching parametric curves.
It is also useful to consider 𝑑2 𝑦⁄𝑑𝑥 2 . This can be found by replacing y by 𝑑𝑦⁄𝑑𝑥 in equation 2:

𝑑 𝑑𝑦
𝑑2𝑦 𝑑 𝑑𝑦 ( )
𝑑𝑡 𝑑𝑥
= ( ) =
𝑑𝑥 2 𝑑𝑥 𝑑𝑥 𝑑𝑥
𝑑𝑡

Example 5.14: A curve C is defined by the parametric equation𝑥 = 𝑡 2 ,𝑦 = 𝑡 3 − 3𝑡.

(a) Show that C has two tangents at the point (3,0) and find their equations
(b) Find the points on C where the tangent is horizontal or vertical

Badri A, Moges B. and Teklebrhan B. 155 AKU

 APPLIED MATHEMATICS III

Solution
(a) Notice that 𝑦 = 𝑡 3 − 3𝑡 = 𝑡(𝑡 2 − 3) = 0 when 𝑡 = 0 or 𝑡 = ±√3.Therefore the point (0,3)
On C arises from two values of the parameter, 𝑡 = √3 and 𝑡 = −√3 .this indicates that C crosses
itself at (0,3).since

𝑑𝑦 𝑑𝑦⁄𝑑𝑡 3𝑡 2 − 3 3 1
= = = (𝑡 − )
𝑑𝑥 𝑑𝑥⁄𝑑𝑡 2𝑡 2 𝑡

The slope of the tangent when 𝑡 = ±√3 is 𝑑𝑦⁄𝑑𝑥 = ±√3(2√3) = ±√3, so the equation of the
tangents at (3,0) are

𝑦 = √3(𝑥 − 3) and 𝑦 = −√3(𝑥 − 3)

(b) C has a horizontal tangent when 𝑑𝑦⁄𝑑𝑥 = 0,that is, 𝑑𝑦⁄𝑑𝑡 = 0 and 𝑑𝑥⁄𝑑𝑡 ≠ 0. Since
𝑑𝑦⁄𝑑𝑡 = 3𝑡 2 − 3,this happens when 𝑡 2 = 1,that is, 𝑡 = ±1.the corresponding point on C are
(1,-2) and (1,2).C has a vertical tangent when 𝑑𝑥⁄𝑑𝑡 = 2𝑡 = 0,that is, 𝑡 = 0.(Note that𝑑𝑦⁄𝑑𝑡 ≠
0 𝑡 there.) the corresponding point on C is (0,0).

Fig 5.21 The curve C; 𝑥 = 𝑡 2 ,𝑦 = 𝑡 3 − 3𝑡.

Example 5.15
(a) Find the tangent to the cycloid 𝑥 = 𝑟(𝜃 − 𝑠𝑖𝑛𝜃), 𝑦 = 𝑟(1 − 𝑐𝑜𝑠𝜃) at the point where 𝜃 =
𝜋⁄
3
(b) At what points is the tangent horizontal? When it is vertical?
Solution
(a) The slope of the tangent line is

𝑑𝑦 𝑑𝑦⁄𝑑 𝜃 𝑟𝑠𝑖𝑛𝜃 𝑠𝑖𝑛𝜃

= = 𝑟(1−𝑐𝑜𝑠𝜃) = (1−𝑐𝑜𝑠𝜃)
𝑑𝑥 𝑑𝑥⁄𝑑𝜃
When 𝜃 = 𝜋⁄3 , we have

Badri A, Moges B. and Teklebrhan B. 156 AKU

 APPLIED MATHEMATICS III

𝑟
𝑥 = 𝑟(𝜋⁄3 − 𝑠𝑖𝑛 𝜋⁄3) = 𝑟 (𝜋⁄3 − √3⁄2) 𝑦 = 𝑟(1 − 𝑐𝑜𝑠 𝜋⁄3) = 2

𝑑𝑦 sin(𝜋⁄3) √3⁄2
And = = = √3
𝑑𝑥 1−cos(𝜋⁄3) 1−
1
2

Therefore the slope of the tangent is √3 and its equation is

𝑟 𝑟𝜋 𝑟√3 𝜋
𝑦 − 2 = √3 (𝑥 − + ) or √3 x − y = r (√3 − 2)
3 2

The tangent is sketched in Fig 5.22 below

Fig 5.22 tangents of 𝑥 = 𝑟(𝜃 − 𝑠𝑖𝑛𝜃), 𝑦 = 𝑟(1 − 𝑐𝑜𝑠𝜃)

(b) The tangent is horizontal when 𝑑𝑦⁄𝑑𝑥 = 0, which occurs when sin 𝜃 = 0 and 1 − cos 𝜃 ≠
0, that is, 𝜃 = (2𝑛 − 1)𝜋, 𝑛 an integer. The corresponding point on the cycloid is
((2𝑛 − 1)𝜋𝑟, 2𝑟).
When 𝜃 = 2𝑛𝜋, both 𝑑𝑥⁄𝑑 𝜃 and 𝑑𝑦⁄𝑑 𝜃 are 0. It appears from the graph that there are
vertical tangents at those points. We can verify this by using L’Hospital’s rule as follows:

𝑑𝑦 sin 𝜃 cos 𝜃
lim + = lim + = lim + =∞
𝜃→2𝑛𝜋 𝑑𝑥 𝜃→2𝑛𝜋 1 − cos 𝜃 𝜃→2𝑛𝜋 sin 𝜃

A similar computation shows that 𝑑𝑦⁄𝑑 𝑥 → −∞ as 𝜃 → 2𝑛𝜋 − , so indeed there are vertical
tangents 𝜃 = 2𝑛𝜋, that is, when 𝑥 = 2𝑛𝜋𝑟

Exercise 5.2

1. Find parametric equations for the circle with center (ℎ, 𝑘) and radius 𝑟

2. What curve is represented by the given parametric equations?

𝑥 = 𝑠𝑖𝑛2𝑡 𝑦 = 𝑐𝑜𝑠2𝑡 0 ≤ 𝑡 ≤ 2𝜋

Badri A, Moges B. and Teklebrhan B. 157 AKU

 APPLIED MATHEMATICS III

3. Eliminate the parameter to find a Cartesian equation of the curve

i) 𝑥 = 3𝑡 − 5, 𝑦 = 2𝑡 + 1 ii) 𝑥 = 𝑡 2 − 2, 𝑦 = 5 − 2𝑡, −3 ≤ 𝑡 ≤ 4

4. In each of the following find an equation of the tangent to the curve at the point corresponding to the
given value of the parameter.

(a) 𝑥 = 𝑡 4 + 1, 𝑦 = 𝑡 3 + 𝑡; 𝑡 = −1

(b) 𝑥 = 𝑡 − 𝑡 −1 , 𝑦 = 1 + 𝑡 2 ; 𝑡=1

(c) 𝑥 = 𝑒 √𝑡 , = 𝑡 − 𝑙𝑛 𝑡 2 ; 𝑡 = 1

5.4 Gradient of a Scalar Field; Divergence and Curl of a Vector

Field

Section Objectives:

Upon successful completion of this chapter, the student will be able to:

 Understand the concept of gradient of a scalar field.

 Determine whether a vector field is conservative.
 Find the curl of a vector field.
 Find the divergence of a vector field.

Gradient of a Scalar Field:

Gradient is one of the simplest and most important types of vector field. We may have noticed
that performing a partial derivative is very much like taking the derivative in a particular
𝑑𝑓
direction, i.e. the partial derivative 𝑑𝑥 measures the rate of increase, or the slope, of the function

,𝑓 in the 𝑥 direction . Since there are only three directions in three- dimensional space there is a
neat and elegant way of summarizing all the information about how the function is increasing,
we simply put all the partial derivatives of the function into a vector. This vector is known as the
gradient of the function.

Badri A, Moges B. and Teklebrhan B. 158 AKU

 APPLIED MATHEMATICS III

Definition: Let 𝑓: 𝑅 3 → 𝑅 be a scalar field, that is a function of three variables. The gradient
of 𝑓, denoted ∇𝑓, is the vector field given by

𝜕𝑓 𝜕𝑓 𝜕𝑓 𝜕𝑓 𝜕𝑓 𝜕𝑓
∇𝑓 = 〈𝜕𝑥 , 𝜕𝑦 , 𝜕𝑧 〉 = 𝜕𝑥 𝒊 + 𝒋 + 𝜕𝑧 𝒌.
𝜕𝑦

Symbolically, we can consider ∇ to be a vector of differential operators, that is

𝜕 𝜕 𝜕
∇= 𝒊 + 𝒋+ 𝒌
𝜕𝑥 𝜕𝑦 𝜕𝑧

Then ∇𝑓 is symbolically the “vector” ∇ “times” the “ scalar” 𝑓.

The gradient has the following properties:

 ∇𝑓 is a vector field
 ∇𝑓 measures the rate of increase of the scalar function 𝑓 in each of the three coordinate
directions.
 ∇𝑓 Points in the direction in which 𝑓 increases the most.

Example 5.16: If 𝑓(𝑥, 𝑦) = 𝑥 2 𝑦 + 3𝑥𝑦 3 then the gradient of 𝑓 is

∇𝑓(𝑥, 𝑦) = 𝑓𝑥 (𝑥, 𝑦)𝒊 + 𝑓𝒚 (𝑥, 𝑦)𝒋

= (2𝑥𝑦 + 3𝑦 3 )𝒊 + (𝑥 2 + 9𝑥𝑦 2 )𝑱

is a vector field in the plane.

Example 5.17: Similarly if 𝑓(𝑥, 𝑦, 𝑧) = 𝑥 2 + 𝑦 2 + 𝑧 2

Then the gradient of 𝑓

∇𝑓(𝑥, 𝑦) = 𝑓𝑥 (𝑥, 𝑦)𝒊 + 𝑓𝒚 (𝑥, 𝑦)𝒋 + 𝑓𝑧 (𝑥, 𝑦, 𝑧)𝒌

= 2𝑥𝒊 + 2𝑦𝒋 + 3z

is a vector field in space

Divergence and Curl

Badri A, Moges B. and Teklebrhan B. 159 AKU

 APPLIED MATHEMATICS III

Divergence and curl are two measurements of vector fields that are very useful in a variety of
applications. Both are most easily understood by thinking of the vector field as representing a
flow of a liquid or gas; that is, each vector in the vector field should be interpreted as a velocity
vector. Roughly speaking, divergence measures the tendency of the fluid to collect or disperse at
a point, and curl measures the tendency of the fluid to swirl around the point. Divergence is a
scalar, that is, a single number, while curl is itself a vector. The magnitude of the curl measures
how much the fluid is swirling, the direction indicates the axis around which it tends to swirl.

Definition: Let F be a vector field given by 𝑭 = f𝒊 + g𝒋 + h𝒌 ,

where, 𝑓, 𝑔, and ℎ are scalar functions. The divergence of 𝑭 is

𝜕𝑓 𝜕𝑔 𝜕ℎ
𝑑𝑖𝑣 𝑭 = + + ,
𝜕𝑥 𝜕𝑦 𝜕𝑧
and the Curl of 𝐹 is
𝜕ℎ 𝜕𝑔 𝜕𝑓 𝜕ℎ 𝜕𝑔 𝜕𝑓
𝐶𝑢𝑟𝑙 𝑭 = ( − )𝒊 + ( − )𝒋 + ( − )𝒌
𝜕𝑦 𝜕𝑧 𝜕𝑧 𝜕𝑥 𝜕𝑥 𝜕𝑦

These are best interpreted in terms of the velocity field of a fluid flow. The divergence is the rate
of expansion of the fluid at a point. The curl is a vector describing the rotation of the fluid near
the point (the direction of the curl is the axis of rotation and the magnitude is a measure of the
rate of rotation). The flow is called incompressible if its divergence is zero, and irrotational if
its curl is zero.

Note: Recall that if 𝑓 is a function, the gradient of 𝑓 is given by

∂f ∂f ∂f
∇f = 〈 , , 〉
∂x ∂y ∂z
A useful mnemonic for the divergence and Curl is , let

Badri A, Moges B. and Teklebrhan B. 160 AKU

 APPLIED MATHEMATICS III

∂ ∂ ∂
∇= 〈 , , 〉,
∂x ∂y ∂z

That is, we pretend that ∇ is a vector with rather odd looking entires . We can then think of the
gradient as
∂ ∂ ∂ ∂f ∂f ∂f
∇f = 〈 , , 〉𝑓 = 〈 , , 〉
∂x ∂y ∂z ∂x ∂y ∂z

that is, we simply multiply the f into the vector.

The divergence and curl can now be defined in terms of this same odd vector 𝛻 by using the
cross product and dot product. The divergence of a vector field 𝐹 = 〈𝑓, 𝑔, h〉 is

∂ ∂ ∂ ∂f ∂g ∂h
∇ ∙ 𝐅 = 〈∂x , ∂y , ∂z〉. 〈𝑓, 𝑔, h〉 = + ∂y + ∂z .
∂x

The curl of 𝑭 is
𝑖 𝑗 𝑘
∂ ∂ ∂ 𝜕ℎ 𝜕𝑔 𝜕𝑓 𝜕ℎ 𝜕𝑔 𝜕𝑓
| |
∇ × 𝐹 = ∂x
| ∂y ∂z | = 〈 𝜕𝑦 − 𝜕𝑧 , 𝜕𝑧 − 𝜕𝑥 , 𝜕𝑥 − 𝜕𝑦〉
𝑓 𝑔 ℎ

Here are two simple but useful facts about divergence and curl.

i. 𝛻 ∙ (𝛻 × 𝐹) = 0.
In words, this says that the divergence of the curl is zero.

j. 𝛻 × (𝛻𝑓) = 0.

That is, the curl of a gradient is the zero vector. Recalling that gradients are
conservative vector fields, this says that the curl of a conservative vector field is the
zero vector. Under suitable conditions, it is also true that if the curl of 𝐹 is 0 then 𝐹 is
conservative.

Example 5.18: Find Curl 𝐹 of the vector field given by:

Badri A, Moges B. and Teklebrhan B. 161 AKU

 APPLIED MATHEMATICS III

𝑓(𝑥, 𝑦, 𝑧) = 2𝑥𝑦𝒊 + (𝑥 2 + 𝑦 2 )𝒋 + 2𝑦𝑧𝒌

Is 𝐹 irrotational ?
Solution: The Curl of 𝐹 is given by
𝑖 𝑗 𝑘
∂ ∂ ∂
| |
𝐶𝑢𝑟𝑙 𝐹(𝑥, 𝑦, 𝑧) = ∇ × 𝐹 = ∂x ∂y ∂z |
|
2𝑥𝑦 𝑥 2 + 𝑦 2 2𝑥𝑦

(After finding the determinant),

= (2𝑧 − 2𝑧)𝑖 − (0 − 0)𝑗 + (2𝑥 − 2𝑥)𝑘
=0
Because, 𝐶𝑢𝑟𝑙 𝐹 = 0, 𝐹 is irrational.
Example 5.19: Find the divergence at (2,1, −1) for the vector field

𝐹(𝑥, 𝑦, 𝑧) = 𝑥 2 𝑦 2 𝑧𝑖 + 𝑥 2 𝑧𝑗 + 𝑥 2 𝑦𝑘.
Solution: The divergence of 𝐹 is
𝜕 3 2 𝜕 2 𝜕
𝑑𝑖𝑣𝐹(𝑥, 𝑦, 𝑧) = [𝑥 𝑦 𝑧] + [𝑥 𝑧] + [𝑥 2 𝑦]
𝜕𝑥 𝜕𝑦 𝜕𝑧
= 3𝑥 2 𝑦 2 𝑧
At the point (2,1, −1), the divergence is
𝑑𝑖𝑣 𝐹(2,1, −1) = 3(22 )(12 )(−1)
= 12
Piecewise Smooth Curves
A classic property of gravitational fields is that, subject to certain physical constraints, the work
done by gravity on an object moving between two points in the field is independent of the path
taken by the object. One of the constraints is that path must be a piecewise smooth curve. Recall
that a plane curve C given by
𝐫(t) = x(t)𝐢 + y(t)𝐣, a≤t≤b
is smooth if
𝑑𝑥 𝑑𝑦
and
𝑑𝑡 𝑑𝑡

are continuous on [𝑎, 𝑏] and not simultaneously 0 on (𝑎, 𝑏). Similarly, a space curve C given

Badri A, Moges B. and Teklebrhan B. 162 AKU

 APPLIED MATHEMATICS III

𝐫(t) = x(t)𝐢 + y(t)𝐣 + z(t)𝐊, a≤t≤b

is smooth if
𝑑𝑥 𝑑𝑦 𝑑𝑧
, and 𝑑𝑡
𝑑𝑡 𝑑𝑡

Are continuous on [𝑎, 𝑏] and not simultaneously 0 on (𝑎, 𝑏). A curve 𝐶 is piecewise smooth if
the interval [𝑎, 𝑏] can be partitioned into a finite number of subintervals, on each which 𝐶 is
smooth.

Example 5.19: Find a piecewise smooth parameterizations of the graph of 𝐶 shown in figure
below

Solution: Since 𝐶 consists of three line segments 𝐶1 , 𝐶2 𝑎𝑛𝑑 𝐶3 , you can construct a smooth
parameterization for each segment and piece them together by making the last 𝑡-value in 𝐶𝑖
Correspond to the first t-value in 𝐶𝑖+1 , as follows.
𝐶1 : 𝑥(𝑡) = 0, 𝑦(𝑡) = 2𝑡, 𝑧(𝑡) = 0 , 0≤𝑡≤1
𝐶2 ∶ 𝑥(𝑡) = 𝑡 − 1, 𝑦(𝑡) = 2 , 𝑧(𝑡) = 0, 1≤𝑡≤2
𝐶3 : 𝑥(𝑡) = 1, 𝑦(𝑡) = 2 , 𝑧(𝑡) = 𝑡 − 2, 2≤𝑡 ≤3

5.5 Line Integrals and Greens Therom

Section Objectives:

Badri A, Moges B. and Teklebrhan B. 163 AKU

 APPLIED MATHEMATICS III

At the end of this section, students should able to:

 Write and evaluate a line integral.
 Write and evaluate a line integral of a vector field.
 Evaluate a line integral in differential form.
 Use Green’s Theorem to evaluate a line integral.
 Use alternative forms of Green’s Theorem

Line Integrals

Introduction: In this section , we consider some new concepts of line integrals. This new kinds
of integrals will be defined as limits of sums in the same general way that single integrals are
defined . An ordinary single integral
𝑏
∫ 𝑓(𝑥) 𝑑𝑥
𝑎

is an integral of a function which is defined along a line segment (an interval of a co –ordinate
axis). There is a corresponding kind of integral for a function which is defined along a curve.
Such an integral might well be called a curvilinear integral; the usual name is line integral, where
line means, in general, a curved line.

Definition: If 𝑓 is defined in a region containing a smooth curve 𝑐 of finte length , then the line
integral of 𝑓 along 𝑐 is given by
∫𝑪 𝒇(𝒙, 𝒚) 𝒅𝒔 = 𝐥𝐢𝐦 ∑𝒏𝒊=𝟏 𝒇(𝒙𝒊 , 𝒚𝒊 ) ∆𝒔𝒊 Plane
||∆||→𝟎

∫𝑪 𝒇(𝒙, 𝒚, 𝒛) 𝒅𝒔 = 𝐥𝐢𝐦 ∑𝒏𝒊=𝟏 𝒇(𝒙𝒊 , 𝒚𝒊 , 𝒛𝒊 ) ∆𝒔𝒊 space

||∆||→𝟎

provided this limit exist.

Evaluation of a line integral is best accomplished by converting it to a definite integral. It can be

shown that if 𝑓 is continuous, the limit given above exists and is the same for all smooth
parameterizations of 𝐶. To evaluate a line integral over a plane curve 𝐶 given by

𝑟(𝑡) = 𝑥(𝑡)𝑖 + 𝑦(𝑡)𝑗

Badri A, Moges B. and Teklebrhan B. 164 AKU

 APPLIED MATHEMATICS III

use the fact that

𝑑𝑠 = ||𝑟 ′ (𝑡)||𝑑𝑡 = √[𝑥 ′ (𝑡)]2 + [𝑦 ′ (𝑡)]2 𝑑𝑡
A similar formula holds for a space curve.

Theorem: (Evaluation of a line integral as a definite integral)

Let 𝑓 be continuous in a region containing a smooth curve 𝐶. If 𝐶 is given by

𝑟(𝑡) = 𝑥(𝑡)𝑖 + 𝑦(𝑡)𝑗 , where 𝑎 ≤ 𝑡 ≤ 𝑏, , then

𝑏
∫𝐶 𝑓(𝑥, 𝑦) 𝑑𝑠 = ∫𝑎 𝑓(𝑥(𝑡), 𝑦(𝑡)) √[𝑥 ′ (𝑡)]2 + [𝑦 ′ (𝑡)]2 𝑑𝑡.
If 𝐶 is given by

𝑟(𝑡) = 𝑥(𝑡)𝑖 + 𝑦(𝑡)𝑗 + 𝑧(𝑡)𝑘 , where 𝑎 ≤ 𝑡 ≤ 𝑏, , then

𝑏
∫𝐶 𝑓(𝑥, 𝑦, 𝑧) 𝑑𝑠 = ∫𝑎 𝑓(𝑥(𝑡), 𝑦(𝑡), 𝑧(𝑡)) √[𝑥 ′ (𝑡)]2 + [𝑦 ′ (𝑡)]2 + [𝑧 ′ (𝑡)]2 𝑑𝑡.

Example 5.20: Evaluate a line integral

∫(𝑥 2 − 𝑦 − 3𝑧) 𝑑𝑠
𝐶

where 𝐶 is the line segment shown in figure below

Figure 5.1

Badri A, Moges B. and Teklebrhan B. 165 AKU

 APPLIED MATHEMATICS III

Solution: A parametric form of the equation of the line segment:

𝑥 = 𝑡, 𝑦 = 2𝑡 , and 𝑧 = 𝑡 , 0≤𝑡≤1
Therefore,
𝑥 ′ (𝑡) = 1, 𝑦 ′ (𝑡) = 2 and 𝑧 ′ (𝑡) = 1
This implies that,

√[𝑥 ′ (𝑡)]2 + [𝑦 ′ (𝑡)]2 + [𝑧 ′ (𝑡)]2 = √12 + 22 + 12 = √6

So the line integral takes the following form;

∫(𝑥 2 − 𝑦 − 3𝑧) 𝑑𝑠 = ∫(𝑡 2 − 2𝑡 − 3𝑡)√6 𝑑𝑡

𝐶 𝐶

= √6 ∫(𝑡 2 + 𝑡)𝑑𝑡
𝐶
5√6
= 6

Suppose 𝐶 is a path composed of smooth curves 𝐶1 , 𝐶2 , . . . 𝐶𝑛 . If 𝑓 is continuous

on 𝐶, it can be shown that

∫𝑓(𝑥, 𝑦) 𝑑𝑠 = ∫ 𝑓(𝑥, 𝑦) 𝑑𝑠 + ∫ 𝑓(𝑥, 𝑦) 𝑑𝑠 + . . . + ∫ 𝑓(𝑥, 𝑦) 𝑑𝑠

𝐶 𝐶1 𝐶2 𝐶𝑛

Example 5.21: Evaluating a line integral over a path

Evaluate ∫𝐶 𝑥𝑑𝑠 , where 𝐶 is the piecewise smooth curve shown below

Badri A, Moges B. and Teklebrhan B. 166 AKU

 APPLIED MATHEMATICS III

Figure 5.2
Solution: Begin by integrating up the line 𝑦 = 𝑥 , using the following parameterization
𝐶1 = 𝑥 = 𝑡, 𝑦 = 𝑡 , 0≤𝑡≤1

For this curve

𝑟(𝑡) = 𝑡𝒊 + 𝑡𝒋
This implies that,
𝑥 ′ (𝑡) = 1 and 𝑦 ′ (𝑡) = 1. So,

√[𝑥 ′ (𝑡)]2 + [𝑦 ′ (𝑡)]2 = √2

and we have
1
√2
∫ 𝑥𝑑𝑠 = ∫ 𝑡√2𝑑𝑡 =
𝐶1 0 2
Next , integrate down the parabola 𝑦 = 𝑥 2 , using the parameterization
𝐶2 : 𝑥 = 1 − 𝑡 , 𝑦 = (1 − 𝑡)2 , 0≤𝑡≤1
For this curve ,
𝑟(𝑡) = (1 − 𝑡)𝑖 + (1 − 𝑡)2 𝑗
This implies that,
𝑥 ′ (𝑡 = −1) and 𝑦 ′ (𝑡) = −2(1 − 𝑡) . So,

Badri A, Moges B. and Teklebrhan B. 167 AKU

 APPLIED MATHEMATICS III

√[𝑥 ′ (𝑡)]2 + [𝑦 ′ (𝑡)]2 = √1 + 4(1 − 𝑡)2

and we have
1
∫ 𝑥𝑑𝑠 = ∫ (1 − 𝑡) √1 + 4(1 − 𝑡)2 𝑑𝑡
𝐶2 0

−1 2 3
1
= ([ (1 + 4(1 − 𝑡)2 )])2 |
8 3 0
3
1
= 12 (5 2 − 1)

Consequently,

∫𝑥 𝑑𝑠 = ∫ 𝑥 𝑑𝑠 + ∫ 𝑥 𝑑𝑠
𝐶 𝐶1 𝐶2

√2 1 3
= + (5 2 − 1) ≅ 1.56
2 12

Example 5.22: Evaluate ∫𝐶(𝑥 + 2) 𝑑𝑠 , where 𝐶 is the cure represented by

4 3 1
𝑟(𝑡) = 𝑡𝒊 + 𝑡2 𝒋 + 𝑡2𝒌 , 0≤𝑡≤2
3 2
1
Solution: Since 𝑟 ′ (𝑡) = 𝒊 + 2𝑡 2 𝒋 + 𝑡𝒌 and

||𝑟 ′ (𝑡)|| = √[𝑥 ′ (𝑡)]2 + [𝑦 ′ (𝑡)]2 + [𝑧 ′ (𝑡)]2 = √1 + 4𝑡 + 𝑡 2

it follows that
2
∫𝐶(𝑥 + 2) 𝑑𝑠 = ∫0 (𝑡 + 2) √1 + 4𝑡 + 𝑡 2 𝑑𝑡
1 2 1
= ∫ 2(𝑡 + 2)(1 + 4𝑡 + 𝑡 2 )2 𝑑𝑡
2 0
1
= 3 (13√13 − 1) ≅ 15.29

Line Integrals of Vector Fields

One of the most important physical applications of line integrals is that of finding the work done
on an object moving in a force field. To see how a line integral can be used to find work done in
a force field 𝑭 , consider an object moving along a path 𝐶 in the field , as shown in figure
below. To determine the work done by the force, you need consider only that part of the force

Badri A, Moges B. and Teklebrhan B. 168 AKU

 APPLIED MATHEMATICS III

that is acting in the same direction as that in which the object is moving (or the opposite
direction). This means that at each point on 𝐶, you can consider the projection 𝑭. 𝑻 of the
force vector 𝑭 onto the unit tangent vector 𝑻. On a small subarc of length ∆𝑠𝑖 , the increment
of work is
∆𝑊𝑖 = (𝑓𝑜𝑟𝑐𝑒)(𝑑𝑖𝑠𝑡𝑎𝑛𝑐𝑒)
≅ [𝐹(𝑥𝑖 , 𝑦𝑖, 𝑧𝑖 ). 𝑇(𝑥𝑖 , 𝑦𝑖, 𝑧𝑖 )]∆𝑠𝑖
Where (𝑥𝑖 , 𝑦𝑖, 𝑧𝑖 ) is a point in the 𝑖 𝑡ℎ subarc. Consequently, the total work done is given by the
following integral.

𝑊 = ∫𝐹(𝑥, 𝑦, 𝑧. 𝑇(𝑥, 𝑦, 𝑧) 𝑑𝑠
𝐶

This line integral appears in other contexts and is the basis of the following definition of the
line integral of a vector field . Note in this definition

𝐫 ′ (t)
𝐅 ∙ 𝐓ds = 𝐅 ∙ ‖𝐫′(t)‖dt
‖𝐫′(t)‖
= 𝐅 ∙ 𝐫′(t)dt
= 𝐅 ∙ d𝐫
Definition: Let 𝑭 be a continuous vector field defined on a smooth curve 𝐶 given by 𝒓(𝑡),
𝑎 ≤ 𝑡 ≤ 𝑏. The line integral of 𝑭 on 𝐶 is given by
𝑏

∫ 𝑭 ∙ 𝑑𝒓 = ∫ 𝑭 ∙ 𝑻𝑑𝑠 = ∫ 𝑭(𝑥(𝑡), 𝑦(𝑡), 𝑧(𝑡)) ∙ 𝒓′ (𝑡)𝑑𝑡

𝐶 𝐶 𝑎

Badri A, Moges B. and Teklebrhan B. 169 AKU

 APPLIED MATHEMATICS III

Example 5.23: find the work done by the force field

1 1 1
𝑭(𝑥, 𝑦, 𝑧) = − 𝑥𝒊 − 𝑦𝒋 + 𝒌
2 2 4
on a particle as it moves along the helix given by
𝒓(𝑡) = 𝑐𝑜𝑠𝑡𝒊 + 𝑠𝑖𝑛𝑡𝒋 + 𝑡𝒌
from the point (1,0,0) to (−1,0,3𝜋) as shown in the figure below.

Figure 5.4
Solution: since
𝒓(𝑡) = 𝑥(𝑡)𝒊 + 𝑦(𝑡)𝒋 + 𝑧(𝑡)𝒌 = 𝒓(𝑡) = 𝑐𝑜𝑠𝑡𝒊 + 𝑠𝑖𝑛𝑡𝒋 + 𝑡𝒌
It follows that
𝑥(𝑡) = 𝑐𝑜𝑠𝑡, 𝑦(𝑡) = 𝑠𝑖𝑛𝑡, and 𝑧(𝑡) = 𝑡.
So, the force field can be written as
1 1 1
𝑭(𝑥(𝑡), 𝑦(𝑡), 𝑧(𝑡)) = − 𝑐𝑜𝑠𝒊 − 𝑠𝑖𝑛𝒋 + 𝒌
2 2 4
To find the work done by the force field in moving a particle along the curve 𝐶, use the fact that
𝒓′(𝑡) = −𝑠𝑖𝑛𝑡𝒊 + 𝑐𝑜𝑠𝑡𝒋 + 𝒌
and write the following.

𝑊 = ∫ 𝑭 ∙ 𝑑𝒓
𝐶

Badri A, Moges B. and Teklebrhan B. 170 AKU

 APPLIED MATHEMATICS III

= ∫ 𝑭(𝑥(𝑡), 𝑦(𝑡), 𝑧(𝑡)) ∙ 𝒓′ (𝑡)𝑑𝑡

𝑎
3𝜋
1 1 1
= ∫ (− 𝑐𝑜𝑠𝑡𝒊 − 𝑠𝑖𝑛𝑡𝒋 + 𝒌) ∙ (−𝑠𝑖𝑛𝑡𝒊 + 𝑐𝑜𝑠𝑡𝒋 + 𝑲)𝑑𝑡
2 2 4
0
3𝜋 3𝜋
1 1 1 1 1 3𝜋
= ∫ ( 𝑠𝑖𝑛𝑡𝑐𝑜𝑠𝑡 − 𝑠𝑖𝑛𝑡𝑐𝑜𝑠𝑡 + ) 𝑑𝑡 = ∫ 𝑑𝑡 = 𝑡]3𝜋
0 =
2 2 4 4 4 4
0 0

Theorem (Fundamental Theorem of Line Integrals)

Let 𝐶 be a piecewise smooth curve lying in an open region 𝑅 and given by
𝑟(𝑡) = 𝑥(𝑡)𝒊 + 𝑦(𝑡)𝒋, 𝑎 ≤ 𝑡 ≤ 𝑏.
If 𝑭(𝑥, 𝑦) = 𝑀𝒊 + 𝑁𝒋 is conservative in 𝑅, and 𝑀 and 𝑁 are continuous in 𝑅,
Then

∫𝑭. 𝑑𝒓 = ∫𝛁𝑓 ∙ 𝑑𝒓 = 𝑓(𝑥(𝑏), 𝑦(𝑏)) − 𝑓(𝑥(𝑎), 𝑦(𝑎))

𝑪 𝐶

Where 𝑓 is a potential function of 𝑭. That is, 𝑭(𝑥, 𝑦) = 𝛁𝑓(𝑥, 𝑦).

Proof: A proof is provided only for a smooth curve. For piecewise smooth curves, the procedure
is carried out separately on each smooth portion. Because
𝐹(𝑥, 𝑦) = 𝛁𝑓𝑥 (𝑥, 𝑦)𝒊 + ∇𝑓𝑦 (𝑥, 𝑦)𝒋
It follows that
𝑏 𝑏
𝑑𝒓 𝑑𝑥 𝑑𝑦
∫ 𝑭 ∙ 𝑑𝒓 = ∫ 𝐹 ∙ 𝑑𝑡 = ∫ [𝑓𝑥 (𝑥, 𝑦) + 𝑓𝑦 (𝑥, 𝑦) ]𝑑𝑡
𝐶 𝑎 𝑑𝑡 𝑎 𝑑𝑡 𝑑𝑡
Using Chain rule, we have
𝑏
𝑑
∫𝑭 ∙ 𝑑𝒓 = ∫ [𝑓(𝑥(𝑡), 𝑦(𝑡))]𝑑𝑡 = 𝑓(𝑥(𝑏), 𝑦(𝑏)) − 𝑓(𝑥(𝑎), 𝑦(𝑎)).
𝐶 𝑎 𝑑𝑡
The last step is an application of the Fundamental Theorem of Calculus. That is, using
𝑏
∫𝑎 𝑓 ′ (𝑥)𝑑𝑥 = 𝑓(𝑏) − 𝑓(𝑎) █
In space, the Fundamental Theorem of Line Integrals takes the following form. Let 𝑐 be a
piecewise smooth curve lying in an open region 𝑄 and given by
𝒓(𝑡) = 𝑥(𝑡)𝒊 + 𝑦(𝑡)𝒋 + 𝑧(𝑡)𝒌, 𝑎 ≤ 𝑡 ≤ 𝑏.
Badri A, Moges B. and Teklebrhan B. 171 AKU
 APPLIED MATHEMATICS III

If 𝐹(𝑥, 𝑦, 𝑧) = 𝑀𝒊 + 𝑁𝒋 + 𝑃𝒌 is conservative and 𝑀, 𝑁, and 𝑃 are continuous , then

∫𝑭 ∙ 𝑑𝒓 = ∫𝛁𝑓 ∙ 𝑑𝒓 = 𝑓(𝑥(𝑏), 𝑦(𝑏), 𝑧(𝑏)) − 𝑓(𝑥(𝑎), 𝑦(𝑎), 𝑧(𝑎))

𝐶 𝐶

where, 𝐹(𝑥, 𝑦, 𝑧) = 𝛁f(x, y, z).

The Fundamental Theorem of Line Integrals states that if the vector field 𝑭 is conservative,
then the line integral between any two points is simply the difference in the values of the
potential function 𝑓 at these points.
Example 5.24: Using the Fundamental Theorem of Line Integrals
Evaluate ∫𝐶 𝑭. 𝑑𝒓 , where 𝐶 is a piecewise smooth curve from (−1, 4) to (1,2) and
𝑭(𝑥, 𝑦) = 2𝑥𝑦 𝒊 + (𝑥 2 − 𝑦)𝒋 as shown in figure below

Figure 5.4
Solution: 𝑭 is the gradient of 𝑓 , where
𝑦2
𝑓(𝑥, 𝑦) = 𝑥 2 𝑦 − +𝒌
2
Consequently, 𝑭 is conservative and by the Fundamental Theorem of Line Integrals, it follows
that

∫𝑭 ∙ 𝑑𝒓 = 𝑓(1,2) − 𝑓(−1,4)
𝐶

22 42
= [12 (2) − ] − [(−1)2 (4) − ]
2 2
= 4.

Badri A, Moges B. and Teklebrhan B. 172 AKU

 APPLIED MATHEMATICS III

Note that it is unnecessary to include a constant 𝑘 as part of 𝑓, because it is canceled by

subtraction.

Example 5.25: Evaluate ∫𝐶 𝑭. 𝑑𝒓 ,

where, 𝐶 is a piecewise smooth curve from (1,1,0) to (0,2,3)
𝐹(𝑥, 𝑦, 𝑧) = 2𝑥𝑦𝒊 + (𝑥 2 + 𝑧 2 )𝒋 + 2𝑦𝑧𝒌

Answer, ∫𝐶 𝑭. 𝑑𝒓 = 17

Green’s Theorem
We now come to the first of three important theorems that extend the Fundamental Theorem of
Calculus to higher dimensions. (The Fundamental Theorem of Line Integrals has already done
this in one way, but in that case we were still dealing with an essentially one-dimensional
integral.) They all share with the Fundamental Theorem the following rather vague description:
To compute a certain sort of integral over a region, we may do a computation on the boundary
of the region that involves one fewer integration. Note that this does indeed describe the
Fundamental Theorem of Calculus and the Fundamental Theorem of Line Integrals: to compute
a single integral over an interval, we do a computation on the boundary (the endpoints) that
involves one fewer integrations, namely, no integrations at all.
In this section, we will study Green’s Theorem, named after the English mathematician George
Green (1793-1841). This theorem states that the value of a double integral over a simply
connected plane region 𝑅 is determined by the value of a line integral around the boundary of
𝑅. A curve 𝐶 given by 𝑟(𝑡) = 𝑥(𝑡)𝒊 + 𝑦(𝑡)𝒋, where 𝑎 ≤ 𝑡 ≤ 𝑏, is simple if it does not
cross itself, that is , 𝑟(𝑐) ≠ 𝑟(𝑑) for all 𝑐 and 𝑑 in the open interval (𝑎, 𝑏). A plane region
𝑅 is simply connected if every simple closed curve in 𝑅 encloses only points that are in 𝑅.

Theorem: (Green’s Theorem)

Let 𝑅 be a simply connected region with a piecewise smooth boundary 𝐶, oriented
counterclockwise (that is, 𝐶 is traversed once so that the region 𝑅 always lies to the left). If 𝑀
and 𝑁 have continuous first partial derivatives in an open region containing 𝑅, then

Badri A, Moges B. and Teklebrhan B. 173 AKU

 APPLIED MATHEMATICS III

𝜕𝑁 𝜕𝑀
∫𝐶 𝑀 𝑑𝑥 + 𝑁 𝑑𝑦 = ∬𝑅( 𝜕𝑥 − ) 𝑑𝐴. █
𝜕𝑦

To indicate that an integral ∫𝐶 is being done over a closed curve in the counter clockwise
direction, we usually write ∮𝐶. we also use the notation 𝜕𝐷 to mean the boundary of D
oriented in the counterclockwise direction. With ∮𝐶 = ∫𝜕𝐷.

Proof: A proof is given only for a region that is both vertically simple and horizontally simple,
as shown in figure below.

Figure
𝑏 𝑎
∫𝐶 𝑀 𝑑𝑥 = ∫𝐶 𝑀 𝑑𝑥 ∫𝐶 𝑀 𝑑𝑥 = ∫𝑎 𝑀(𝑥, 𝑓1 (𝑥))𝑑𝑥 + ∫𝑏 𝑀(𝑥, 𝑓2 (𝑥)) 𝑑𝑥
1 2

𝑏
= ∫𝑎 [𝑀(𝑥, 𝑓1 (𝑥))𝑑𝑥 − 𝑀(𝑥, 𝑓2 (𝑥))]𝑑𝑥
On the other hand
𝑏 𝑓2 (𝑥)
𝜕𝑀 𝜕𝑀
∬ 𝑑𝐴 = ∫ ∫ 𝑑𝑦𝑑𝑥
𝜕𝑦 𝑎 𝑓1 (𝑥) 𝜕𝑦
𝑅
𝑏
𝑓2
= ∫ 𝑀(𝑥, 𝑦) | 𝑑𝑥
𝑎 𝑓1
𝑏
= ∫ [𝑀(𝑥, 𝑓2 (𝑥)) − 𝑀(𝑥, 𝑓1 (𝑥))]𝑑𝑥
𝑎
𝜕𝑀
Consequently, ∫𝐶 𝑀 𝑑𝑥 = ∬𝑅 𝜕𝑦 𝑑𝐴.
𝜕𝑁
Similarly, you can use 𝑔1 (𝑦) 𝑎𝑛𝑑 𝑔2 (𝑦) to show that ∫𝐶 𝑁𝑑𝑦 = ∫𝑅 ∫ 𝜕𝑥 𝑑𝐴. By adding the

integrals ∫𝐶 𝑀𝑑𝑥 and ∫𝐶 𝑁𝑑𝑦, you obtain the collection stated in the theorem.
▄
Example 5.26:
Use Green’s Theorem to evaluate the line integral

Badri A, Moges B. and Teklebrhan B. 174 AKU

 APPLIED MATHEMATICS III

∫𝑦 3 𝑑𝑥 + (𝑥 3 + 3𝑥𝑦 2 )𝑑𝑦
𝐶

Where 𝐶 is the path from (0,0) to (1,1) along the graph of 𝑦 = 𝑥 3 and from (1,1) to (0,0)
along the graph of 𝑦 = 𝑥 as shown in figure below

𝑪 is simple and closed, and the region 𝑅 always lies to the left of 𝐶
Figure 5.5
Solution:
Since, 𝑀 = 𝑦 3 and 𝑁 = 𝑥 3 + 3𝑥𝑦 2 , it follows that
𝜕𝑁 𝜕𝑀
= 3𝑥 2 + 3𝑦 2 and = 3𝑦 2 .
𝜕𝑥 𝜕𝑦

Applying Green’s Theorem, we then have

𝜕𝑁 𝜕𝑀
∫𝑦 3 𝑑𝑥 + (𝑥 3 + 3𝑥𝑦 2 )𝑑𝑦 = ∬( − ) 𝑑𝐴
𝐶 𝜕𝑥 𝜕𝑦
𝑅
1 𝑥 1
𝑥 1
𝑥
= ∫ ∫[(3𝑥 2 + 3𝑦 2 ) − 3𝑦 2 ] 𝑑𝑦𝑑𝑥 = ∫ ∫ 3𝑥 2 𝑑𝑦𝑑𝑥 = ∫ 3𝑥 2 𝑦] | 3 𝑑𝑥
𝑥3 0 𝑥
0 𝑥3 0
1
2 5)
3𝑥 2 𝑥 6 1 1
= ∫(3𝑥 − 3𝑥 𝑑𝑥 = [ − ]| =
4 2 0 4
0

Example 5.27: While subject to the force

𝑭(𝑥, 𝑦) = 𝑦 3 𝒊 + (𝑥 3 + 3𝑥𝑦 2 )𝒋
a particles travels once around the circle of radius 3, as shown in figure below, using Green’s
Theorem find the work done by 𝑭.

Badri A, Moges B. and Teklebrhan B. 175 AKU

 APPLIED MATHEMATICS III

𝑭(𝑥, 𝑦) = 𝑦 3 𝒊 + (𝑥 3 + 3𝑥𝑦 2 )𝒋

Figure 5.6
Solution: From example one above, (using Green’s Theorem), we have

∫𝑦 3 𝑑𝑥 + (𝑥 3 + 3𝑥𝑦 2 ) 𝑑𝑦 = ∬ 3𝑥 2 𝑑𝐴.
𝐶
𝑅

In polar coordinates, using 𝑥 = 𝑟𝑐𝑜𝑠𝜃, and 𝑑𝐴 = 𝑟𝑑𝑟𝑑𝜃 , the work done is

2𝜋 3

𝑊 = ∬ 3𝑥 2 𝑑𝐴 = ∫ ∫ 3(𝑟𝑐𝑜𝑠𝜃)2 𝑟𝑑𝑟𝑑𝜃
𝑅 0 0
2𝜋 3 2𝜋 𝑟4 2 81
= 3 ∫0 ∫0 𝑟 3 𝑐𝑜𝑠 2 𝜃𝑑𝑟𝑑𝜃 = 3 ∫0 𝑐𝑜𝑠 2 𝜃]30 𝑑𝜃 = 3 ∫0 𝑐𝑜𝑠 2 𝜃𝑑𝜃
4 4
2𝜋
243 243 𝑠𝑖𝑛2𝜃 2𝜋 243𝜋
= ∫ (1 + 𝑐𝑜𝑠2𝜃)𝑑𝜃 = [𝜃 + ]| =
8 8 2 0 4
0

Note: When evaluating line integrals over closed curves, remember that for conservative vector
𝜕𝑁 𝜕𝑀
fields (those for which = ), the value of the line integral is 0. This is easily seen from the
𝜕𝑥 𝜕𝑦

statement of Green’s Theorem:

𝜕𝑁 𝜕𝑀
∫ 𝑀𝑑𝑥 + 𝑁 𝑑𝑦 = ∬( − ) 𝑑𝐴 = 0.
𝜕𝑥 𝜕𝑦
𝐶 𝑅

Example 5.28: Evaluate the line integral

∫ 𝑦 3 𝑑𝑥 + 3𝑥𝑦 2 𝑑𝑦
𝐶

Where 𝐶 is a closed path .

Badri A, Moges B. and Teklebrhan B. 176 AKU

 APPLIED MATHEMATICS III

𝜕𝑁 𝜕𝑀
Solution: From this line integral , 𝑀 = 𝑦 3 and 𝑁 = 3𝑥𝑦 2 . So, = 3𝑦 2 and = 3𝑦 2 .
𝜕𝑥 𝜕𝑦

This implies that the vector field 𝑭 = 𝑀𝒊 + 𝑁𝒋, is conservative, and because 𝐶 is closed, you
can conclude that

∫ 𝑦 3 𝑑𝑥 + 3𝑥𝑦 2 𝑑𝑦 = 0.
𝐶

Theorem: (Line Integral For Area)

If 𝑅 is a plane region bounded by a piecewise smooth simple closed curve 𝐶, oriented
counterclockwise , then the area of 𝑅 is given by
1
𝐴= ∫ 𝑥 𝑑𝑦 − 𝑦 𝑑𝑥 .
2
𝐶

Example 5.29: Using a line integral find the area of the ellipse
𝑥2 𝑦2
+ = 1.
𝑎2 𝑏 2
Solution : we can induce a counterclockwise orientation to yhe elliptical path by letting
x = acost and y = bsint, 0 ≤ t ≤ 2π.
So the area is
2π
1 1
A = ∫ xdy − ydx = ∫ [(acost)(bcot)dt − (bsint)(−asint)dt]
2 2
C 0
2π
ab ab 2π
= ∫ (cos2 t + sin2 t)dt = [t] | = πab.
2 2 0
0

Alternative Forms of Green’s Theorem

This section concludes with the derivation of two vector forms of Green’s Theorem for regions
in the plane . The extension of these vector forms to three dimensions is the basis for the
discussion in the remaining sections of this chapter . If 𝑭 is a vector field in the plane, we can
write
𝑭(𝑥, 𝑦, 𝑧) = 𝑀𝒊 + 𝑁𝒋 + 0𝒌

So that the curl of 𝑭 is given by

Badri A, Moges B. and Teklebrhan B. 177 AKU

 APPLIED MATHEMATICS III

𝒊 𝒋 𝒌
𝜕 𝜕 𝜕
| |
𝑐𝑢𝑟𝑙 𝑭 = 𝛁 × 𝑭 = 𝜕𝑥 𝜕𝑦 𝜕𝑧
| |
𝑀 𝑁 0

𝜕𝑁 𝜕𝑀 𝜕𝑁 𝜕𝑀
= − 𝜕𝑧 𝒊 + 𝒋 + ( 𝜕𝑥 − )𝒌
𝜕𝑧 𝜕𝑦

Consequently,
𝜕𝑁 𝜕𝑀 𝜕𝑁 𝜕𝑀
(𝑐𝑢𝑟𝑙 𝑭) ∙ 𝒌 = [− 𝒊+ 𝒋+( − ) 𝒌] ∙ 𝒌
𝜕𝑧 𝜕𝑧 𝜕𝑥 𝜕𝑦
𝜕𝑁 𝜕𝑀
= − .
𝜕𝑥 𝜕𝑦

With appropraite conditions on 𝑭, 𝐶, and 𝑅, we can write Green’s Theorem in the vector form
𝜕𝑁 𝜕𝑀
∫ 𝑭 ∙ 𝑑𝒓 = ∬ ( − ) 𝑑𝐴
𝜕𝑥 𝜕𝑦
𝑅 𝑹

= ∬𝑅(𝑐𝑢𝑟𝑙 𝑭) ∙ 𝒌 𝑑𝐴 First alternative form

The extension of this vector form of Green’s Theorem to surfaces in space produces Stokes’s
Theorem,we will discusse in the next section.
For the second vector form of Green’s , assume the same conditions for 𝑭, 𝐶 and 𝑅.
Using the arc length parameter 𝑠 for 𝐶 , we have 𝒓(𝑠) = 𝑥(𝑠)𝒊 + 𝑦(𝑠)𝒋. So, a unit tangent
vector 𝑻 to curve 𝐶 is given by 𝒓′ (𝑠) = 𝑻 = 𝑥 ′ (𝑠)𝒊 + 𝑦 ′ (𝑠)𝒋. Using the figure below

figure
𝑻 = 𝑐𝑜𝑠𝜃𝒊 + 𝑠𝑖𝑛𝜃𝒋
𝜋 𝜋
𝒏 = cos (0 + ) 𝒊 + sin(0 + )𝒋 = −𝑠𝑖𝑛𝜃𝒊 + 𝑐𝑜𝑠𝜃𝒋
2 2
𝑵 = 𝑠𝑖𝑛𝜃𝒊 − 𝑐𝑜𝑠𝜃𝒋

Badri A, Moges B. and Teklebrhan B. 178 AKU

 APPLIED MATHEMATICS III

We can see that the outward unit normal vector 𝑵 can then be written as
𝑵 = 𝑦 ′ (𝑠)𝒊 + 𝑥′(𝑠)𝒋
Consequently, for 𝑭(𝑥, 𝑦) = 𝑴𝒊 + 𝑵𝒋, we can apply Green’s Theorem to obtain
𝑏

∫ 𝑭. 𝑵𝑑𝑠 = ∫(𝑀𝒊 + 𝑁𝒋). (𝑦 ′ (𝑠)𝒊 − 𝑥′(𝑠)𝒋) 𝑑𝑠

𝐶 𝑎
𝑏
𝑑𝑦 𝑑𝑥
= ∫(𝑀 − 𝑁 )𝑑𝑠 = ∫ 𝑀𝑑𝑦 − 𝑁𝑑𝑥 = ∫ −𝑁𝑑𝑥 + 𝑀𝑑𝑦
𝑑𝑠 𝑑𝑠
𝑎 𝐶 𝐶
𝜕𝑀 𝜕𝑁
= ∬𝑅( 𝜕𝑥 + 𝜕𝑦 )𝑑𝐴 Green’s Theorem

= ∬ 𝑑𝑖𝑣 𝑭 𝑑𝐴.
𝑅

Therefore,
∫𝐶 𝑭 ∙ 𝑵 𝑑𝑠 = ∬𝑅 𝑑𝑖𝑣 𝑭 𝑑𝐴. Second alternative form
The extension of this form to three dimensions is called the Divergence Theorem.

Exercises 3.3

I. Verify Green’s Theorem by evaluating both integrals

𝜕𝑁 𝜕𝑀
∫ 𝑦 2 𝑑𝑥 + 𝑥 2 𝑑𝑦 = ∬( − )𝑑𝐴
𝜕𝑥 𝜕𝑦
𝐶 𝑅

For the given path

1. 𝐶: boundary of the region lying between the graphs of 𝑦 = 𝑥 and 𝑦 = 𝑥 2
2. C: boundary of the region laying between the graphs of 𝑦 = 𝑥 and 𝑦 = √𝑥
3. 𝐶: square with vertices (0, 0), (1, 0), (1, 1), (0, 1)
4. 𝐶: rectangle with vertices (0, 0), (3, 0), (3, 4) and (0, 4)
I. Use Green’s Theorem to evaluate the integral

∫(𝑦 − 𝑥)𝑑𝑥 + (2𝑥 − 𝑦) 𝑑𝑦

𝐶

for the given path.

1.𝐶: boundary of the region laying between the graphs of 𝑦 = 𝑥 and 𝑦 = 𝑥 2 − 2𝑥
2. 𝐶: 𝑥 = 2𝑐𝑜𝑠𝜃, 𝑦 = 𝑠𝑖𝑛𝜃

Badri A, Moges B. and Teklebrhan B. 179 AKU

 APPLIED MATHEMATICS III

3. 𝐶: boundary of the region lying inside the rectangle bounded by 𝑥 = −5,

𝑥 = 5, 𝑦 = −3 and 𝑦 = 3, and outside the square bounded by
𝑥 = −1, 𝑥 = 1 , and 𝑦 = 1
4.boundary of the region laying inside the semicircle
𝑦 = √25 − 𝑥 2 and outside the semicircle 𝑦 = √9 − 𝑥 2
III. Use Green’s Theorem to evaluate the line integral
1. ∫𝐶 2𝑥𝑦𝑑𝑥 + (𝑥 + 𝑦) 𝑑𝑦
C: boundary of the region lying between the graphs of 𝑦 = 0 and 𝑦 = 1 − 𝑥 2
2. ∫𝐶 𝑦 2 𝑑𝑥 + 𝑥𝑦 𝑑𝑦

𝐶: boundary of the region lying between the graphs of 𝑦 = 0. 𝑦 = √𝑥, and 𝑥 = 9

5.6. Surface Integral; Divergence Theorem of Gauss; Applications;

5.6.1. Surface Integrals

Section Objectives:

At the end of this subtopic, students will be able to:

 Evaluate a surface integral as a double integral.

 Evaluate a surface integral for a parametric surface.
 Determine the orientation of a surface.
 Understand the concept of a flux integral.

1.6.1. Surface Integrals

Definition: Let S be a surface given by 𝑧 = 𝑔(𝑥, 𝑦) and let 𝑅 be its projection on to the 𝑥𝑦 −
plane. Suppose that 𝑔, 𝑔𝑥 , 𝑎𝑛𝑑 𝑔𝑦 are continuous at all points in R and that 𝑓 is defined on S.
𝑛

∬ 𝑓(𝑥, 𝑦, 𝑧) 𝑑𝑠 = lim ∑ 𝑓(𝑥𝑖 , 𝑦𝑖 , 𝑧𝑖 )∆𝑆𝑖

‖∆‖→0
𝑆 𝑖=1

Badri A, Moges B. and Teklebrhan B. 180 AKU

 APPLIED MATHEMATICS III

2
Where ∆𝑆𝑖 ≈ √1 + [𝑔𝑥 (𝑥𝑖 , 𝑦𝑖 )]2 + [𝑔𝑦 (𝑥𝑖 , 𝑦𝑖 )] ∆𝐴𝑖 and surface area of 𝑓 at (𝑥𝑖 , 𝑦𝑖 , 𝑧𝑖 ) and from

the sum ∑𝑛𝑖=1 𝑓(𝑥𝑖 , 𝑦𝑖 , 𝑧𝑖 )∆𝑆𝑖 provided the limit of this sum as ‖∆‖ approaches 0 exists, then
it is called a surface integral of 𝑓 over 𝑆

Fig. 5.6.1.1

Theorem5.6.1: (Evaluating a Surface Integral)

Let 𝑆 be a surface with equation 𝑧 = 𝑔(𝑥, 𝑦) and let 𝑅 be its projection on to the 𝑥𝑦 − plane. If
𝑔, 𝑔𝑥 , 𝑎𝑛𝑑 𝑔𝑦 are continuous at all points in R and that 𝑓 is defined on S.

Then the surface integral of 𝑓 over 𝑆 is

2
∬ 𝑓(𝑥, 𝑦, 𝑧) 𝑑𝑠 = ∬ 𝑓(𝑥, 𝑦, 𝑔(𝑥, 𝑦))√1 + [𝑔𝑥 (𝑥, 𝑦)]2 + [𝑔𝑦 (𝑥, 𝑦)] 𝑑𝐴
𝑆 𝑅

Remark1: If 𝑆 is the graph of 𝑦 = 𝑔(𝑥, 𝑧) and 𝑅 is its projection on to the 𝑥𝑧 − plane, then

∬ 𝑓(𝑥, 𝑦, 𝑧) 𝑑𝑠 = ∬ 𝑓(𝑥, 𝑔(𝑥, 𝑧), 𝑧)√1 + [𝑔𝑥 (𝑥, 𝑧)]2 + [𝑔𝑧 (𝑥, 𝑧)]2 𝑑𝐴
𝑆 𝑅

Remark2: If 𝑆 is the graph of 𝑥 = 𝑔(𝑦, 𝑧) and 𝑅 is its projection on to the 𝑦𝑧 − plane, then

2
∬𝑆 𝑓(𝑥, 𝑦, 𝑧) 𝑑𝑠 = ∬𝑅 𝑓(𝑔(𝑦, 𝑧), 𝑦, 𝑧)√1 + [𝑔𝑦 (𝑦, 𝑧)] + [𝑔𝑧 (𝑦, 𝑧)]2 𝑑𝐴

Example 5.30: Evaluate the surface integral

∬(𝑦 2 + 2𝑦𝑧) 𝑑𝑠
𝑆

Where 𝑆 is the first octant portion of the plane 2𝑥 + 𝑦 + 2𝑧 = 6 Fig. 5.6.1.2

Badri A, Moges B. and Teklebrhan B. 181 AKU

 APPLIED MATHEMATICS III

1 1
Solution: begin by writing 𝑆 as 𝑧 = 2 (6 − 2𝑥 − 𝑦) and 𝑔(𝑥, 𝑦) = 2 (6 − 2𝑥 − 𝑦)

1
Using the partial derivatives 𝑔𝑥 (𝑥, 𝑦) = −1 and 𝑔𝑦 (𝑥, 𝑦) = − 2,

2 1 3
you can write √1 + [𝑔𝑥 (𝑥, 𝑦)]2 + [𝑔𝑦 (𝑥, 𝑦)] = √1 + 1 + 4 = 2

Using the theorem5.6.1 we have

2
∬(𝑦 2 + 2𝑦𝑧) 𝑑𝑠 = ∬ 𝑓(𝑥, 𝑦, 𝑔(𝑥, 𝑦))√1 + [𝑔𝑥 (𝑥, 𝑦)]2 + [𝑔𝑦 (𝑥, 𝑦)] 𝑑𝐴
𝑆 𝑅

1 3
= ∬𝑅 [𝑦 2 + 2𝑦 (2) (6 − 2𝑥 − 𝑦)] (2) 𝑑𝐴

3 2(3−𝑥) 3 3 3 243
= 3 ∫0 ∫0 𝑦(3 − 𝑥)𝑑𝑦𝑑𝑥 = 6 ∫0 (3 − 𝑥)3 𝑑𝑥 = [− 2 (3 − 𝑥)3 ] =
0 2

Example 5.31: Evaluate the surface integral ∬𝑆(𝑥 + 𝑧) 𝑑𝑠

Where 𝑆 is the first- octant portion of the cylinder 𝑦 2 + 𝑧 2 = 9

between 𝑥 = 0 and 𝑥 = 4, as shown in Fig. 5.6.1.3

Solution: project 𝑆on to the 𝑥𝑦 − plane so that 𝑧 = 𝑔(𝑥, 𝑦) = √9 − 𝑦 2 , Fig. 5.6.1.3

2 −𝑦 3
and obtain √1 + [𝑔𝑥 (𝑥, 𝑦)]2 + [𝑔𝑦 (𝑥, 𝑦)] = √1 + ( )2 = figure
√9−𝑦 2 √9−𝑦2

Now theorem5.6.1 does not apply directly, because 𝑔𝑦 is not continuous when 𝑦 = 3.
however, you can apply theorem 5.6.1 for 0 ≤ 𝑏 < 3 and then take the limit as 𝑏 approaches 3.
As follows

𝑏 4
3
∬(𝑥 + 𝑧) 𝑑𝑠 = lim− ∫ ∫ (𝑥 + √9 − 𝑦 2 ) 𝑑𝑥𝑑𝑦
𝑏→3 √9 − 𝑦 2
𝑆 0 0

Badri A, Moges B. and Teklebrhan B. 182 AKU

 APPLIED MATHEMATICS III

𝑏 4 𝑏 4
𝑥 𝑥2
= lim− 3 ∫ ∫ ( + 1) 𝑑𝑥𝑑𝑦 = lim−3 ∫ [ + 𝑥] 𝑑𝑦
𝑏→3 √9 − 𝑦 2 𝑏→3 2√9 − 𝑦 2
0
0 0 0

𝑏
8 𝑦𝑏
= lim− 3 ∫ ( + 4) 𝑑𝑦 = lim− 3 [4𝑦 + 8 𝑎𝑟𝑐 sin ]
𝑏→3 √9 − 𝑦 2 𝑏→3 30
0

𝑏 𝜋
= lim− 3(4𝑏 + 8 𝑎𝑟𝑐 sin ) = 36 + 24 ( ) = 36 + 12𝜋
𝑏→3 3 2

Parametric Surface and Surface Integrals

For a surface 𝑆 given by the vector-valued function

𝑟(𝑢, 𝑣) = 𝑥(𝑢, 𝑣)𝑖 + 𝑦(𝑢, 𝑣)𝑗 + 𝑧(𝑢, 𝑣)𝑘 is the parametric surface defined over a region D in
the 𝑢𝑣 −plane . you can show that the surface of 𝑓(𝑥, 𝑦, 𝑧)over S is given by

∬ 𝑓(𝑥, 𝑦, 𝑧) 𝑑𝑆 = ∬ 𝑓(𝑥(𝑢, 𝑣), 𝑦(𝑢, 𝑣), 𝑧(𝑢, 𝑣))‖𝑟𝑢 (𝑢, 𝑣) × 𝑟𝑣 (𝑢, 𝑣) ‖𝑑𝐴
𝑆 𝑅

Note the similarity to a line integral over a space curve 𝐶

∬ 𝑓(𝑥, 𝑦, 𝑧) 𝑑𝑠 = ∫ 𝑓(𝑥(𝑡), 𝑦(𝑡), 𝑧(𝑡))‖𝑟 ′ (𝑡)‖𝑑𝑡

𝑆 𝑎

Note: 𝑑𝑠 𝑎𝑛𝑑 𝑑𝑆 can be written as 𝑑𝑠 = ‖𝑟 ′ (𝑡)‖𝑑𝑡 and 𝑑𝑆 = ‖𝑟𝑢 (𝑢, 𝑣) × 𝑟𝑣 (𝑢, 𝑣) ‖𝑑𝐴.

Example 5.32: Evaluating the Surface integral

∬(𝑥 + 𝑧) 𝑑𝑆
𝑆

Where 𝑆 is the first- octant portion of the cylinder

𝑦 2 + 𝑧 2 = 9 between 𝑥 = 0 and 𝑥 = 4, as shown

in Fig.5.5.1.4 by parametric form.

Fig.5.5.1.4

Badri A, Moges B. and Teklebrhan B. 183 AKU

 APPLIED MATHEMATICS III

Solution: in parametric form, the surface is given by 𝑟(𝑥, 𝜃) = 𝑥𝑖 + 3 cos 𝜃 𝑗 + 3 sin 𝜃 𝑘

𝜋
Where 0 ≤ 𝑥 ≤ 4 𝑎𝑛𝑑 0 ≤ 𝜃 ≤ to evaluate the surface integral in parametric form we have
2

𝑟𝑥 = 𝑖 and 𝑟𝜃 = −3 sin 𝜃 𝑗 + 3 cos 𝜃 𝑘

𝑖 𝑗 𝑘
𝑟𝑥 × 𝑟𝜃 = |1 0 0 | = −3 cos 𝜃 𝑗 − 3 sin 𝜃 𝑘
0 −3 sin 𝜃 3 cos 𝜃

‖𝑟𝑥 × 𝑟𝜃 ‖ = √9 cos 2 𝜃 + 9 sin2 𝜃 = 3

So, the surface integral can be evaluated as follows:

𝜋
4 2

∬(𝑥 + 3 sin 𝜃)3 𝑑𝐴 = ∫ ∫(3𝑥 + 9 sin 𝜃)𝑑𝜃𝑑𝑥

𝑆 0 0

4
𝜋
= ∫[3𝑥𝜃 − 9 cos 𝜃 𝑗]02 𝑑𝑥
0

4 3𝜋
= ∫0 ( 2 𝑥 + 9) 𝑑𝑥]

3𝜋 4
= [ 2 𝑥 + 9] = 12𝜋 + 36
0

Orientation of a Surface

Unit normal vectors are used to induce an orientation to a surface 𝑠 in space. A surface is called
orientable if a unit normal vector 𝑁 can be defined at every nonboundary point of 𝑆 in such away
that the noprmal vector vary continuously over the surface 𝑆. If this is possible, 𝑆 is called an
oriented surface.

Most common surfaces, such as sphers, paraboloids, ellipses, and planes, are orientable.

Moreover, for an orientable surface, the gradient vector provides a convenient way to find a unit
normal vector. That is, for an orientable surface 𝑆 given by 𝑧 = 𝑔(𝑥, 𝑦)

Badri A, Moges B. and Teklebrhan B. 184 AKU

 APPLIED MATHEMATICS III

Let 𝐺(𝑥, 𝑦, 𝑧) = 𝑧 − 𝑔(𝑥, 𝑦) then 𝑆 can be either the unit normal vector

∇𝐺(𝑥, 𝑦, 𝑧) −𝑔𝑥 (𝑥, 𝑦)𝑖 − 𝑔𝑦 (𝑥, 𝑦)𝑗 + 𝑘

𝑁= =
‖∇𝐺(𝑥, 𝑦, 𝑧)‖ 2
√1 + [𝑔𝑥 (𝑥, 𝑦)]2 + [𝑔𝑦 (𝑥, 𝑦)]

𝑆 is oriented in a upward direction .

Or the unit normal vector Fig. 5.6.1.5

−∇𝐺(𝑥,𝑦,𝑧) 𝑔𝑥 (𝑥,𝑦)𝑖+𝑔𝑦 (𝑥,𝑦)𝑗−𝑘

𝑁 = ‖∇𝐺(𝑥,𝑦,𝑧)‖ = 2
√1+[𝑔𝑥 (𝑥,𝑦)]2 +[𝑔𝑦 (𝑥,𝑦)]

𝑆 is oriented in a downward direction.

Fig.5.6.1.6

If the smooth orientable surface 𝑆 is given in parametric form by 𝑟(𝑢, 𝑣) = 𝑥(𝑢, 𝑣)𝑖 +
𝑦(𝑢, 𝑣)𝑗 + 𝑧(𝑢, 𝑣)𝑘

𝑟 ×𝑟
the unit normal vectors are given by 𝑁 = ‖𝑟𝑢 ×𝑟𝑣‖
𝑢 𝑣

Note: suppose that the orientable surface is given by 𝑦 = 𝑔(𝑥, 𝑧) 𝑜𝑟 𝑥 = 𝑔(𝑦, 𝑧). then you can
use the gradient vector ∇𝐺(𝑥, 𝑦, 𝑧) = −𝑔𝑥 (𝑥, 𝑧)𝑖 + 𝑗 − 𝑔𝑧 (𝑥, 𝑧)𝑘 , 𝐺(𝑥, 𝑦, 𝑧) = 𝑦 − 𝑔(𝑥, 𝑧)

Or ∇𝐺(𝑥, 𝑦, 𝑧) = 𝑖 − 𝑔𝑦 (𝑦, 𝑧)𝑗 − 𝑔𝑧 (𝑦, 𝑧)𝑘 , 𝐺(𝑥, 𝑦, 𝑧) = 𝑥 − 𝑔(𝑦, 𝑧)

to oriented the surface.

Flux Integrals

Let ∆𝑆 be the area of a small patch of the surface 𝑆 over which 𝐹 is

nearly constant. Then the amount of fluid crossing this region per unit

of time is approximated by the volume of the column of height 𝐹⦁𝑁,

Badri A, Moges B. and Teklebrhan B. 185 AKU

 APPLIED MATHEMATICS III

as shown in fig.5.6.7. That is ∆𝑉 = (ℎ𝑒𝑖𝑔ℎ𝑡)(𝑎𝑟𝑒𝑎 𝑜𝑓 𝑏𝑎𝑠𝑒) = (𝐹⦁𝑁)∆𝑆 Fig. 5.6.1.7

Consequently, the volume of 𝑆 fluid crossing the surface 𝑆 per unit of time (called the flux of 𝐹
across 𝑆) is given by the surface integral in the following definition.

Definition of Flux Integral

Let 𝐹(𝑥, 𝑦, 𝑧) = 𝑀𝑖 + 𝑁𝑗 + 𝑃𝑘, where M, N and P have continuous first partial derivatives on
the surface 𝑆 oriented by a unit normal vector N. the flux integral of F across 𝑆 is given by

∬ 𝐹⦁𝑁 𝑑𝑆
𝑆

Geometrically, a flux integral is the surface integral over 𝑆 of the normal component of F. if
𝜌(𝑥, 𝑦, 𝑧) is the density of the fluid at (𝑥, 𝑦, 𝑧), the flux integral

∬ 𝜌𝐹⦁𝑁 𝑑𝑆
𝑆

represents the mass of the fluid flowing across 𝑆 per unit of time. To evaluate a flux integral for a
surface given by 𝑧 = 𝑔(𝑥, 𝑦), let 𝐺(𝑥, 𝑦, 𝑧) = 𝑧 − 𝑔(𝑥, 𝑦) then, 𝑁𝑑𝑆 can be written as follows.

∇𝐺(𝑥, 𝑦, 𝑧) ∇𝐺(𝑥, 𝑦, 𝑧) 2
𝑁𝑑𝑆 = 𝑑𝑆 = √1 + [𝑔𝑥 (𝑥, 𝑦)]2 + [𝑔𝑦 (𝑥, 𝑦)] 𝑑𝐴
‖∇𝐺(𝑥, 𝑦, 𝑧)‖ 2
√1 + [𝑔𝑥 (𝑥, 𝑦)]2 + [𝑔𝑦 (𝑥, 𝑦)]

= ∇𝐺(𝑥, 𝑦, 𝑧)𝑑𝐴

Theorem 5.6.2: (Evaluating a Flux Integral)

Let 𝑆 be an oriented surface given by 𝑧 = 𝑔(𝑥, 𝑦) and let 𝑅 be its projection onto the 𝑥𝑦 −plane.

∬ 𝐹⦁𝑁 𝑑𝑆 = ∬ 𝐹⦁[−𝑔𝑥 (𝑥, 𝑦)𝑖 − 𝑔𝑦 (𝑥, 𝑦)𝑗 + 𝑘] 𝑑𝐴

𝑆 𝑅

∬ 𝐹⦁𝑁 𝑑𝑆 = ∬ 𝐹⦁[𝑔𝑥 (𝑥, 𝑦)𝑖 + 𝑔𝑦 (𝑥, 𝑦)𝑗 − 𝑘] 𝑑𝐴

𝑆 𝑅

Badri A, Moges B. and Teklebrhan B. 186 AKU

 APPLIED MATHEMATICS III

For the first integral, the surface is oriented upward and for the second integral, the surface is

oriented downward

Example 5.33: using a flux integral to find the rate of mass flow

Let 𝑆 be the portion of the paraboloid 𝑧 = 𝑔(𝑥, 𝑦) = 4 − 𝑥 2 − 𝑦 2 lying

above the 𝑥𝑦 −plane, oriented by an upward unit normal vector,

as shown in Fig.5.6.1.8. A fluid of constant density 𝜌 is following through

the surface 𝑆 according to the vector field

𝐹(𝑥, 𝑦, 𝑧) = 𝑥𝑖 + 𝑦𝑗 + 𝑧𝑘. Find the rate of mass flow through 𝑆. Fig.5.6.1.8

Solution: Begin by computing the partial derivatives of 𝑔.

𝑔𝑥 (𝑥, 𝑦) = −2𝑥 𝑎𝑛𝑑 𝑔𝑦 (𝑥, 𝑦) = −2𝑦

The rate of mass flow through the surface 𝑆 is

∬ 𝜌𝐹⦁𝑁 𝑑𝑆 = 𝜌 ∬ 𝐹⦁[−𝑔𝑥 (𝑥, 𝑦)𝑖 − 𝑔𝑦 (𝑥, 𝑦)𝑗 + 𝑘] 𝑑𝐴

𝑆 𝑅

= 𝜌 ∬𝑅[𝑥𝑖 + 𝑦𝑗 + (4 − 𝑥 2 − 𝑦 2 )𝑘]⦁(2𝑥𝑖 + 2𝑦𝑗 + 𝑘) 𝑑𝐴

= 𝜌 ∬𝑅[2𝑥 2 + 2𝑦 2 + (4 − 𝑥 2 − 𝑦 2 )] 𝑑𝐴

= 𝜌 ∬𝑅[(4 + 𝑥 2 + 𝑦 2 )] 𝑑𝐴

2𝜋 2
= 𝜌 ∫0 ∫0 (4 + 𝑟 2 )𝑟 𝑑𝑟𝑑𝜃 (∵ 𝑝𝑜𝑙𝑎𝑟 𝑐𝑜𝑜𝑟𝑑𝑖𝑛𝑎𝑡𝑒𝑠)

2𝜋
= 𝜌 ∫0 12 𝑑𝜃 = 24𝜋𝜌

Example 5.34: Evaluating the Surface integral

Compute the flux of water through the

parabolic cylinder 𝑆: 𝑦 = 𝑥 2 , 0 ≤ 𝑥 ≤ 2, 0 ≤ 𝑧 ≤ 3

Badri A, Moges B. and Teklebrhan B. 187 AKU

 APPLIED MATHEMATICS III

If the velocity vector is 𝑣 = 𝐹 = (3𝑧 2 , 6,6𝑧𝑥) Fig.5.6.1.9

Solution: Writing 𝑥 = 𝑢 and 𝑧 = 𝑣, we have𝑦 = 𝑥 2 = 𝑢2 .

Hence a representation of S is

𝑆: 𝑟 = [𝑢, 𝑢2 , 𝑣] (0 ≤ 𝑢 ≤ 2,0 ≤ 𝑣 ≤ 3).

By differentiation and by the definition of the cross product,

𝑁 = 𝑟𝑢 × 𝑟𝑣 = [1,2𝑢, 0] × [0,0,1] = [2𝑢, −1,0].

On S, writing simply 𝐹(𝑠) for 𝐹[𝑟(𝑢, 𝑣)] , we have 𝐹(𝑠) = (3𝑣 2 , 6,6𝑢𝑣).

Hence 𝐹(𝑠)⦁𝑁 = 6𝑢𝑣 2 − 6,

By define the surface integral

3 2 3

∬ 𝐹⦁𝑛 𝑑𝐴 = ∫ ∫(6𝑢𝑣 2 − 6) 𝑑𝑢𝑑𝑣 = ∫[3𝑢2 𝑣 2 − 6𝑢]20 𝑑𝑣

𝑆 0 0 0

3
= ∫0 (12𝑣 2 − 12)𝑑𝑣 = [4𝑣 3 − 12𝑣]3𝑣=0 = 108 − 36 = 72

Example 5.35: Find the Surface integral when 𝐹 = (𝑥 2 , 0,3𝑦 2 ) and 𝑆 is the portion of the plane
𝑥 + 𝑦 + 𝑧 = 1 in the first octant (Fig.5.6.1.10)

Solution: writing 𝑥 = 𝑢 and 𝑦 = 𝑣, we have 𝑧 = 1 − 𝑥 − 𝑦 = 1 − 𝑢 − 𝑣.

Hence we can represent the plane 𝑥 + 𝑦 + 𝑧 = 1 in the form

𝑟(𝑢, 𝑣) = [𝑢, 𝑣, 1 − 𝑢 − 𝑣].

Fig.5.6.1.10

We obtain the first-octant portion S of this plane by restricting 𝑥 = 𝑢 and 𝑦 = 𝑣to the projection
R of S in the 𝑥𝑦-plane. R is the triangle bounded by the two coordinate axes and the straight line
𝑥 + 𝑦 = 1, obtained from 𝑥 + 𝑦 + 𝑧 = 1 by setting 𝑧 = 0. Thus0 ≤ 𝑥 ≤ 1 − 𝑦, 0 ≤ 𝑦 ≤ 1.

By inspection or by differentiation, 𝑁 = 𝑟𝑢 × 𝑟𝑣 = [1,0, −1] × [0,1, −1] = [1,1,1]

Badri A, Moges B. and Teklebrhan B. 188 AKU

 APPLIED MATHEMATICS III

Hence 𝐹(𝑠)⦁𝑁 = [𝑢2 , 0, 3𝑣 2 ]⦁[1,1,1] = 𝑢2 + 3𝑣 2

1 1−𝑣

∬ 𝐹⦁𝑛 𝑑𝐴 = ∬(𝑢2 + 3𝑣 2 )𝑑𝑢 𝑑𝑣 = ∫ ∫ (𝑢2 + 3𝑣 2 ) 𝑑𝑢𝑑𝑣

𝑆 𝑅 0 0
1
2
1 1
= ∫ [ (1 − 𝑣)3 + 3𝑣 2 (1 − 𝑣)] 𝑑𝑣 =
3 0 3
0

Exercises 3.3

1. Find the Surface integral for the given data.

a. 𝐹 = [−𝑥 2 , 𝑦 2 , 0], 𝑆: 𝑟 = [𝑢, 𝑣, 3𝑢 − 2𝑣], 0 ≤ 𝑢 ≤ 1.5 , −2 ≤ 𝑣 ≤ 2

b. 𝐹 = [𝑒 𝑦 , 𝑒 𝑥 , 1] 𝑆: 𝑥 + 𝑦 + 𝑧 = 1, 𝑥 ≥ 0, 𝑦 ≥ 0 , 𝑧 ≥ 0
c. 𝐹 = [0, 𝑥, 0] 𝑆: 𝑥 2 + 𝑦 2 + 𝑧 2 = 1, 𝑥 ≥ 0, 𝑦 ≥ 0 , 𝑧 ≥ 0
d. 𝐹 = [0, sin 𝑦 , cos 𝑧] 𝑆 the cylinder 𝑥 = 𝑦 2 , where0 ≤ 𝑦 ≤ 𝜋⁄4 and 0 ≤ 𝑧 ≤ 𝑦

e. 𝐹 = [𝑦 2 𝑥 2 , 𝑧 2 ], 𝑆: 𝑧 = 4√𝑥 2 + 𝑦 2 , 0 ≤ 𝑧 ≤ 8, 𝑦 ≥ 0

2. Find the flux of F through S,∬𝑆 𝐹⦁𝑁 𝑑𝑆,where N is the upward unit normal vector to S

a. 𝐹(𝑥, 𝑦, 𝑧) = 3𝑧𝑖 − 4𝑗 + 𝑦𝑘 , 𝑠: 𝑧 = 1 − 𝑥 − 𝑦, 𝑓𝑟𝑖𝑠𝑡 𝑜𝑐𝑡𝑎𝑛𝑡

b. 𝐹(𝑥, 𝑦, 𝑧) = 𝑥𝑖 + 𝑦𝑗, 𝑠: 𝑧 = 6 − 3𝑥 − 2𝑦, 𝑓𝑟𝑖𝑠𝑡 𝑜𝑐𝑡𝑎𝑛𝑡
c. 𝐹(𝑥, 𝑦, 𝑧) = 𝑥𝑖 + 𝑦𝑗 + 𝑧𝑘 , 𝑠: 𝑧 = 1 − 𝑥 2 − 𝑦 2 , 𝑧 ≥ 0
d. 𝐹(𝑥, 𝑦, 𝑧) = 𝑥𝑖 + 𝑦𝑗 + 𝑧𝑘 , 𝑠: 𝑧 = 𝑥 2 + 𝑦 2 + 𝑧 2 , 𝑧 ≥ 𝑓𝑟𝑖𝑠𝑡 𝑜𝑐𝑡𝑎𝑛𝑡
e.

6.5.2. Divergence Theorem of Gauss

In this section we discuss another “big” integral theorem, the divergence theorem, which
transforms surface integrals into triple integrals. So let us begin with a review of the latter.

A triple integral is an integral of a function 𝑓 (𝑥, 𝑦, 𝑧) taken over a closed bounded, three-
dimensional region T in space. (Note that “closed” and “bounded” are defined in the same way
with “sphere” substituted for “circle”). Triple integrals can be evaluated by three successive
integrations. This is similar to the evaluation of double integrals by two successive integrations.

Badri A, Moges B. and Teklebrhan B. 189 AKU

 APPLIED MATHEMATICS III

Divergence Theorem of Gauss

Triple integrals can be transformed into surface integrals over the boundary surface of a region in
space and conversely. Such a transformation is of practical interest because one of the two kinds
of integral is often simpler than the other. It also helps in establishing fundamental equations in
fluid flow, heat conduction, etc., as we shall see. The transformation is done by the divergence
theorem, which involves the divergence of a vector function

𝐹 = [𝐹1 , 𝐹2 , 𝐹3 ] = 𝐹1 𝑖 + 𝐹2 𝑗 + 𝐹3 𝑘

𝜕𝐹1 𝜕𝐹2 𝜕𝐹3

Namely, 𝑑𝑖𝑣 𝐹 = + +
𝜕𝑥 𝜕𝑦 𝜕𝑧

Theorem: Divergence Theorem of Gauss

(Transformation between Triple and Surface Integrals)

Let T be a closed bounded region in space whose boundary is a piecewise smooth orientable
surface S. Let 𝐹 (𝑥, 𝑦, 𝑧) be a vector function that is continuous and has continuous first partial
derivatives in some domain containing T. Then

∭ 𝑑𝑖𝑣 𝐹 𝑑𝑣 = ∬ 𝐹⦁𝑛 𝑑𝐴
𝑇 𝑆

In components of 𝐹 = [𝐹1 , 𝐹2 , 𝐹3 ] and of the outer unit normal vector 𝑛 = [cos 𝛼 , cos 𝛽 , cos 𝛾]
of 𝑆. Now this becomes

𝜕𝐹1 𝜕𝐹2 𝜕𝐹3

∭( + + ) 𝑑𝑥𝑑𝑦𝑑𝑧 = ∬(𝐹1 cos 𝛼 + 𝐹2 cos 𝛽 + 𝐹3 cos 𝛾)𝑑𝐴
𝜕𝑥 𝜕𝑦 𝜕𝑧
𝑇 𝑆

= ∬𝑆(𝐹1 𝑑𝑦𝑑𝑧 + 𝐹2 𝑑𝑧𝑑𝑥 + 𝐹3 𝑑𝑥𝑑𝑦)

Badri A, Moges B. and Teklebrhan B. 190 AKU

 APPLIED MATHEMATICS III

Example 5.36: Evaluation of a surface integral by the Divergence Theorem

Evaluate 𝐼 = ∬𝑆(𝑥 3 𝑑𝑦𝑑𝑧 + 𝑥 2 𝑦𝑑𝑧𝑑𝑥 + 𝑥 2 𝑧𝑑𝑥𝑑𝑦) where 𝑆the closed surface

in this Fig. is consisting of the cylinder 𝑥 2 + 𝑦 2 = 𝑎2 (0 ≤ 𝑧 ≤ 𝑏)and

the disks 𝑧 = 0 and 𝑧 = 𝑏(𝑥 2 + 𝑦 2 ≤ 𝑎2 )

Fig.5.6.2.1

Solution: 𝐹1 = 𝑥 3 , 𝐹1 = 𝑥 2 𝑦, 𝐹1 = 𝑥 2 𝑧. Hence 𝑑𝑖𝑣 𝐹 = 3𝑥 2 + 𝑥 2 + 𝑥 2 = 5𝑥 2

The form of the surface suggests that we introduce polar coordinates 𝑟, 𝜃 defined by

𝑥 = 𝑟 cos 𝜃 , 𝑦 = 𝑟 sin 𝜃 (thus cylindrical coordinates 𝑟, 𝜃, 𝑧 ). Then the volume element is

𝑑𝑥 𝑑𝑦 𝑑𝑧 = 𝑟 𝑑𝑟 𝑑𝜃 𝑑𝑧, and we obtain

𝑏 2𝜋 𝑎

𝐼 = ∭ 5𝑥 2 𝑑𝑥𝑑𝑦𝑑𝑧 = ∫ ∫ ∫(5𝑥 2 cos 2 𝜃)𝑟𝑑𝑟𝑑𝜃𝑑𝑧

𝑇 𝑧=0 𝜃=0 𝑟=0

𝑏 2𝜋 𝑎4 𝑏 𝑎4 5𝜋
= 5 ∫𝑧=0 ∫𝜃=0 cos2 𝜃 𝑑𝜃𝑑𝑧 = 5 ∫𝑧=0 𝑑𝑧 = 𝑎4 𝑏.
4 4 4

Example 5.37: Evaluate ∬𝑆(7𝑥𝑖 − 𝑧𝑘)⦁𝑛𝑑𝐴 over the sphere 𝑆: 𝑥 2 + 𝑦 2 + 𝑧 2 = 4

Solution: 𝑑𝑖𝑣 𝐹 = 𝑑𝑖𝑣[7𝑥, 0, −𝑧] = 𝑑𝑖𝑣[7𝑥𝑖 − 𝑧𝑘] = 7 − 1 = 6

We can represent 𝑆 with 𝑎 = 2 and use 𝑛𝑑𝐴 = 𝑁𝑑𝑢𝑑𝑣

𝑆: 𝑟 = [2 cos 𝑣 cos 𝑢 , 2 cos 𝑣 sin 𝑢 , 2 sin 𝑢]

𝑟𝑢 = [−2 cos 𝑣 sin 𝑢 , 2 cos 𝑣 cos 𝑣 cos 𝑢 , 0]

𝑟𝑣 = [−2 sin 𝑣 cos 𝑢 , −2 sin 𝑣 sin 𝑢 , 2 cos 𝑣]

Then 𝑟𝑢 × 𝑟𝑣 = [4 cos 2 𝑣 cos 𝑢 , 4 cos2 𝑣 sin 𝑢 , 4 cos 𝑣 sin 𝑣]

Badri A, Moges B. and Teklebrhan B. 191 AKU

 APPLIED MATHEMATICS III

Now on 𝑆 we have 𝑥 = 2 cos 𝑣 cos 𝑢 , 𝑧 = 2 sin 𝑢 so that 𝐹 = [7𝑥, 0, −𝑧] becomes on 𝑆 𝐹(𝑆) =
[14 cos 𝑣 cos 𝑢 , 0, −2 sin 𝑣]

And 𝐹(𝑆)⦁𝑁 = (14 cos 𝑣 cos 𝑢)⦁4 cos 2 𝑣 cos 𝑢 + (−2 sin 𝑣)⦁4 cos 𝑣 sin 𝑣

= 56 cos 3 𝑣 cos2 𝑢 − 8 cos 𝑣 sin2 𝑣

On 𝑆 we have to integrate ver 𝑢 from 0 to 2𝜋. this gives

𝜋⦁56 cos3 𝑣 − 2𝜋⦁8 cos 𝑣 sin2 𝑣

The integral of cos 𝑣 sin2 𝑣 equals (sin2 𝑣)/3 and that of cos3 𝑣 = cos 𝑣 (1 − sin2 𝑣) equals
sin 𝑣 − (sin2 𝑣)/3 on 𝑆 we have −𝜋⁄2 ≤ 𝑣 ≤ 𝜋⁄2

2
So that by substituting these limits we get 56𝜋(2 − 2⁄3) − 16𝜋 ∙ 3 = 64𝜋

Exercises 3.3

Evaluate the surface integral ∬𝑆 𝐹⦁𝑛 𝑑𝐴 by the divergence theorem

1. 𝐹 = [𝑥 2 , 0, 𝑧 2 ] 𝑆 the surface of the box |𝑥| ≤ 1, |𝑦| ≤ 3, 0 ≤ 𝑧 ≤ 2

2. 𝐹 = [𝑒 𝑥 , 𝑒 𝑦 , 𝑒 𝑧 ] 𝑆 the surface of the cube |𝑥| ≤ 1, |𝑦| ≤ 1, |𝑧| ≤ 1
3. 𝐹 = [𝑥 3 − 𝑦 3 , 𝑦 3 − 𝑧 3 , 𝑧 3 − 𝑥 3 ] 𝑆 the surface of 𝑥 2 + 𝑦 2 + 𝑧 2 ≤ 25, 𝑧 ≥ 0
4. 𝐹 = [sin 𝑦 , cos 𝑥 , cos 𝑧] 𝑆 the surface of 𝑥 2 + 𝑦 2 +≤ 4, |𝑧| ≤ 2
5. 𝐹 𝑎𝑠 𝑖𝑛 𝑝𝑟𝑜𝑏. 4, S the surface of 𝑥 2 + 𝑦 2 +≤ 9, 0≤𝑧≤2

5.6.3. Applications of the Divergence Theorem

The divergence theorem has many important applications: In fluid flow, it helps characterize
sources and sinks of fluids. In heat flow, it leads to the heat equation. In potential theory, it gives
properties of the solutions of Laplace’s equation. In this section, we assume that the region T and
its boundary surface S are such that the divergence theorem applies.

Badri A, Moges B. and Teklebrhan B. 192 AKU

 APPLIED MATHEMATICS III

5.7 Stokes’s Theorem; Applications

Overview

In this subtopic we are going to learn that we can transform surface integrals into line integrals
and conversely, line integrals into surface integrals is called Stokes’s Theorem and we will see
examples.

Section Objectives:

At the end of this subtopic, students will be able to:

 Understand and use Stokes’s Theorem.

 Use curl to analyze the motion of a rotating liquid.

A second higher dimension analog of Green’s Theorem is called Stokes’s Theorem, after the
English mathematical physicist George Gabriel Stokes. In addition to making contributions to
physics, stokes worked with infinite series and differential equation, as well as with the
integration result presented in this section.

To complete our discussion on transforming integrals, that allows us to transform surface

integrals into line integrals and conversely, line integrals into surface integrals. It is called
Stokes’s Theorem, and it generalizes Green’s theorem in the plane.

Stokes’s Theorem gives the relationship between a surface integral over an oriented surface 𝑆
and a line integral along a closed space curve 𝐶 forming the boundary of , as shown

Fig.5.7.1.1. The positive direction along 𝐶 is counterclockwise relative to the normal vector 𝑁.
That is, if you imagine grasping the normal vector N with your right hand, with your thumb
pointing in

the direction of N, your fingers will point in the positive direction C, as shown Fig. 5.7.1.2.

Badri A, Moges B. and Teklebrhan B. 193 AKU

 APPLIED MATHEMATICS III

Fig. 5.7.1.1 Fig. 5.7.1.2

Theorem 5.7.1:( Stokes’s Theorem)

Let S be a piecewise smooth oriented surface with unit normal vector N, and let the boundary of
S be a piecewise smooth simple closed curve C. Let 𝐹(𝑥, 𝑦, 𝑧) be a continuous vector function
that has continuous first partial derivatives in a domain in space containing S. Then

∫𝐶 𝐹⦁ 𝑑𝑟 = ∬𝑆(𝑐𝑢𝑟𝑙 𝐹)⦁𝑁𝑑𝑠

𝑖 𝑗 𝑘
Where 𝑐𝑢𝑟𝑙 𝐹 = |𝜕⁄𝜕𝑥 𝜕⁄
𝜕𝑦
𝜕⁄ |
𝜕𝑧
𝐹1 𝐹2 𝐹3

OR

Badri A, Moges B. and Teklebrhan B. 194 AKU

 APPLIED MATHEMATICS III

Theorem 5.7.1: Stokes’s Theorem (Transformation between Surface and Line Integrals)

Let S be a piecewise smooth oriented surface in space and let the boundary of S be a piecewise smooth
simple closed curve C. Let 𝐹(𝑥, 𝑦, 𝑧) be a continuous vector function that has continuous first partial
derivatives in a domain in space containing S. Then

∬(𝑐𝑢𝑟𝑙 𝐹)⦁𝑛 𝑑𝐴 = ∮ 𝐹⦁𝑟 ′ (𝑠)𝑑𝑠 (∗)

𝑆 𝐶

Here n is a unit normal vector of S and, depending on 𝒏, the integration around C is taken in the sense
shown in Fig. 5.6.4. Furthermore, 𝑟 ′ = 𝑑𝑟⁄𝑑𝑠 is the unit tangent vector and 𝑠 the arc length of C. In
components, formula (*) becomes

𝜕𝐹 𝜕𝐹2 𝜕𝐹 𝜕𝐹3 𝜕𝐹 𝜕𝐹1

∬𝑆 [( 𝜕𝑦3 − 𝜕𝑧
) 𝑁1 + ( 𝜕𝑧1 − 𝜕𝑥
) 𝑁2 + ( 𝜕𝑥2 − 𝜕𝑦
) 𝑁3 ] 𝑑𝑢𝑑𝑣 = ∮𝐶(𝐹1 𝑑𝑥 + 𝐹2 𝑑𝑦 + 𝐹3 𝑑𝑧) (**)

Here, 𝐹 = [𝐹1 , 𝐹2 , 𝐹3 ], 𝑁 = [𝑁1 , 𝑁2 , 𝑁3 ], 𝑛𝑑𝑎 = 𝑁𝑑𝑢𝑑𝑣 , 𝑟 ′ 𝑑𝑠 = [𝑑𝑥, 𝑑𝑦, 𝑑𝑧], and R is the region
with boundary curve 𝐶̅ in the 𝑢𝑣-plane corresponding to S represented by 𝑟 (𝑢, 𝑣)

Example 5.38: Using Stokes’s Theorem evaluate

Let C be the oriented triangle lying in the plane 2𝑥 + 2𝑦 + 𝑧 = 6,

as shown in Fig. 5.7.1.3. Where 𝐹(𝑥, 𝑦, 𝑧) = −𝑦 2 𝑖 + 𝑧𝑗 + 𝑥𝑘

Solution: Using Stokes’s Theorem, benign by 𝑐𝑢𝑟𝑙 𝑜𝑓 𝐹

𝑖 𝑗 𝑘
𝜕
𝑐𝑢𝑟𝑙 𝐹 = | ⁄𝜕𝑥 𝜕⁄ 𝜕⁄ | i.e. Fig. 5.7.1.3
𝜕𝑦 𝜕𝑧
𝐹1 𝐹2 𝐹3

𝑖 𝑗 𝑘
𝑐𝑢𝑟𝑙 𝐹 = |𝜕⁄𝜕𝑥 𝜕⁄
𝜕𝑦
𝜕⁄ | = −𝑖 − 𝑗 + 2𝑦𝑘
𝜕𝑧
−𝑦 2 𝑧 𝑥

Considering 𝑧 = 6 − 2𝑥 − 2𝑦 = 𝑔(𝑥, 𝑦, ) you can use theorem 5.6.2 for an upward normal
vector to obtain

Badri A, Moges B. and Teklebrhan B. 195 AKU

 APPLIED MATHEMATICS III

∫ 𝐹⦁ 𝑑𝑟 = ∬(𝑐𝑢𝑟𝑙 𝐹)⦁𝑁𝑑𝑠
𝐶 𝑆

= ∬𝑆(−𝑖 − 𝑗 + 2𝑦𝑘)[−𝑔𝑥 (𝑥, 𝑦)𝑖 − 𝑔𝑦 (𝑥, 𝑦)𝑗 + 𝑘]𝑑𝐴

= ∬𝑅(−𝑖 − 𝑗 + 2𝑦𝑘)[2𝑖 + 2𝑗 + 𝑘]𝑑𝐴

3 3−𝑦
= ∫0 ∫0 (2𝑦 − 4) 𝑑𝑥𝑑𝑦

3
= ∫0 (−2𝑦 2 + 10𝑦 − 12)𝑑𝑦

𝟑
2𝑦 3
= [− + 5𝑦 2 − 12𝑦] = −𝟗
3 𝟎

Example 5.39: An application of curl

A liquid is swirling around in a cylindrical container of radius 2,

so that its described by the velocity field

𝐹(𝑥, 𝑦, 𝑧) = −𝑦√𝑥 2 + 𝑦 2 𝑖 + 𝑥√𝑥 2 + 𝑦 2 𝑗 as shown in fig .

Find ∬𝑆(𝑐𝑢𝑟𝑙 𝐹)⦁𝑁𝑑𝑠 where 𝑆 the upper surface Fig. 5.7.1.4

of the cylindrical container.

𝑖 𝑗 𝑘
Solution: the 𝑐𝑢𝑟𝑙 𝑜𝑓 𝐹 is given by 𝑐𝑢𝑟𝑙 𝐹 = |𝜕⁄𝜕𝑥 𝜕⁄
𝜕𝑦
𝜕⁄ | =
𝜕𝑧
𝐹1 𝐹2 𝐹3
𝑖 𝑗 𝑘
𝜕⁄ 𝜕⁄ 𝜕⁄
| 𝜕𝑥 𝜕𝑦 𝜕𝑧| = 3√𝑥 2 + 𝑦 2 𝑘.
−𝑦√𝑥 2 + 𝑦 2 𝑥√𝑥 2 + 𝑦 2 0

Letting 𝑁 = 𝑘, you have

∬(𝑐𝑢𝑟𝑙 𝐹)⦁𝑁𝑑𝑠 = ∬ 3√𝑥 2 + 𝑦 2 𝑑𝐴

𝑆 𝑅

Badri A, Moges B. and Teklebrhan B. 196 AKU

 APPLIED MATHEMATICS III

2𝜋 2 2𝜋 2𝜋

= ∫ ∫ 3𝑟)𝑟𝑑𝑟𝑑𝜃 = ∫ [𝑟 3 ]20 𝑑𝜃 = ∫ 8𝑑𝜃 = 16𝜋

0 0 0 0

Example 5.40: using Stokes’s Theorem evaluate where 𝐹 = [𝑦, 𝑧, 𝑥]

and 𝑆 the paraboloid 𝑧 = 𝑓(𝑥, 𝑦) = 1 − (𝑥 2 + 𝑦 2 ), 𝑧 ≥ 0.

Solution: The curve C, oriented as in Fig.5.7.1.4, Fig.5.7.1.4

is the circle 𝑟(𝑠) = [cos 𝑠 , sin 𝑠 , 0]. It’s unit tangent vector is 𝑟 ′ (𝑠) = [−sin 𝑠 , cos 𝑠 , 0] . The
function 𝐹 = [𝑦, 𝑧, 𝑥] on C

is 𝐹(𝑟(𝑠)) = [sin 𝑠 , 0, cos 𝑠]. Hence

2𝜋 2𝜋

∮ 𝐹⦁𝑑𝑟 = ∮ 𝐹(𝑟(𝑠))⦁𝑟 ′ (𝑠)𝑑𝑠 = ∮ [(sin 𝑠)(− sin 𝑠) + 0 + 0] 𝑑𝑠 = −𝜋

𝐶 0 0

We now consider the surface integral. We have 𝐹1 = 𝑦, 𝐹2 = 𝑧, 𝐹3 = 𝑥 , so that in (∗∗) we

obtain

𝑐𝑢𝑟𝑙 𝐹 = 𝑐𝑢𝑟𝑙 [𝐹1 , 𝐹2 , 𝐹3 ] = 𝑐𝑢𝑟𝑙 [𝑦, 𝑧, 𝑥] = [−1, −1, −1].

A normal vector of S is 𝑁 = 𝑔𝑟𝑎𝑑(𝑧 − 𝑓(𝑥, 𝑦)) = [2𝑥, 2𝑦, 1].

Hence (𝑐𝑢𝑟𝑙 𝐹)⦁𝑁 = −2𝑥 − 2𝑦 − 1.

Now 𝑛 𝑑𝐴 = 𝑁 𝑑𝑥𝑑𝑦 with 𝑥, 𝑦 instead of 𝑢, 𝑣. Using polar coordinates 𝑟, 𝜃 defined by𝑥 =

𝑟 cos 𝜃 , 𝑦 = 𝑟 sin 𝜃 and denoting the projection of S into the 𝑥𝑦 −plane by R, we thus obtain

∬(𝑐𝑢𝑟𝑙 𝐹)⦁𝑛 𝑑𝐴 = ∬(𝑐𝑢𝑟𝑙 𝐹)⦁𝑁𝑑𝑥 𝑑𝑦 = ∬(−2𝑥 − 2𝑦 − 1)𝑑𝑥 𝑑𝑦

𝑆 𝑅 𝑅
2𝜋 1

= ∫ ∫(−2𝑟(cos 𝜃 + sin 𝜃) − 1) 𝑑𝑟𝑑𝜃

𝜃=0 𝑟=0

2𝜋
2 1 1
= ∫ (− (cos 𝜃 + sin 𝜃) − )𝑑𝜃 = 0 + 0 − (2𝜋) = −𝜋
3 2 2
𝜃=0

Badri A, Moges B. and Teklebrhan B. 197 AKU

 APPLIED MATHEMATICS III

Example 5.41: Evaluation of a Line Integral by Stokes’s Theorem

Evaluate ∮𝐶 𝐹⦁𝑟 ′ 𝑑𝑠, where C is the circle 𝑥 2 + 𝑦 2 = 4, 𝑧 = −3 , oriented counterclockwise as

seen by a person standing at the origin, and, with respect to right-handed Cartesian coordinates,

Solution: As a surface S bounded by C we can take the plane circular disk 𝑥 2 + 𝑦 2 = 4 in the
plane 𝑧 = −3.

Then 𝑛 in Stokes’s theorem points in the positive 𝑧 −direction; thus 𝑛 = 𝑘 .

Hence (𝑐𝑢𝑟𝑙 𝐹)⦁𝑛 is simply the component of 𝑐𝑢𝑟𝑙 𝐹 in the positive 𝑧 −direction. Since 𝐹 with
𝑧 = −3 has the components 𝐹1 = 𝑦, 𝐹2 = −27𝑥, 𝐹3 = 3𝑦 3 , we thus obtain

𝜕𝐹1 𝜕𝐹2
(𝑐𝑢𝑟𝑙 𝐹)⦁𝑛 = − = −27 − 1 = −28.
𝜕𝑥 𝜕𝑦

Hence the formula in Stokes’s theorem now takes the form

𝜕𝐹1 𝜕𝐹2
∬( − ) 𝑑𝐴 = ∮ 𝐹1 𝑑𝑥 + 𝐹2 𝑑𝑦 = −28
𝜕𝑥 𝜕𝑦
𝑆 𝐶

Note: if 𝑐𝑢𝑟𝑙 𝐹 = 0 throuhot region R , the rotation of F about each unit normal N is 0.

5.8Line Integral Independent of Path

Path dependence of line integrals is practically and theoretically so important that we formulate it
as a theorem.

Definition: A line integral is written in the form of

∫𝐹(𝑟)⦁𝑑𝑟 = ∫(𝐹1 𝑑𝑥 + 𝐹2 𝑑𝑦 + 𝐹3 𝑑𝑧) (𝑑𝑟 = [𝑑𝑥, 𝑑𝑦, 𝑑𝑧] (1)

𝐶 𝐶

The line integral (1) is said to be path independent in a domain D in space if

for every pair of endpoints A, B in domain D, (1) has the same value for all paths

in D that begin at A and end at B.

Badri A, Moges B. and Teklebrhan B. 198 AKU

 APPLIED MATHEMATICS III

Theorem1: (Path Independence)

A line integral (1) with continuous𝐹1 , 𝐹2 , 𝐹3 in a domain D in space is path independent in D if

and only if 𝐹 = [𝐹1 , 𝐹2 , 𝐹3 ] is the gradient of some function 𝑓 in D,

𝜕𝑓 𝜕𝑓 𝜕𝑓
𝐹 = 𝑔𝑟𝑎𝑑 𝑓 thus 𝐹1 = 𝜕𝑥 , 𝐹2 = 𝜕𝑦 , 𝐹3 = (2)
𝜕𝑧

Then ∫𝐶(𝐹1 𝑑𝑥 + 𝐹2 𝑑𝑦 + 𝐹3 𝑑𝑧) = 𝑓(𝐵) − 𝑓(𝐴) where

Proof : We assume that (2) holds for some function 𝑓 in D and show that this implies path
independence. Let C be any path in D from any point A to any point B in D, given by𝑟(𝑡) =
[𝑥(𝑡), 𝑦(𝑡), 𝑧(𝑡)], where 𝑎 ≤ 𝑡 ≤ 𝑏. Then from (2), the chain rule

𝜕𝑓 𝜕𝑓 𝜕𝑓
∫(𝐹1 𝑑𝑥 + 𝐹2 𝑑𝑦 + 𝐹3 𝑑𝑧) = ∫ ( 𝑑𝑥 + 𝑑𝑦 + 𝑑𝑧)
𝐶 𝐶 𝜕𝑥 𝜕𝑦 𝜕𝑧

𝑏
𝜕𝑓 𝑑𝑥 𝜕𝑓 𝑑𝑦 𝜕𝑓 𝑑𝑧
= ∫( + + )
𝜕𝑥 𝑑𝑡 𝜕𝑦 𝑑𝑡 𝜕𝑧 𝑑𝑡
𝑎

𝑏
𝑑𝑓
=∫ 𝑑𝑡 = 𝑓[𝑥(𝑡), 𝑦(𝑡), 𝑧(𝑡)]𝑡=𝑏
𝑡=𝑎
𝜕𝑥
𝑎

= 𝑓(𝑥(𝑏), 𝑦(𝑏), 𝑧(𝑏)) − 𝑓(𝑥(𝑎), 𝑦(𝑎), 𝑧(𝑎)) = 𝑓(𝐵) − 𝑓(𝐴)

Example 5.42: Show that the integral ∫𝐶 𝐹(𝑟)⦁𝑑𝑟 = ∫𝐶(2𝑥𝑑𝑥 + 2𝑦𝑑𝑦 + 4𝑧𝑑𝑧) is path
independent in any domain in space and find its value in the integration from A: (0, 0, 0) to B:
(2, 2, 2).

Solution: 𝐹 = [2𝑥, 2𝑦, 4𝑧 ] = 𝑔𝑟𝑎𝑑 𝑓, where 𝑓 = 𝑥 2 + 𝑦 2 + 2𝑧 2 because

𝜕𝑓 𝜕𝑓 𝜕𝑓
= 2𝑥 = 𝐹1 , 𝜕𝑦 = 2𝑦 = 𝐹2 , 𝜕𝑧 = 4𝑧 = 𝐹3 . Hence the integral is independent of path according
𝜕𝑥

to Theorem 1, and gives𝑓(𝐵) − 𝑓(𝐴) = 𝑓(2,2,2) − 𝑓(0,0,0) = 4 + 4 + 8 = 16

Badri A, Moges B. and Teklebrhan B. 199 AKU

 APPLIED MATHEMATICS III

If you want to check this, use the most convenient path 𝑐: 𝑟(𝑡) = [𝑡, 𝑡, 𝑡], 0 ≤ 𝑡 ≤ 2 on
which𝐹(𝑟(𝑡)) = [2𝑡, 2𝑡, 4𝑡] so that 𝐹(𝑟(𝑡))⦁𝑟 ′ (𝑡) = 2𝑡 + 2𝑡 + 4𝑡 = 8𝑡 and integration from 0
2
2 8 𝑡2
to 2 gives ∫0 8𝑡 𝑑𝑡 = [ ] = 16
2 0

Example 5.43: Evaluate the integral 𝐼 = ∫𝐶(3𝑥 2 𝑑𝑥 + 2𝑦𝑧𝑑𝑦 + 𝑦 2 𝑑𝑧)from 𝐴: (0,1,2)

to𝐵: (1, −1,7)

Solution: If F has a potential 𝑓, we should have

𝑓𝑥 = 𝐹1 = 3𝑥 2 , 𝑓𝑦 = 𝐹2 = 2𝑦𝑧, 𝑓𝑧 = 𝐹3 = 𝑦 2

We show that we can satisfy these conditions. By integration of 𝑓𝑥 and differentiation,

𝑓 = 𝑥 3 + 𝑔(𝑦, 𝑧), 𝑔 = 𝑦 2 + ℎ(𝑧), 𝑓 = 𝑥 3 + 𝑦 2 𝑧 + ℎ(𝑧),

𝑓𝑦 = 𝑔𝑦 = 2𝑦𝑧, 𝑓𝑧 = 𝑦 2 + ℎ′ = 𝑦 2 , ℎ′ = 0, ℎ = 0

This gives 𝑓(𝑥, 𝑦, 𝑧) = 𝑥 3 + 𝑦 3 𝑧

𝐼 = 𝑓(1, −1,7) − 𝑓(0,1,2) = 1 + 7 − (0 + 2) = 6

Exercise: Evaluate the path independent integrals

𝜋 1 1 1
1. ∫𝜋/2,𝜋 (2 cos 2 𝑥 cos 2𝑦 𝑑𝑥 − 2 sin 2 𝑥 sin 2𝑦 𝑑𝑦)
(6,1)
2. ∫(4,0) 𝑒 4𝑦 (2𝑥 𝑑𝑥 + 4𝑥 2 𝑑𝑦)
(2,1/2,𝜋/2)
3. ∫(0,0,𝜋) 𝑒 𝑥𝑦 (𝑦 sin 𝑧 𝑑𝑥 + 𝑥 sin 𝑧 𝑑𝑦 + cos 𝑧 𝑑𝑧)
(1,1,0) 2 +𝑦 2 +𝑧 2
4. ∫(0,0,0) 𝑒 𝑥 (𝑥 𝑑𝑥 + 𝑦 𝑑𝑦 + 𝑧 𝑑𝑧)
(1,1,1)
5. ∫(0,2,3) (𝑦𝑧 sinh 𝑥𝑧 𝑑𝑥 + cosh 𝑥𝑧 𝑑𝑦 + 𝑥𝑦 sinh 𝑥𝑧 𝑑𝑧)

Badri A, Moges B. and Teklebrhan B. 200 AKU

 APPLIED MATHEMATICS III

Unit Summary

 Vector calculus deals with the application of calculus operations on vectors (vector).
 if 𝐷 is a subset of 𝑅 𝑛 , then a scalar field in 𝐷 is a function
𝑓: 𝐷 ⟶ 𝑅 and a vector field in 𝐷 is a function 𝑭: 𝐷 ⟶ 𝑅 𝑛 .
 A curve in 𝑅 2 (or𝑅 3 ) is a differentiable function 𝒓: [𝑎, 𝑏] ⟶ 𝑅 2(or𝑅 3 ). The initial point is
𝒓(𝒂) and the final point is 𝒓(𝑏).the domain of the curve is the interval [𝑎, 𝑏].
 If a curve C is described by the parametric equation 𝑥 = 𝑓 (𝑡), 𝑦 = 𝑔(𝑡) and z = ℎ(𝑡),
𝑎 ≤ 𝑡 ≤ 𝑏, where 𝑓 ′ , 𝑔′ and ℎ′ are continuous on [𝑎, 𝑏] and C is traversed exactly once as
𝑡 increases from 𝑎 to 𝑏, then the length 𝐿 of C is.

𝑏 𝑑𝑥 2 𝑑𝑦 2 𝑏 𝑑𝑥 𝑑𝑦 𝑑𝑧2 2 2
𝐿 = ∫𝑎 √( 𝑑𝑡 ) + ( 𝑑𝑡 ) 𝑑𝑡 or 𝐿 = ∫𝑎 √( 𝑑𝑡 ) + ( 𝑑𝑡 ) + ( 𝑑𝑡 ) 𝑑𝑡

 Let 𝑓: 𝑅 3 → 𝑅 be a scalar field, that is a function of three variables. The gradient of 𝑓,

denoted ∇𝑓, is the vector field given by
𝜕𝑓 𝜕𝑓 𝜕𝑓 𝜕𝑓 𝜕𝑓 𝜕𝑓
∇𝑓 = 〈 , , 〉= 𝒊 + 𝒋+ 𝒌.
𝜕𝑥 𝜕𝑦 𝜕𝑧 𝜕𝑥 𝜕𝑦 𝜕𝑧

Symbolically, we can consider ∇ to be a vector of differential operators, that is

𝜕 𝜕 𝜕
∇= 𝒊 + 𝒋+
𝜕𝑥 𝜕𝑦 𝜕𝑧

 Let F be a vector field given by 𝑭 = f𝒊 + g𝒋 + h𝒌 ,where, 𝑓, 𝑔, and ℎ are scalar

functions. The divergence of 𝑭 is

𝜕𝑓 𝜕𝑔 𝜕ℎ
𝑑𝑖𝑣 𝑭 = + + ,
𝜕𝑥 𝜕𝑦 𝜕𝑧
and the Curl of 𝐹 is
𝜕ℎ 𝜕𝑔 𝜕𝑓 𝜕ℎ 𝜕𝑔 𝜕𝑓
𝐶𝑢𝑟𝑙 𝑭 = ( − ) 𝒊 + ( − ) 𝒋 + ( − ) 𝒌
𝜕𝑦 𝜕𝑧 𝜕𝑧 𝜕𝑥 𝜕𝑥 𝜕𝑦
 If 𝑓 is defined in a region containing a smooth curve 𝑐 of finte length , then the line
integral of 𝑓 along 𝑐 is given by
∫𝑪 𝒇(𝒙, 𝒚) 𝒅𝒔 = 𝐥𝐢𝐦 ∑𝒏𝒊=𝟏 𝒇(𝒙𝒊 , 𝒚𝒊 ) ∆𝒔𝒊 Plane
||∆||→𝟎

∫𝑪 𝒇(𝒙, 𝒚, 𝒛) 𝒅𝒔 = 𝐥𝐢𝐦 ∑𝒏𝒊=𝟏 𝒇(𝒙𝒊 , 𝒚𝒊 , 𝒛𝒊 ) ∆𝒔𝒊 Space

||∆||→𝟎

Badri A, Moges B. and Teklebrhan B. 201 AKU

 APPLIED MATHEMATICS III

provided this limit exist.

 If 𝑭 is a continuous vector field defined on a smooth curve 𝐶 given by 𝒓(𝑡),
𝑎 ≤ 𝑡 ≤ 𝑏. The line integral of 𝑭 on 𝐶 is given by
𝑏

∫ 𝑭 ∙ 𝑑𝒓 = ∫ 𝑭 ∙ 𝑻𝑑𝑠 = ∫ 𝑭(𝑥(𝑡), 𝑦(𝑡), 𝑧(𝑡)) ∙ 𝒓′ (𝑡)𝑑𝑡

𝐶 𝐶 𝑎

 If 𝑅 be a simply connected region with a piecewise smooth boundary 𝐶, oriented

counterclockwise (that is, 𝐶 is traversed once so that the region 𝑅 always lies to the left).
If 𝑀 and 𝑁 have continuous first partial derivatives in an open region containing 𝑅,
then, Green’s Theorem states that
𝜕𝑁 𝜕𝑀
∫𝐶 𝑀 𝑑𝑥 + 𝑁 𝑑𝑦 = ∬𝑅( 𝜕𝑥 − 𝜕𝑦
) 𝑑𝐴.

 Let S be a surface given by 𝑧 = 𝑔(𝑥, 𝑦) and let 𝑅 be its projection on to the 𝑥𝑦 − plane.
Suppose that 𝑔, 𝑔𝑥 , 𝑎𝑛𝑑 𝑔𝑦 are continuous at all points in R and that 𝑓 is defined on S.
𝑛

∬ 𝑓(𝑥, 𝑦, 𝑧) 𝑑𝑠 = lim ∑ 𝑓(𝑥𝑖 , 𝑦𝑖 , 𝑧𝑖 )∆𝑆𝑖

‖∆‖→0
𝑆 𝑖=1

 Let 𝐹(𝑥, 𝑦, 𝑧) = 𝑀𝑖 + 𝑁𝑗 + 𝑃𝑘, where M, N and P have continuous first partial

derivatives on the surface 𝑆 oriented by a unit normal vector N. the flux integral of F
across 𝑆 is given by

∬ 𝐹⦁𝑁 𝑑𝑆
𝑆

 Let S be a piecewise smooth oriented surface with unit normal vector N, and let the
boundary of S be a piecewise smooth simple closed curve C. Let 𝐹(𝑥, 𝑦, 𝑧) be a
continuous vector function that has continuous first partial derivatives in a domain in
space containing S. Then

∫𝐶 𝐹⦁ 𝑑𝑟 = ∬𝑆(𝑐𝑢𝑟𝑙 𝐹)⦁𝑁𝑑𝑠

𝑖 𝑗 𝑘
Where 𝑐𝑢𝑟𝑙 𝐹 = |𝜕⁄𝜕𝑥 𝜕⁄
𝜕𝑦
𝜕⁄ |
𝜕𝑧
𝐹1 𝐹2 𝐹3

Badri A, Moges B. and Teklebrhan B. 202 AKU

 APPLIED MATHEMATICS III

Miscellaneous Exercises

1. What are scalar fields? Vector fields? Potential? Give examples.

2. What is the gradient? How is it related to directional derivatives?
3. What is the definition of the divergence? Its physical meaning?
4. How do you express the result of forces, the moment of a force, and the work done by a force
in terms of vector?
5. List the kind of integrals in this chapter and how the integral theorems relate some of them.
6. State from memory how can you can evaluate a line integral. A double integral.
7. What do you remember about path independence? Why is it important?
8. How did we use Stokes’s theorem in connection with path independence?
9. State the definition of curl. Why is it important in this chapter?
10. How can you transform a double integral or a surface integral into a line integral ?
11. Let 𝑓 = 𝑧𝑦 + 𝑦𝑥, 𝒗 = [𝑦, 𝑧, 4𝑧 − 𝑥], 𝒘 = [𝑦 2 , 𝑧 2 , 𝑥 2 ]. Find
a. grad 𝑓 at (3,4,0)
b. (grad f)×grad 𝑓, (grad f) ∙ grad𝒇
c. div 𝒗, div 𝒘
d. curl v, curl w
e. curl(grad 𝒇), div(grad 𝒇)
f. 𝛁 2 (𝒇), 𝛁 2 (𝒇2 )
g. div(𝒗 × 𝒘)
h. 𝑐𝑢𝑟𝑙(𝒗 × 𝒘) + 𝑐𝑢𝑟𝑙(𝒘 × 𝒗)

12. Evaluate line integral, with 𝑭 and 𝐶 as given , by the method that seems most suitable.
Recall that if 𝑭 is a force, the integral gives the work done in a displacement along 𝐶.
a. 𝑭 = [𝑥 2 , 𝑦 2 , 𝑧 2 ]
𝐶the straight line segment from (4,1,8) to (0,2,3)
b. 𝑭 = [𝑦𝑧, 2𝑧𝑥, 𝑥𝑦],
𝐶The circle 𝑥 2 + 𝑦 2 = 9, 𝑧 = 1, counterclockwise
c. 𝑭 = [𝑠𝑖𝑛𝜋𝑦, 𝑐𝑜𝑠𝜋𝑥, 𝑠𝑖𝑛𝜋𝑥],
𝐶the boundary of 0 ≤ 𝑥 ≤ 1⁄2 , 0 ≤ 𝑦 ≤ 2, 𝑧 = 2𝑥
d. 𝑭 = [𝑥 − 𝑦, 0, 𝑒 𝑥 ]
𝐶: 𝑦 = 3𝑥 2 , 𝑧 = 2𝑥for𝑥 from 0 to 2
13. Using Green’s Theorem evaluate the line integral
3. ∫𝐶(𝑥 2 − 𝑦 2 ) 𝑑𝑥 + 2𝑥𝑦𝑑𝑦 , 𝐶 = 𝑥 2 + 𝑦 2 = 16
4. ∫𝐶 𝑒 𝑥 𝑐𝑜𝑠2𝑦𝑑𝑥 − 2𝑒 𝑥 𝑠𝑖𝑛2𝑦𝑑𝑦 , 𝐶 = 𝑥 2 + 𝑦 2 = 𝑎2
5. ∫𝐶(𝑥 − 3𝑦) 𝑑𝑥 + (𝑥 + 𝑦)𝑑𝑦
𝐶:boundary of the region lying between the graphs of

Badri A, Moges B. and Teklebrhan B. 203 AKU

 APPLIED MATHEMATICS III

𝑥 2 + 𝑦 2 = 1and𝑥 2 + 𝑦 2 = 9
14. Evaluate the integral ∬𝑆(𝑐𝑢𝑟𝑙 𝑭) ∙ 𝒏dA directly for the give: 𝑭and S.
a. 𝑭 = [4𝑧 2 , 16𝑥, 0], 𝑆: 𝑧 = 𝑦 (0 ≤ 𝑥 ≤ 1, 0 ≤ 𝑦 ≤ 1)
1
b. 𝑭 = [0,0,5𝑥𝑐𝑜𝑠𝑧], 𝑆 = 𝑥 2 + 𝑦 2 = 4, 𝑦 ≥ 0, 0 ≤ 𝑧 ≤ 2 𝜋
c. 𝑭 = [−𝑒 𝑦 , 𝑒 𝑧 , 𝑒 𝑥 ], 𝑆: 𝑧 = 𝑥 + 𝑦 (0 ≤ 𝑥 ≤ 1, 0 ≤ 𝑦 ≤ 1)
d. 𝑭 = [3𝑐𝑜𝑠𝑦, 𝑐𝑜𝑠ℎ𝑧, 𝑥], 𝑆 the square 0 ≤ 𝑥 ≤ 2, 0 ≤ 𝑦 ≤ 2, 𝑧 = 4
15. Calculate this integral by Stokes’s theorem, clockwise as seen by a person standing at the
origin, for the following 𝐹and 𝐶. Assume the Cartesian coordinates to be right handed.
a. 𝑭 = [−3𝑦, 3𝑥, 𝑧], 𝐶 the circle 𝑥 2 + 𝑦 2 = 4, 𝑧 = 1
b. 𝑭 = [4𝑧, −2𝑥, 2𝑥], 𝐶 the intersection of 𝑥 2 + 𝑦 2 = 1 and 𝑧 = 𝑦 + 1
c. 𝑭 = [𝑦 2 , 𝑥 2 , −𝑥 + 𝑧], around the triangle with vertices(0,0,1), (1,0,1), (1,1,1)
d. 𝑭 = [𝑦, 𝑥𝑦 3 , −𝑧𝑦 3 ] , 𝐶 the circle 𝑥 2 + 𝑦 2 = 𝑎2 , 𝑧 = 𝑏 (> 0)
16. Evaluate the surface integral directly or, if possible , by the divergence theorem.
a. 𝑭 = [2𝑥 2 , 4𝑦, 0], 𝑆 = 𝑥 + 𝑦 + 𝑧 = 1, 𝑥 ≥ 0, 𝑦 ≥ 0, 𝑧 ≥ 0
b. 𝑭 = [𝑦, −𝑥, 0], 𝑆 = 3𝑥 + 2𝑦 + 𝑧 = 6, 𝑥 ≥ 0, 𝑦 ≥ 0, 𝑧 ≥ 0
c. 𝑭 = [𝑥 − 𝑦, 𝑦 − 𝑧, 𝑧 − 𝑥], 𝑆 the sphere of radius 5 and center 0
d. 𝑭 = [𝑦 2 , 𝑥 2 , 𝑧 2 ], 𝑆the surface of 𝑥 2 + 𝑦 2 ≤ 4, 0 ≤ 𝑧 ≤ 5
e. 𝑭 = [𝑦 3 , 𝑥 3 , 3𝑧 2 ], 𝑆 the portion of the paraboloid 𝑧 = 𝑥 2 + 𝑦 2 , 𝑧 ≤ 4

Badri A, Moges B. and Teklebrhan B. 204 AKU

 APPLIED MATHEMATICS III

References

1. Erwin Kreyszing,Advanced engineering mathematics, ,John Wiley and Sons,Inc,2006

2. Larson Edwards, Calculus,Brooks/Cole CengageLearning,2010

3. A. Ganesh and Etla, Engineering Mathematics II, New age International press,2009

4. Wilfred Kaplan,Advanced Calculus, 5th edition, publishing house of electronics industry

5.Salas Hille Etgen, Calculus – One and Several variables,10th edition, WILLEY PLUS

6.Knopp, K., Theory f functions, 2 parts.New York: Dover,1996

7.Krantz, S. G., Complex Analysis: The Geometric Viewpoint. Washington, DC,1999

8.Kaplan, W.: "Advanced Calculus," 5th ed., Addison-Wesley Higher Mathematics, Boston,
2003
9. Knopp, K., Theory of Functions. 2 parts. New York:Dover, Reprinted 1996.
10. Krantz, S. G., Complex Analysis: The GeometricViewpoint. Washington, DC: The
MathematicalAssociation of America, 1990.
11.Lang, S., Complex Analysis. 4th ed. New York:Springer, 1999.
12. ] Narasimhan, R., Compact Riemann Surfaces. NewYork: Springer, 1996.
13. Nehari, Z., Conformal Mapping. Mineola, NY:Dover, 1975.
14. Springer, G., Introduction to Riemann Surfaces.Providence, RI: American Mathematical
Society, 2001

Badri A, Moges B. and Teklebrhan B. 205 AKU

 APPLIED MATHEMATICS III

CHAPTER-6

6. COMPLEX ANALYTIC FUNCTIONS

Introduction

The transition from “real calculus” to “complex calculus” starts with a discussion of complex
numbers and their geometric representation in the complex plane. We desire functions to be
analytic because these are the “useful functions” in the sense that they are differentiable in some
domain and operations of complex analysis can be applied to them. The most important
equations are the Cauchy–Riemann equations because they allow a test of analyticity of such
functions. Moreover, we show how the Cauchy–Riemann equations are related to the important
Laplace equation.

The remaining sections of the chapter are devoted to elementary complex functions (exponential,
trigonometric, hyperbolic, and logarithmic functions). These generalize the familiar real
functions of calculus.

Unit Objectives:

On the completion of this unit, students should be able to:

 understand the definition of a complex analytic functions;

 find the limit, derivative, analytic function of complex variable;
 Understand the Cauchy-Riemann equation;
 Understand the exponential and trigonometric in the elementary function.
 Understand the hyperbolic function, logarithm function and general power.

Badri A, Moges B. and Teklebrhan B. 206 AKU

 APPLIED MATHEMATICS III

6.1. Complex Numbers; the Triangle Inequality

6.1. 1. Complex Numbers

Overview:

In this section, we are going to deal with the definition and notation of the complex numbers by
considering various examples.

Section Objectives:

At the end of this subtopic, students will be able to:

 define a complex numbers;

 represent a triangle inequality using the notation;

Definition: A complex number is an order pair (𝑥, 𝑦) of real number x and y that is 𝑧 = (𝑥, 𝑦)

𝑥 is called the real part and y is called the imaginary part of z, written 𝑥 = 𝑅𝑒 𝑧 and 𝑦 = 𝐼𝑚 𝑧

The order pair (𝑥, 0) = 𝑥 and (0,1) = 𝑖 by definition, two complex number are equal if and
only if their real parts are equal and their imaginary parts are equal.

Now the complex number 𝑧 = (𝑥, 𝑦) = 𝑥 + 𝑖𝑦 if 𝑥 = 0, then 𝑧 = 𝑖𝑦 and is called pure

imaginary.

Notation 𝑧 = 𝑥 + 𝑖𝑦

Addition of two complex numbers 𝑧1 = (𝑥1 , 𝑦1 ) and 𝑧2 = (𝑥2 , 𝑦2 ) is defined by

𝑧1 + 𝑧2 = (𝑥1 , 𝑦1 ) + (𝑥2 , 𝑦2 ) = (𝑥1 + 𝑥2 , 𝑦1 + 𝑦2 )

Multiplication is defined by

𝑧1 𝑧2 = (𝑥1 , 𝑦1 )(𝑥2 , 𝑦2 ) = (𝑥1 𝑥2 − 𝑦1 𝑦2 , 𝑥1 𝑦2 + 𝑥2 𝑦1 )

Based on the multiplication property, we find that

Badri A, Moges B. and Teklebrhan B. 207 AKU

 APPLIED MATHEMATICS III

𝑖 2 = −1 𝑎𝑠 𝑖 2 = 𝑖𝑖 = (0,1)(0,1) = (−1,0) = −1

From this we see continued multiplication by positive power of 𝑖 leads to the following pattern:

𝑖 = 𝑖, 𝑖 2 = −1, 𝑖 3 = −𝑖, 𝑖 4 = 1, 𝑖 5 = 𝑖, …

Example 6.1: - Find Real part, Imaginary part, Sum and Product of Complex Numbers

Let 𝑧1 = 8 + 3𝑖 and 𝑧2 = 9 − 2𝑖.

Then 𝑅𝑒𝑧1 = 8, 𝑅𝑒𝑧2 = 9 , 𝐼𝑚𝑧1 = 3, 𝐼𝑚𝑧2 = −2

And 𝑧1 + 𝑧2 = (8,3) + (9, −2) = (8 + 9, 3 + (−2)) = (17,1) = 17 + 𝑖

𝑧1 𝑧2 = (8,3)(9, −2) = (8(9) − 3(−2), 8(−2) + 9(3)) = (78,11) = 78 + 11𝑖

Subtraction and Division are defined as the inverse operation of addition and multiplication,
respectively. Thus the difference is 𝑧 = 𝑧1 − 𝑧2 the complex number 𝑧 for which 𝑧1 = 𝑧 + 𝑧2

Hence 𝑧1 − 𝑧2 = (𝑥1 − 𝑥2 , 𝑦1 − 𝑦2 ) = 𝑥1 − 𝑥2 + 𝑖(𝑦1 − 𝑦2 )

𝑧1
The quotient 𝑧 = ⁄𝑧2 (𝑧2 ≠ 0)is the complex number 𝑧1 = 𝑧𝑧2 .

If we equate the real and the imaginary parts on both sides of this equation, setting 𝑧 = 𝑥 + 𝑖𝑦

We obtain

𝑥1 = 𝑥2 𝑥 − 𝑦2 𝑦, 𝑦1 = 𝑦2 𝑥 + 𝑥2 𝑦

The solution is

𝑧1 𝑥1 𝑥2 + 𝑦1 𝑦2 𝑥2 𝑦1 − 𝑥1 𝑦2
𝑍= = 𝑋 + 𝑖𝑌 , 𝑋 = 2 2 𝑎𝑛𝑑 𝑌 =
𝑧2 𝑥2 + 𝑦2 𝑥22 + 𝑦22

𝑧1
The practical rule used to get this is by multiplying numerator and denominator of ⁄𝑧2
By 𝑥2 − 𝑖𝑦2 and simplifying

𝑥1 + 𝑖𝑦1 (𝑥1 + 𝑖𝑦1 ) (𝑥2 − 𝑖𝑦2 ) 𝑥1 𝑥2 + 𝑦1 𝑦2 𝑥2 𝑦1 − 𝑥1 𝑦2

𝑍= = = 𝑖
𝑥2 + 𝑖𝑦2 (𝑥2 + 𝑖𝑦2 ) (𝑥2 − 𝑖𝑦2 ) 𝑥22 + 𝑦22 𝑥22 + 𝑦22

Badri A, Moges B. and Teklebrhan B. 208 AKU

 APPLIED MATHEMATICS III

Example 6.2: - Let 𝑧1 = 8 + 3𝑖 and 𝑧2 = 9 − 2𝑖. Find the difference and quotient.

We get 𝑧1 − 𝑧2 = (𝑥1 − 𝑥2 , 𝑦1 − 𝑦2 ) = 8 − 9 + 𝑖(3 − (−2)) = −1 + 5𝑖 and

𝑧1 8 + 3𝑖 (8 + 3𝑖 )(9 + 2𝑖) 66 + 43𝑖 66 43𝑖

= = = = +
𝑧2 9 − 2𝑖 (9 − 2𝑖) (9 + 2𝑖) 81 + 4 85 85

Complex Plane
So far we discussed the algebraic manipulation of complex numbers. Consider the geometric
representation of complex numbers, which is of great practical importance. We choose two
perpendicular coordinate axes, the horizontal 𝑥-axis, called the real axis, and the vertical 𝑦-axis,
called the imaginary axis. On both axes we choose the same unit of length (Fig. 6.1.1.1). This is
called a Cartesian coordinate system.

The complex plane fig.6.1.1.1 The complex plane in 4 − 3𝑖 in fig.6.1.1.2

Definition: We now plot a given complex number 𝑧 = (𝑥, 𝑦) = 𝑥 + 𝑖𝑦 as the point P with
coordinates 𝑥, 𝑦. The 𝑥𝑦 plane in which the complex numbers are represented in this way is
called the complex plane.
Addition and subtraction can now be visualized as illustrated in Figs. 6.1.1.3 and 6.1.1.4

Badri A, Moges B. and Teklebrhan B. 209 AKU

 APPLIED MATHEMATICS III

Fig. 6.1.1.3Addition of complex numbers Fig. 6.1.1.4 subtraction of complex numbers

Complex Conjugate Numbers: The complex conjugate 𝑧̅ of a complex number 𝑧 = 𝑥 + 𝑖𝑦 is

defined by
𝑧̅ = 𝑥 − 𝑖𝑦
It is obtained geometrically by reflecting the point 𝑧 in the real axis. Figure 6.1.1.5 shows this for
𝑧 = 5 + 2𝑖 and it’s conjugate 𝑧̅ = 5 − 2𝑖

Fig. 6.1.1.5 Complex conjugate numbers

The complex conjugate is important because it permits us to switch from complex to real.
Indeed, by multiplication, 𝑧𝑧̅ = 𝑥 2 + 𝑦 2 (verify!). By addition and subtraction,
𝑧 + 𝑧̅ = 2𝑥, 𝑧 − 𝑧̅ = 2𝑖𝑦. We thus obtain for the real part 𝑥 and the imaginary part 𝑦 (not 𝑖𝑦!) of
𝑧 = 𝑥 + 𝑖𝑦 . The important formulas are
1 1
𝑅𝑒𝑧 = 𝑥 = (𝑧 + 𝑧̅), 𝐼𝑚𝑧 = (𝑧 − 𝑧̅)
2 2𝑖
If 𝑧 is real, 𝑧 = 𝑥 then 𝑧̅ = 𝑧 by the definition of 𝑧̅ and conversely. Working with Conjugates is
easy, since we have

Badri A, Moges B. and Teklebrhan B. 210 AKU

 APPLIED MATHEMATICS III

(𝑧̅̅̅̅̅̅̅̅̅
1 + 𝑧2 ) = 𝑧̅1 + 𝑧̅2 , (𝑧̅̅̅̅̅̅̅̅̅)
1 − 𝑧2 = 𝑧̅1 − 𝑧̅2

̅̅̅̅̅̅
𝑧1 𝑧̅1
𝑧1 𝑧2 = 𝑧̅1 𝑧̅2 ,
̅̅̅̅̅̅ ( )=
𝑧2 𝑧̅2
Example 6.3: Let 𝑧1 = 4 + 3𝑖 𝑎𝑛𝑑 𝑧2 = 2 + 5𝑖 .
1 3𝑖+3𝑖
Then 𝐼𝑚𝑧1 = 2𝑖 [(4 + 3𝑖) − (4 − 31)] = =3
2𝑖

Also the multiplication of

𝑧1 𝑧2 = ̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅
̅̅̅̅̅̅ (4 + 3𝑖)(2 + 5𝑖) = ̅̅̅̅̅̅̅̅̅̅̅̅
−7 + 26𝑖 = −7 − 26𝑖
𝑧̅1 𝑧̅2 = (4 − 3𝑖)(2 + 5𝑖) = −7 − 26𝑖
Exercise
1. show that i2 = −1, i3 = −i, i5 = i
2. Let z = x + iy. Find in terms of x and y
a) Im(1/z) and Im(1/z 2 )
b) Re[(1 + i)16 z 2 ]
c) Rez 4 − (Rez 2 )2

Polar Form of Complex Numbers: We gain further insight into the arithmetic operations of
complex numbers if, in addition to the 𝑥𝑦-coordinates in the complex plane, we also employ the
usual polar coordinates 𝜃 defined by 𝑥 = 𝑟 cos 𝜃 , 𝑦 = 𝑟 sin 𝜃 then 𝑧 = 𝑥 + 𝑖𝑦 hence
𝑧 = 𝑟(𝑐𝑜𝑠𝜃 + 𝑖𝑠𝑖𝑛𝜃) is called polar form
𝑟 is called the absolute value or modulus of 𝑧 and is denoted by |𝑧|. Hence
|𝑧| = 𝑟 = √𝑥 2 + 𝑦 2 = √𝑧𝑧̅
Geometrically, |𝑧| is the distance of the point 𝑧 from the origin .Similarly, |𝑧1 − 𝑧2 |is the
distance between 𝑧1 𝑎𝑛𝑑 𝑧2 .
𝜃 is called the argument of 𝑧 and is denoted by 𝜃 = 𝑎𝑟𝑔 𝑧.
𝑦
tan 𝜃 = (𝑧 ≠ 0)
𝑥
Geometrically, 𝜃 is the directed angle from the positive 𝑥-axis to 𝑂𝑃 in Fig.6.01. Here, as in
calculus, all angles are measured in radians and positive in the counterclockwise sense.

Badri A, Moges B. and Teklebrhan B. 211 AKU

 APPLIED MATHEMATICS III

For 𝑧 = 0 this angle 𝜃 is undefined. (Why?) For a given 𝑧 ≠ 0 it is determined only up to integer
multiples of 2𝜋 since cosine and sine are periodic with period 2𝜋 . But one often wants to
specify a unique value of 𝑎𝑟𝑔𝑧 of a given 𝑧 ≠ 0 . For this reason one defines the principal
value 𝐴𝑟𝑔 𝑧 (with capital A!) of arg 𝑧 by the double inequality −𝝅 < 𝑎𝑟𝑔 𝑧 ≤ 𝜋
Then we have 𝐴𝑟𝑔 𝑧 = 0 for positive real𝑧 = 𝑥 which is practical, and 𝐴𝑟𝑔𝑧 = 0 (not −𝜋 ) for
negative real 𝑧. Obviously, for a given 𝑧 ≠ 0, the other values of arg 𝑧 are
𝑎𝑟𝑔 𝑧 = 𝐴𝑟𝑔 𝑧 ± 2𝑛𝜋 (𝑛 = ±1, ±2, ⋯ )

Fig. 6.1.1.6.Complex plane, Fig. 6.1.1.7 Distance between two points in the complex plane
Polar form of a complex number

Example 6.4: Find the Polar Form of Complex Numbers of 𝑧 =1+𝑖

Solution: 𝑟 = |1 + 𝑖| = |𝑧| = √2
Now
1 1
Polar form 𝑧 = √2 (𝑐𝑜𝑠 4 𝜋 + 𝑖 𝑠𝑖𝑛 4 𝜋), Fig 6.1.1.8

𝐻𝑒𝑛𝑐𝑒 𝑤𝑒 𝑜𝑏𝑡𝑎𝑖𝑛
1
arg 𝑧 = 𝜋 ± 2𝑛𝜋 (𝑛 = 0,1, ⋯ ),
4
1
𝑎𝑛𝑑 arg 𝑧 = 4 𝜋 and (the principal value).

Example 6.5: Find the Polar Form of Complex Numbers of 𝑧 = 3 − 3𝑖

𝑟 = |3 − 3𝑖| = √9 + 9 = √18 = 3√2
And
𝜋
𝜃𝑝 = Arg(z), 𝜃 = tan−1 −1 = − 4

Hence

Badri A, Moges B. and Teklebrhan B. 212 AKU

 APPLIED MATHEMATICS III

π π
𝑧 = 3 − 3𝑖 = 3√2 (cos − i sin )
4 4
Similarly,
1 1 1
𝑧 = 3 + 3√3𝑖 = 6 (𝑐𝑜𝑠 𝜋 + 𝑖 𝑠𝑖𝑛 𝜋) , |𝑧| = 6, 𝑎𝑛𝑑 𝐴𝑟𝑔 𝑧 = 𝜋
3 3 3
6.1.2. Triangle Inequality
Inequalities such as 𝑥1 < 𝑥2 make sense for real numbers, but not in complex because there is no
natural way of ordering complex numbers. However, inequalities between absolute values
(which are real!), such as|𝑧1 < |𝑧2 || (meaning that𝑧1 is closer to the origin than 𝑧2 ) are of great
importance. The daily bread of the complex analyst is the triangle inequality
Theorem6.1.2: |𝑧1 + 𝑧2 | ≤ |𝑧1 | + |𝑧2 |
Proof: This inequality follows by noting that the three points are the vertices of a triangle with
sides |𝑧1 | , |𝑧2 | and |𝑧1 + 𝑧2 | and one side cannot exceed the sum of the other two sides.
Now
|𝑧1 + 𝑧2 |2 = (𝑧1 + 𝑧2 )(𝑧̅1 + 𝑧̅2 ) = 𝑧1 𝑧̅1 + (𝑧1 𝑧̅2 + ̅̅̅̅̅̅
𝑧1 𝑧̅2 ) + 𝑧2 𝑧̅2 (∗)
But
𝑧1 𝑧̅2 + ̅̅̅̅̅̅
𝑧1 𝑧̅2 = 2𝑅𝑒(𝑧1 𝑧̅2 ) ≤ 2|𝑧1 ||𝑧2 |
This is substitute in to (*) we have,
|𝑧1 + 𝑧2 |2 ≤ (|𝑧1 | + |𝑧2 |)2
Therefore
|𝑧1 + 𝑧2 | ≤ |𝑧1 | + |𝑧2 |

Multiplication and Division in polar form

If 𝑧1 = 𝑟1 (cos 𝜃1 + 𝑖 sin 𝜃1 )𝑎𝑛𝑑 𝑧2 = 𝑟2 (cos 𝜃2 + 𝑖 sin 𝜃2 )
Then
𝑧1 𝑧2 = 𝑟1 𝑟2 (cos 𝜃1 cos 𝜃2 − sin 𝜃1 sin 𝜃2 ) + 𝑖(sin 𝜃1 cos 𝜃2 + cos 𝜃1 sin 𝜃2
In addition rules for the sine and cosine, now
𝑧1 𝑧2 = 𝑟1 𝑟2 [cos(𝜃1 + 𝜃2 ) − 𝑖 sin(𝜃1 + 𝜃2 ]
𝑧1 𝑟
And = 𝑟1 [cos(𝜃1 − 𝜃2 ) − 𝑖 sin(𝜃1 − 𝜃2 ]
𝑧2 2

The argument of a product equals the sum of the argument of the factors,
arg (𝑧1 𝑧2 ) = arg 𝑧1 + arg 𝑧2

Badri A, Moges B. and Teklebrhan B. 213 AKU

 APPLIED MATHEMATICS III

Similarly, arg 𝑧1 = arg[(𝑧1 /𝑧2 )𝑧2 ] = arg(𝑧1 /𝑧2 ) + arg 𝑧2 and by subtraction of arg 𝑧2
𝑧1
arg = arg 𝑧1 − arg 𝑧2
𝑧2
Euler formulas: the known as Euler’s formula is 𝑒 𝑖𝜃 = cos 𝜃 + isin 𝜃

We can now express the polar representation of a complex number in the form of 𝑧 = 𝑟𝑒 𝑖𝜃
we now that |𝑒 𝑖𝜃 | = 1
Powers and Roots of Complex numbers
The polar representation of 𝑧 = 𝑟𝑒 𝑖𝜃 is particularly useful in finding powers and roots of various
complex number.
Example 6.6: 𝑧 2 = 𝑟 2 𝑒 2𝑖𝜃 , where as repeatedly multiplying by 𝑧 and 𝑟𝑒 𝑖𝜃 ,we obtain
𝑧 𝑛 = 𝑟 𝑛 (𝑒 𝑖𝜃 )𝑛 = 𝑟 𝑛 𝑒 𝑛𝑖𝜃 , 𝑛 = 1,2, ⋯
𝑧 𝑛 = 𝑟 𝑛 (cos 𝑛𝜃 + 𝑖 sin 𝑛𝜃)
which in terms of trigonometric functions if |𝑧| = 𝑟 = 1 becomes the famous De Moive’s
formula 𝑧 𝑛 = (cos 𝜃 + 𝑖 sin 𝜃)𝑛 = 𝑧 𝑛 = cos 𝑛𝜃 + 𝑖 sin 𝑛𝜃, 𝑛 = 1,2, ⋯
One of the principal uses of De Moiver’s formula is finding fractional powers of complex
numbers. for example suppose we wish to find the solution of the equation 𝑧 𝑛 = 𝑧0
1⁄
Formally, we represent the solution as 𝑧 = 𝑧0 𝑛 but we don’t know how to find the 𝑛𝑡ℎ root of
a complex number. To do so, we first write 𝑧0 = 𝑟0 𝑒 𝑖𝜃0 and 𝑧 𝑛 = 𝑧0
𝑟 𝑛 𝑒 𝑖𝑛𝜃 = 𝑟0 𝑒 𝑖𝜃0
From this relation it now follows that 𝑟 𝑛 = 𝑟0 and 𝑛𝜃 = 𝜃0
𝑛 𝜃0 +2𝑘𝜋
From which we deduce 𝑟 = √|𝑧|, 𝜃= 𝑘 = 0, ±1, ±2, ⋯
𝑛

Thus
𝜃0 +2𝑘𝜋
𝑧 = 𝑛√𝑟0 𝑒 𝑖( 𝑛
)
, 𝑘 = 0,1,2, ⋯ , 𝑛 − 1
The 𝑛𝑡ℎ 𝑟𝑜𝑜𝑡 of any complex number 𝑧 can expressed as
𝜃0 +2𝑘𝜋
1⁄
𝑧 𝑛 = 𝜔𝑘+1 = 𝑒 𝑖( 𝑛
)
, 𝑘 = 0, ±1, ±2, ⋯ Where |𝑧| = 𝑟 and 𝜃0 = 𝐴𝑟𝑔(𝑧) or
𝑚⁄ 𝑚 𝜃0 +2𝑘𝜋
= ( √|𝑧|) 𝑒 𝑖𝑚( )
𝑛
𝑧 𝑛 𝑛 𝑚 = 1,2, ⋯ 𝑎𝑛𝑑 𝑘 = 0,1,2, ⋯ , 𝑛 − 1

Badri A, Moges B. and Teklebrhan B. 214 AKU

 APPLIED MATHEMATICS III

Example 6.7: Find the cube roots of unity

Solution: For 𝑧 = 1 , it follows that 𝜃0 = 0
Then
1⁄ 2𝑘𝜋
𝑖
1 3 = 𝜔𝑘+1 = 𝑒 3 , 𝑘 = 0,1,2
By writing 𝑒 𝑖𝜃 = cos 𝜃 + 𝑖 sin 𝜃, we find
𝑘0 , 𝜔1 = cos 0 + 𝑖 sin 0 = 1

𝑘1 , 𝜔2 = cos 2𝜋⁄3 + 𝑖 sin 2𝜋⁄3 = − 1⁄2 + 𝑖√3⁄2

𝑘2 , 𝜔3 = cos 4𝜋⁄3 + 𝑖 sin 4𝜋⁄3 = − 1⁄2 − 𝑖√3⁄2

Fig.6.1.2.1
3⁄
Example 6.8: Find all values of (−1 + 𝑖√3) 2

2𝜋
Solution: 𝑧 = −1 + 𝑖√3 , |𝑧| = 2, 𝜃0 = 𝑡𝑎𝑛−1 (√3) =
3
2𝜋
+2𝑘𝜋
𝑖3( 3 )
𝜔𝑘+1 = 2√2 𝑒 2 𝑘 = 0,1
𝑘0 , 𝜔1 = 2√2 𝑒 𝑖𝜋 = −2√2
𝑘1 , 𝜔2 = 2√2 𝑒 𝑖4𝜋 = 2√2
Exercise: Find all roots in the complex plane
3
1) √1 + 𝑖
8
2) √1
4
3) √𝑖
5
4) √−1
12
5) √1
7
6) √3 + 4𝑖

Badri A, Moges B. and Teklebrhan B. 215 AKU

 APPLIED MATHEMATICS III

6.2. Functions of Complex Variable; Limit, Derivative, Analytic Function

Overview:

In this section, we are going to deal with the definition and notation of the limit, derivative,
analytic function by considering various examples.

Section Objectives:

At the end of this subtopic, students will be able to:

 Define the function of complex variable;

 Define a limit of functions of complex variable;
 Define the derivative of functions of complex variable;
 Define the analytic function of complex variable.
 Represent the example of a limit, derivative and analytic function using the notation;

6.2.1 Function of Complex Variable

Let S be a set of complex numbers. A function 𝑓 defined on 𝑆 is a rule that assigns to every 𝑧
in 𝑆 a complex number 𝑤, called the value of 𝑓 at 𝑧. We write 𝑤 = 𝑓(𝑧)
Here 𝑧 varies in 𝑆 and is called a complex variable. The set 𝑆 is called the domain of definition
of 𝑓 or, briefly, the domain of 𝑓. (In most cases 𝑆 will be open and connected, thus a domain as
defined just before.)
The set of all values of a function 𝑓 is called the range of 𝑓. 𝑤 is complex, and we write 𝑤 =
𝑢 + 𝑖𝑣 where 𝑢 and 𝑣 are the real and imaginary parts, respectively. Now 𝑤 depends on 𝑧 = 𝑥 +
𝑖𝑦. Hence 𝑢 becomes a real function of 𝑥 and 𝑦, and so does 𝑣. We may thus write
𝑤 = 𝑓(𝑧) = 𝑢(𝑥, 𝑦) + 𝑖𝑣(𝑥, 𝑦)
This shows that a complex function 𝑓 (𝑧) is equivalent to a pair of real functions 𝑢(𝑥, 𝑦) and
𝑣(𝑥, 𝑦), each depending on the two real variables 𝑥 and 𝑦.
Example 6.9: Let 𝑤 = 𝑓(𝑧) = 𝑧 2 + 3𝑧. Find 𝑢(𝑥, 𝑦)and 𝑣(𝑥, 𝑦) the value of 𝑓 at 𝑧 = 1 + 3𝑖
Solution: 𝑓(𝑧) = 𝑧 2 + 3𝑧 = (𝑥 + 𝑖𝑦)2 + 3(𝑥 + 𝑖𝑦)
= 𝑥 2 − 𝑦 2 + 3𝑥 + 𝑖(2𝑥𝑦 + 3𝑦)
Now
𝑢(𝑥, 𝑦) = 𝑅𝑒 𝑓(𝑧) = 𝑥 2 − 𝑦 2 + 3𝑥and 𝑣(𝑥, 𝑦) = 2𝑥𝑦 + 3𝑦 ,

Badri A, Moges B. and Teklebrhan B. 216 AKU

 APPLIED MATHEMATICS III

Also
𝑓(1 + 3𝑖) = (1 + 3𝑖)2 + 3(1 + 3𝑖)
= 1 − 9 + 6𝑖 + 3 + 9𝑖 = −5 + 15𝑖
So
𝑢(1,3) = −5 and 𝑣(1,3) = 15
Example 6.10: Let 𝑤 = 𝑓(𝑧) = 𝑧 2 . Find 𝑢(𝑥, 𝑦)and 𝑣(𝑥, 𝑦) the value of 𝑓 at 𝑧 = 5𝑖
Solution: 𝑓(𝑧) = 𝑧 2 = (𝑥 + 𝑖𝑦)2
= 𝑥 2 − 𝑦 2 + +𝑖(2𝑥𝑦)
Now
𝑢(𝑥, 𝑦) = 𝑅𝑒 𝑓(𝑧) = 𝑥 2 − 𝑦 2 +and 𝑣(𝑥, 𝑦) = 2𝑥𝑦,

Also
𝑓(5𝑖) = (5𝑖)2 = −25

1
Exercise: Let 𝑤 = 𝑓(𝑧) = 2𝑖𝑧 + 6𝑧̅. Find 𝑢(𝑥, 𝑦)and 𝑣(𝑥, 𝑦) the value of 𝑓 at 𝑧 = 2 + 4𝑖

6.2.2. Limits and Continuity of Complex Variable

Definition 1: A sequence of complex numbers {𝑧𝑛 }1∞ is said to have the limit 𝑧0 or to converges
to 𝑧0 , and we write lim 𝑧𝑛 = 𝑧0
𝑛→∞

Or equivalently, 𝑧𝑛 → 𝑧0 as 𝑛 → ∞ if for any 𝜀 > 0 there exists an integer N such that |𝑧𝑛 −
𝑧0 | < 𝜀 for all 𝑛 > 𝑁.

Definition 2: Let 𝑓 be a function defined in some neighborhood of 𝑧0 itself. We say that the limit
of 𝑓(𝑧) as 𝑧 approaches 𝑧0 is the number 𝑤0 and write lim 𝑓(𝑧) = 𝑤0
𝑧→𝑧0

Or equivalently, 𝑓(𝑧) → 𝑤0 as 𝑧 → 𝑧0 if for any 𝜀 > 0 there exists a positive number 𝛿 such that
|𝑓(𝑧) − 𝑤0 | < 𝜀 whenever 0 < |𝑧 − 𝑧0 | < 𝛿 .

Badri A, Moges B. and Teklebrhan B. 217 AKU

 APPLIED MATHEMATICS III

Example 6.11: prove that lim 𝑧 2 = −1

𝑧→𝑖

Solution: we must show that for any given 𝜀 > 0 there is a positive number 𝛿 such that
|𝑧 2 − (−1)| < 𝜀 𝜀 Whenever 0 < |𝑧 − 𝑖| < 𝛿 .
So we express |𝑧 2 − (−1)| in terms of |𝑧 − 𝑖|
𝑧 2 − (−1) = 𝑧 2 + 1 = (𝑧 − 𝑖)(𝑧 + 𝑖) = (𝑧 − 𝑖)(𝑧 − 𝑖 + 2𝑖)
Now
if |𝑧 − 𝑖| < 𝛿 hence |𝑧 2 − (−1)| = |𝑧 − 𝑖||𝑧 − 𝑖 + 2𝑖|
≤ |𝑧 − 𝑖|(|𝑧 − 𝑖| + |2𝑖|)
= |𝑧 − 𝑖|(|𝑧 − 𝑖| + 2)
< 𝛿(𝛿 + 2)
So to ensure that it is less than 𝜀, we choose 𝛿 to be smaller than each either of the number
𝜀 𝜀
and 1. |𝑧 − 𝑖|(|𝑧 − 𝑖| + 2) < 3 (1 + 2) = 𝜀
3
𝑖
Exercise: prove that 𝑙𝑖𝑚 𝑓(𝑧) = where 𝑓(𝑧) = 𝑖𝑧/2 in the open disk |𝑧| < 1,
𝑧→1 2

Definition 3: Let 𝑓 be a function defined in some neighborhood of 𝑧0 . Then 𝑓 is continuous at 𝑧0

if lim 𝑓(𝑧) = 𝑓(𝑧0 )
𝑧→𝑧0

In other words, for 𝑓 to be continuous at 𝑧0 , it must have a limiting value at 𝑧0 and this limiting
value must be 𝑓(𝑧0 ). A function 𝑓 is said to be continuous on a set 𝑆 if it is continuous at each
point of 𝑆.

Note that by definition of a limit this function 𝑓(𝑧) is said to be continuous in a domain if it is
continuous at each point of this domain.

Fig.6.2.2.1 Limit

Badri A, Moges B. and Teklebrhan B. 218 AKU

 APPLIED MATHEMATICS III

Theorem 6.2.2.1: If 𝑙𝑖𝑚 𝑓(𝑧) = 𝐴 𝑎𝑛𝑑 𝑙𝑖𝑚 𝑔(𝑧) = 𝐵 , then

𝑧→𝑧0 𝑧→𝑧0

I. 𝑙𝑖𝑚 (𝑓(𝑧) ± 𝑔(𝑧)) = 𝐴 ± 𝐵

𝑧→𝑧0

II. 𝑙𝑖𝑚 (𝑓(𝑧)𝑔(𝑧)) = 𝐴𝐵

𝑧→𝑧0
𝑓(𝑧) 𝐴
III. 𝑙𝑖𝑚 𝑔(𝑧) = 𝐵 𝑖𝑓 𝐵 ≠ 0
𝑧→𝑧0
z+i
Example 6.12: Evaluate the limit lim
z→−i z2 +1
z+i z+i 1 1
Solution: lim = lim = lim = − 2i
z→−i z2 +1 z→−i (z+i)(z−i) z→−i (z−i)

Example 6.13: Evaluate the limit

𝑧−𝑖
lim
𝑧→𝑖 𝑧 2 + 1

𝑧−𝑖 𝑧−𝑖 1 1
Solution: lim 𝑧 2 +1 = lim (𝑧+𝑖)(𝑧−𝑖) = lim 𝑧+1 = 2𝑖
𝑧→𝑖 𝑧→𝑖 𝑧→𝑖

Definition: (Continuity) A function 𝑤 = 𝑓(𝑧) is said to be continuous at 𝑧 = 𝑧0 , provided

i. 𝑓(𝑧0 ) is defined
ii. lim 𝑓(𝑧0 ) exists
𝑧→𝑧0

iii. lim 𝑓(𝑧) = 𝑓(𝑧0 )

𝑧→𝑧0

6.2.3. Derivative of Complex Variable

Definition: The derivative of a complex function 𝑤 = 𝑓(𝑧) at a fixed point 𝑧0 is written 𝑓′(𝑧0 )
and is defined by

𝑓(𝑧0 + ∆𝑧) − 𝑓(𝑧0 )

𝑓′(𝑧0 ) = lim … … … … … … … … … … … … … … … . . (∗)
∆𝑧→0 ∆𝑧

Provided this limit exists. Then 𝑓 is said to be differentiable at 𝑧0 . if we write ∆𝑧 = 𝑧 − 𝑧0 ,

we have 𝑧 = 𝑧0 + ∆𝑧 and (*) takes the form

𝑓(𝑧) − 𝑓(𝑧0 )
𝑓′(𝑧0 ) = lim
𝑧→𝑧0 𝑧 − 𝑧0

Badri A, Moges B. and Teklebrhan B. 219 AKU

 APPLIED MATHEMATICS III

Theorem 6.2.2.2: If 𝑓(𝑧) and 𝑔(𝑧) are differentiable function at a given point 𝑧, then
𝑑
i. ((𝑓(𝑧) ± 𝑔(𝑧))) = 𝑓′(𝑧) ± 𝑔′(𝑧)
𝑑𝑧
𝑑
ii. ((𝑓(𝑧)𝑔(𝑧))) = 𝑓′(𝑧)𝑔(𝑧) + 𝑓(𝑧)𝑔′(𝑧)
𝑑𝑧
𝑑
iii. (𝑐 𝑓(𝑧)) = 𝑐𝑓′(𝑧) where c is a constant
𝑑𝑧
𝑑 𝑓(𝑧) 𝑓′(𝑧)𝑔(𝑧)−𝑓(𝑧)𝑔′(𝑧)
iv. ( )= 2 𝑖𝑓 𝑔(𝑧) ≠ 0.
𝑑𝑧 𝑔(𝑧) (𝑔(𝑧))

Example 6.14: given 𝑓(𝑧) = 𝑧 2 , show that 𝑓′(𝑧) = 2𝑧

(𝑧+∆𝑧)2 −𝑧 2
Solution: 𝑓′(𝑧) = lim
∆𝑧→0 ∆𝑧

= lim (2𝑧 + ∆𝑧) = 2𝑧

∆𝑧→0

Example: Show that 𝑓(𝑧) = 𝑧̅ is not differentiable at any point in the complex plane.
̅̅̅̅̅̅̅)−𝑧̅
(𝑧+∆𝑧 ̅̅̅̅
∆𝑧
Solution: 𝑓′(𝑧) = lim = lim
∆𝑧→0 ∆𝑧 ∆𝑧→0 ∆𝑧

To evaluate this last limit, we construct to 𝑥 and 𝑦 coordinate to obtain

̅̅̅
∆𝑧 ∆𝑥 − 𝑖∆𝑦
lim = lim
∆𝑧→0 ∆𝑧 ∆𝑥→0,∆𝑦→0 ∆𝑥 + 𝑖∆𝑦

If we allow ∆𝑦 → 0 first
̅̅̅
∆𝑧 ∆𝑥
lim = lim =1
∆𝑧→0 ∆𝑧 ∆𝑥→0 ∆𝑥
̅∆𝑧
̅̅̅ −𝑖∆𝑦
When ∆𝑥 → 0 first lim = lim = −1
∆𝑧→0 ∆𝑧 ∆𝑦→0 𝑖∆𝑦

Therefore the limit is not unique (does note exists) 𝑓(𝑧) = 𝑧̅ does not derivative.
6.2.4. Analytic Function of Complex Variable

Definition: (Analytic function) A function 𝑤 = 𝑓(𝑧) is said to be analytic at a point 𝑧0 if it is

defined at the point and 𝑓 ′ (𝑧) exists at every point in some neighborhood of 𝑧0 .
That is Analytic Function⇒ Differentiability
Differentiability ⇏ Analytic Function

Badri A, Moges B. and Teklebrhan B. 220 AKU

 APPLIED MATHEMATICS III

Example 6.15: The non-negative of power integer 1, 𝑧, 𝑧 2 , … is analytic in the entire complex
plane

6.3. Cauchy-Riemann Equation; Laplace Equation

Overview:

In this section, we are going to deal with the definition and notation of the Cauchy-Riemann
equation and Laplace equation by considering various examples.

Section Objectives:

At the end of this subtopic, students will be able to:

 Define and proof the Cauchy-Riemann equation;

 Define Laplace equation;
 Represent the example of Cauchy-Riemann equation and Laplace equation using the
notation;

6.3.1. Cauchy-Riemann Equation

Let us consider the function 𝑤 = 𝑓(𝑧) defined in the neighborhood of some point 𝑧. if it has a
derivative,
∆𝑓 ∆𝑢 + 𝑖∆𝑣
𝑓′(𝑧) = lim = lim
∆𝑧→0 ∆𝑧 ∆𝑥→0,∆𝑦→0 ∆𝑥 + 𝑖∆𝑦

Where were using the notation ∆𝑓 = 𝑓(𝑧 + ∆𝑧) − 𝑓(𝑧) and 𝑓(𝑧) = 𝑢(𝑥, 𝑦) + 𝑖𝑣(𝑥, 𝑦)
If we first let ∆𝑦 → 0
∆𝑢 𝑖∆𝑣
𝑓′(𝑧) = lim ( + )
∆𝑥→0 ∆𝑥 ∆𝑥

𝜕𝑢 𝜕𝑣
𝑓′(𝑧) = 𝜕𝑥 + 𝑖 𝜕𝑥 ……………………………………... (1)

And let ∆𝑥 → 0

Badri A, Moges B. and Teklebrhan B. 221 AKU

 APPLIED MATHEMATICS III

1 ∆𝑢 ∆𝑣
𝑓′(𝑧) = lim ( + )
∆𝑦→0 𝑖 ∆𝑦 ∆𝑦
𝜕𝑣 𝜕𝑢
𝑓′(𝑧) = 𝜕𝑦 − 𝑖 𝜕𝑦 ……………………………………….. (2)

Now combine equation (1) and (2) ,We have

𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑣
= − , =
𝜕𝑦 𝜕𝑦 𝜕𝑥 𝜕𝑦
This condition is called the Cauchy-Riemann equation.
Example 6.17: using the Cauchy-Riemann equation Show that
𝑓(𝑧) = 𝑧 3 is analytic everywhere.
Solution: we have, 𝑓(𝑧) = 𝑧 3 = (𝑥 + 𝑖𝑦)3
= (𝑥 3 − 3𝑥𝑦 2 ) + 𝑖(3𝑥 2 𝑦 − 𝑦 3 )
Here
𝑢(𝑥, 𝑦) = 𝑥 3 − 3𝑥𝑦 2 and 𝑣(𝑥, 𝑦) = 3𝑥 2 𝑦 − 𝑦 3
Thus
𝑢𝑥 = 3𝑥 2 − 3𝑦 2 , 𝑣𝑥 = 6𝑥𝑦
𝑢𝑦 = −6𝑥𝑦, 𝑣𝑦 = 3𝑥 2 − 3𝑦 2
We observe that 𝑢𝑥 = 𝑣𝑦 and 𝑢𝑦 = −𝑣𝑥 at a points
Hence
𝑓(𝑧) = 𝑧 3 is analytic for every z and further
𝑓 ′(𝑧) = 𝑢𝑥 + 𝑖𝑣𝑥 = 3𝑥 2 − 3𝑦 2 + 𝑖6𝑥𝑦
= 3[𝑥 2 + (𝑖𝑦)2 + 2𝑖𝑥𝑦] = 3(𝑥 + 𝑖𝑦)2 = 3𝑍 2 .
Example 6.18: using the Cauchy-Riemann equation Show that
𝑓(𝑧) = |𝑧|2 is analytic nowhere.

Solution: we have 𝑓(𝑧) = |𝑧|2 = 𝑧𝑧̅ = 𝑥 2 + 𝑦 2

Here
𝑢(𝑥, 𝑦) = 𝑥 2 + 𝑦 2 and 𝑣(𝑥, 𝑦) = 0
Badri A, Moges B. and Teklebrhan B. 222 AKU
 APPLIED MATHEMATICS III

𝑢𝑥 = 2𝑥, 𝑣𝑥 = 0 And 𝑢𝑦 = 2𝑦, 𝑣𝑦 = 0

We observe that 𝑢𝑥 , 𝑢𝑦 , 𝑣𝑥 , and 𝑣𝑦 are continuous everywhere
Moreover 𝑢𝑥 = 𝑣𝑦 is satisfied at all points on 𝑥 = 0, (the imaginary axis) and 𝑢𝑦 = −𝑣𝑥 is
satisfied at all points on 𝑦 = 0, (the real axis)
So both Cauchy-Riemann equations
Satisfied only at (0, 0)
Thus
𝑓(𝑧) = |𝑧|2 is differentiable only at (0, 0) and hence it analytic nowhere.
Exercise: Show that the function f(z) = √xy is not analytic at the origin even though the Cauchy-
Riemann equations are satisfied.
Note: also the Cauchy-Riemann equations in polar coordinates
𝜕𝑢 𝜕𝑣
𝑓 ′ (𝑧) = (cos 𝜃 − 𝑖 sin 𝜃) ( + 𝑖 )
𝜕𝑟 𝜕𝑟
𝜕𝑢 𝜕𝑣
= 𝑒 −𝑖𝜃 ( 𝜕𝑟 + 𝑖 𝜕𝑟 )

And that the Cauchy-Riemann equations assume the function

𝜕𝑢 1 𝜕𝑣
=
𝜕𝑟 𝑟 𝜕𝜃
1 𝜕𝑢 𝜕𝑣
=−
𝑟 𝜕𝜃 𝜕𝑟
6.3.2. Laplace Equation; Harmonic function
Theorem: (Laplace’s equation) if 𝑓(𝑧) = 𝑢(𝑥, 𝑦) + 𝑖𝑣(𝑥, 𝑦) is analytic in a domain D, then both
𝑢 and 𝑣 satisfy Laplace’s equation.
∇2 u = uxx + uyy = 0
And ∇2 v = vxx + vyy = 0
in D and have continuous second partial derivatives in D.
Solution of Laplace’s equation having continuous second order partial derivatives is called
Harmonic function.
Theorem: If 𝑓(𝑧) = 𝑢(𝑥, 𝑦) + 𝑖𝑣(𝑥, 𝑦) is analytic in some domain D, then both 𝑢(𝑥, 𝑦) and
𝑣(𝑥, 𝑦)are Harmonic functions in D.
Theorem 6.3.2: If 𝑓(𝑧) = 𝑢(𝑥, 𝑦) + 𝑖𝑣(𝑥, 𝑦) is analytic in some domain D, then both
(𝑥, 𝑦)and 𝑣(𝑥, 𝑦) are harmonic functions in D.

Badri A, Moges B. and Teklebrhan B. 223 AKU

 APPLIED MATHEMATICS III

Example 6.19: construct an analytic function whose real part is 𝑢(𝑥, 𝑦) = 𝑦 3 − 3𝑥 2 𝑦.

Solution: 𝑢(𝑥, 𝑦) is Harmonic and derivatives
Hence
𝜕2 𝑢 𝜕2 𝑢
+ 𝜕𝑦 2 = −6𝑦 + 6𝑦 = 0, Then 𝑢 is harmonic
𝜕𝑥 2

Using the Cauchy-Riemann equations, we obtained

𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑣
= −6𝑥𝑦 = 𝜕𝑦 and = 3𝑦 2 − 3𝑥 2 = − 𝜕𝑦
𝜕𝑥 𝜕𝑦

Integrating the first relation with respect to y, we get

𝑣(𝑥, 𝑦) = 3𝑦 2 + 𝑔(𝑥) 𝑔(𝑥) is constant.
Using the second Cauchy-Riemann equation
𝜕𝑣
= −3𝑦 2 + 𝑔′ (𝑧) = −3𝑦 2 + 3𝑥 2
𝜕𝑥
𝑔′(𝑧) = 3𝑥 2 , Or 𝑔(𝑧) = 𝑥 3 + 𝑐, where 𝑐 is any constant.
𝑣(𝑥, 𝑦) = 3𝑦 2 + 𝑥 3 + 𝑐, 𝑐 = 0
The analytic function 𝑓(𝑧) = 𝑢(𝑥, 𝑦) + 𝑖𝑣(𝑥, 𝑦)
= 𝑦 3 − 3𝑥 2 𝑦 + 𝑖(3𝑦 2 + 𝑥 3 )

6.4. Elementary Functions: Exponential, Trigonometric

Overview:

In this section, we are going to deal with the definition and notation of the elementary
functions, exponential functions and trigonometric functions by considering various examples.

Section Objectives:

At the end of this subtopic, students will be able to:

 Represent the elementary functions, exponential functions and trigonometric

functions
6.4.1. Elementary Functions: In introducing functions of a complex variable we wish to define
them in such a way that the definition is constant with the real function that arises when 𝑧 is
replaced by the real variable 𝑥.

Badri A, Moges B. and Teklebrhan B. 224 AKU

 APPLIED MATHEMATICS III

If to each complex number 𝑧 there is but one value 𝑤 = 𝑓(𝑧), we say that 𝑓(𝑧) is single-valued.
Multiple valued functions which are not single-valued.

6.4.2. Complex Exponential function

We define the complex exponential function by
𝑒 𝑧 = 𝑒 𝑥+𝑖𝑦 = 𝑒 𝑥 (cos 𝑦 + 𝑖 sin 𝑦).

Properties of 𝑒 𝑧 are
I. 𝑒 𝑧1 𝑒 𝑧2 = 𝑒 𝑧1 +𝑧2 , 𝑒 𝑧 ≠ 0 ∀𝑧
II. |𝑒 𝑧 | = 𝑒 𝑥
III. 𝑒 𝑧 = 1, 𝑦 𝑟𝑒𝑎𝑙 and (𝑒̅̅̅𝑧 ) = 𝑒 𝑧̅
Example 6.20: show that |𝑒 −3𝑖𝑧+5𝑖 | = 𝑒 3𝑦 = |𝑒 −3𝑖𝑧+5𝑖 | = 𝑒 3𝑦
−3𝑖𝑧+5𝑖 |=𝑒 3𝑦
Solution: |𝑒 −3𝑖𝑧+5𝑖 | = |𝑒 3𝑦+|𝑒 | = 𝑒 3𝑦 |𝑒 𝑖(5−3𝑥) | = 𝑒 3𝑦

Note: For all finite𝑥, this implies that 𝑒 𝑧 is non zero for all finite𝑧, also
arg 𝑒 𝑧 = 𝑦 ± 2𝑛𝜋 𝑛 = 0,1,2, ⋯
6.4.3. Trigonometric Function
If we odd and subtract the Euler formulas
𝑒 𝑖𝑦 = cos 𝑦 + 𝑖 sin 𝑦
𝑒 −𝑖𝑦 = cos 𝑦 − 𝑖 sin 𝑦
We are led to the real trigonometric functions
1 𝑖𝑦
cos 𝑦 = (𝑒 + 𝑒 −𝑖𝑦 )
2
1
sin 𝑦 = (𝑒 𝑖𝑦 − 𝑒 −𝑖𝑦 )
2𝑖
To define the complex trigonometric functions………………………………………………(*)
1
cos 𝑧 = (𝑒 𝑖𝑧 + 𝑒 −𝑖𝑧 )
2
1
sin 𝑧 = (𝑒 𝑖𝑧 − 𝑒 −𝑖𝑧 )
2𝑖
𝑑 𝑑
And Formulas for the derivatives follow readily from 𝑑𝑧 cos 𝑧 = − sin 𝑧 , 𝑑𝑧 sin 𝑧 = cos 𝑧

Using the Periodicity

Badri A, Moges B. and Teklebrhan B. 225 AKU

 APPLIED MATHEMATICS III

cos(𝑧 + 2𝜋) = cos 𝑧

sin(𝑧 + 2𝜋) = sin 𝑧
And even and odd function
cos(−𝑧) = cos 𝑧
sin(−𝑧) = − sin 𝑧
Example 6.21: Real, Imaginary Parts and Absolute value.
Show that
a. Cos 𝑧 = cos 𝑥 cosh 𝑦 − 𝑖 sin 𝑥 sinh 𝑦
b. sin 𝑧 = sin 𝑥 cosh 𝑦 + 𝑖 cos 𝑥 sinh 𝑦
c. cos 𝑖𝑧 = cosh 𝑦
d. sin 𝑖𝑧 = 𝑖 sinh 𝑦
e. cos2 𝑧 + sin2 𝑧 = 1
f. |Cos 𝑧|2 = cos2 𝑥 + sinh2 𝑦
g. |sin 𝑧|2 = sin2 𝑥 + sinh2 𝑦
1
Solution a: cos 𝑧 = 2 (𝑒 𝑖(𝑥+𝑖𝑦) + 𝑒 −𝑖(𝑥+𝑖𝑦) )
1 1
= 2 𝑒 −𝑦 (cos 𝑥 + 𝑖 sin 𝑥) + 2 𝑒 𝑦 (cos 𝑥 − 𝑖 sin 𝑥)
1 1
= 2 (𝑒 𝑦 + 𝑒 −𝑦 ) cos 𝑥 − 2 𝑖(𝑒 𝑦 − 𝑒 −𝑦 ) sin 𝑥.
1 1
We know from calculus, cosh 𝑦 = 2 (𝑒 𝑦 + 𝑒 −𝑦 ) and sinh 𝑦 = 2 (𝑒 𝑦 − 𝑒 −𝑦 )

Therefore cos 𝑧 = cos 𝑥 cosh 𝑦 − 𝑖 sin 𝑥 sinh 𝑦

Similar the other properties
General formulas for the real trigonometric functions continue to hold for complex values. This
follows immediately from the definitions. We mention in particular the addition rules
cos(𝑧1 ± 𝑧2 ) = cos 𝑧1 cos 𝑧2 ∓ 𝑖 sin 𝑧1 sin 𝑧2
sin(𝑧1 ± 𝑧2 ) = sin 𝑧1 cos 𝑧2 ± 𝑖 sin 𝑧2 cos 𝑧1

Badri A, Moges B. and Teklebrhan B. 226 AKU

 APPLIED MATHEMATICS III

6.5 Hyperbolic and Logarithm function; General power

Overview:

In this section, we are going to deal with the definition and notation of the hyperbolic,
logarithm function and general power by considering various examples.

Section Objectives:

At the end of this subtopic, students will be able to:

 Define and proof the hyperbolic function ;

 Define logarithm function and general power;
 Represent the example of hyperbolic , logarithm function and general power using the
notation;

5.5.1. Hyperbolic functions

we define complex hyperbolic function in the same fashion as the real hyperbolic functions. The
complex hyperbolic cosine and sine are defined by the formulas
1 1
cosh 𝑧 = 2 (𝑒 𝑧 + 𝑒 −𝑧 ), sinh 𝑦 = 2 (𝑒 𝑦 − 𝑒 −𝑦 ) ……………………………………………..(**)

This is suggested by the familiar definitions for a real variable. These functions are entire, with
derivatives
(cosh 𝑧)′ = sinh 𝑦
(sinh 𝑦)′ = cosh 𝑧

Complex Trigonometric and Hyperbolic Functions Are Related. If in (**), we replace 𝑧

by 𝑖𝑧 and then use (*), we obtain cosh 𝑖𝑧 = cos 𝑧, sinh 𝑖𝑧 = 𝑖 sin 𝑧
Similarly, if in (*) we replace z by 𝑖𝑧 and then use (**), we obtain conversely
cos 𝑖𝑧 = cosh 𝑦, sin 𝑖𝑧 = 𝑖 sinh 𝑦
Here we have another case of unrelated real functions that have related complex analogs,
pointing again to the advantage of working in complex in order to get both a more unified

Badri A, Moges B. and Teklebrhan B. 227 AKU

 APPLIED MATHEMATICS III

formalism and a deeper understanding of special functions. This is one of the main reasons for
the importance of complex analysis to the engineer and physicist.
Example 6.22: Real, Imaginary Parts and Absolute value. Show that:-
1) Cosh 𝑧 = cosh 𝑥 cos 𝑦 + 𝑖 sinh 𝑥 sin 𝑦
2) sinh 𝑧 = sinh 𝑥 cos 𝑦 + 𝑖 cosh 𝑥 sin 𝑦
3) cosh(𝑧1 + 𝑧2 ) = cosh 𝑧1 cosh 𝑧2 + 𝑖 sinh 𝑧1 sinh 𝑧2
4) sinh(𝑧1 + 𝑧2 ) = sinh 𝑧1 cosh 𝑧2 ± 𝑖 sinh 𝑧2 cosh 𝑧1
5) cosh2 𝑧 − sinh2 𝑧 = 1
6) |Cosh 𝑧|2 = cosh2 𝑥 + sin2 𝑦
7) |sinh 𝑧|2 = sinh2 𝑥 + sin2 𝑦
Solution: 1) since 𝑧 = 𝑥 + 𝑖𝑦
𝑒 𝑧 +𝑒 −𝑧 𝑒 (𝑥+𝑖𝑦) +𝑒 −(𝑥+𝑖𝑦)
Cosh 𝑧 = =
2 2
1
Cosh 𝑧 = 2 (𝑒 (𝑥+𝑖𝑦) + 𝑒 −(𝑥+𝑖𝑦) )
1 𝑥 1
= 𝑒 (cos 𝑦 + 𝑖 sin 𝑦) + 𝑒 −𝑥 (cos 𝑦 − 𝑖 sin 𝑦)
2 2
1 1
= (𝑒 𝑥 + 𝑒 −𝑥 ) cos 𝑦 + 𝑖(𝑒 𝑥 − 𝑒 −𝑥 ) sin 𝑦.
2 2
1 1
We know from calculus, cosh 𝑥 = 2 (𝑒 𝑥 + 𝑒 −𝑥 ) and sinh 𝑥 = 2 (𝑒 𝑥 − 𝑒 −𝑥 )

Therefore
Cosh 𝑧 = cosh 𝑥 cos 𝑦 + 𝑖 sinh 𝑥 sin 𝑦
2) Since 𝑧 = 𝑥 + 𝑖𝑦
1 𝑧 1
sinh 𝑧 = (𝑒 − 𝑒 −𝑧 ) = (𝑒 (𝑥+𝑖𝑦) − 𝑒 −(𝑥+𝑖𝑦) )
2 2
1 1
= 𝑒 𝑥 (cos 𝑦 + 𝑖 sin 𝑦) − 𝑒 −𝑥 (cos 𝑦 − 𝑖 sin 𝑦)
2 2
1 1
= (𝑒 𝑥 − 𝑒 −𝑥 ) cos 𝑦 + 𝑖(𝑒 𝑥 + 𝑒 −𝑥 ) sin 𝑦.
2 2
1 1
We know from calculus, sinh 𝑥 = 2 (𝑒 𝑥 − 𝑒 −𝑥 ) and cosh 𝑥 = 2 (𝑒 𝑥 + 𝑒 −𝑥 )

Therefore
sinh 𝑧 = sinh 𝑥 cos 𝑦 + 𝑖 cosh 𝑥 sin 𝑦
Similar the other properties, do these?

Badri A, Moges B. and Teklebrhan B. 228 AKU

 APPLIED MATHEMATICS III

6.5.2. Complex Logarithm

We finally introduce the complex logarithm, which is more complicated than the real logarithm
(which it includes as a special case). The natural logarithm of 𝑧 = 𝑥 + 𝑖𝑦 is denoted by ln 𝑧
(sometimes also by 𝑙𝑜𝑔 𝑧) and
is defined as the inverse of the exponential function; that is, 𝑤 = ln 𝑧 is defined for 𝑧 ≠ 0 by the
relation 𝑒 𝑤 = 𝑧.
(Note that 𝑧 = 0 is impossible, since 𝑒 𝑤 ≠ 0 for all 𝑤.) If we set 𝑤 = 𝑢 + 𝑖𝑣 and 𝑧 = 𝑟𝑒 𝑖𝜃 , this
becomes 𝑒 𝑤 = 𝑒 𝑢+𝑖𝑣 = 𝑟𝑒 𝑖𝜃
Now the absolute value of |𝑒 𝑢+𝑖𝑣 | = 𝑒 𝑢 and the argument 𝑣.
These must be equal to the absolute value and argument on this 𝑒 𝑢+𝑖𝑣 = 𝑟𝑒 𝑖𝜃 :
Since 𝑒 𝑖𝜃 = 𝑒 𝑖(𝜃+2𝑛𝜋) for any integer n,
Thus 𝑟𝑒 𝑖(𝜃+2𝑛𝜋) = 𝑒 𝑢+𝑖𝑣
Now 𝑒 𝑢 = 𝑟, 𝑜𝑟 𝑢 = ln 𝑟 and 𝑣 = 𝜃 + 2𝑛𝜋, 𝑛 ∈ ℤ
Hence for any complex 𝑧 ≠ 0, the solution of 𝑒 𝑤 = 𝑧are given by
𝑤 = ln 𝑧 = ln 𝑟 + 𝑖(𝜃 + 2𝑛𝜋), 𝑛 ∈ ℤ
= ln 𝑟 + 𝑖𝑎𝑟𝑔(𝑧), 𝑧 ≠ 0

Now comes an important point (without analog in real calculus). Since the argument of z is
determined only up to integer multiples of 2𝜋. the complex natural logarithm ln 𝑧 (𝑧 ≠ 0) is
infinitely many-valued.
The value of ln 𝑧 corresponding to the principal value 𝐴𝑟𝑔 𝑧 is denoted by 𝐿𝑛 𝑧 (𝐿𝑛 with
capital𝐿) and is called the principal value of ln z.
Thus restricted by – 𝜋 < arg 𝑧 ≤ 𝜋.
ln 𝑧 = ln|𝑧| + 𝑖 𝐴𝑟𝑔 𝑧, 𝑧≠0

Badri A, Moges B. and Teklebrhan B. 229 AKU

 APPLIED MATHEMATICS III

The uniqueness of 𝐴𝑟𝑔 𝑧 for given 𝑧 (𝑧 ≠ 0 ) implies that 𝐿𝑛 𝑧 is single-valued, that is, a
function in the usual sense. Since the other values of 𝑎𝑟𝑔 𝑧 differ by integer multiples of 2𝜋, the
other values of 𝑙𝑛 𝑧 are given by
ln 𝑧 = ln 𝑧 ± 𝑖 2𝑛𝜋, (𝑛 = 1,2, ⋯ )
They all have the same real part, and their imaginary parts differ by integer multiples of 2𝜋.
If 𝑧 is positive real, then 𝐴𝑟𝑔 𝑧 = 𝑜 and ln 𝑧 becomes identical with the real natural logarithm we
know from calculus. if z is negative real (so that the natural logarithm of calculus is not
defined!), then 𝐴𝑟𝑔 𝑧 = 𝜋 and ln 𝑧 = ln|𝑧| + 𝜋𝑖 , (𝑧 negative real).
𝑒 ln 𝑟 = 𝑟 For positive real r we obtain 𝑒 ln 𝑧 = 𝑧 as expected, but since 𝑎𝑟𝑔(𝑒 𝑧 ) = 𝑦 ± 2𝑛𝜋 is
multivalve, so is 𝑎𝑟𝑔(𝑒 𝑧 ) = 𝑧 ± 2𝑛𝜋𝑖, 𝑛 = 0,1, ⋯

6.5.3. General Powers

General Powers of a complex number 𝑧 = 𝑥 + 𝑖𝑦 are defined by the formula
𝑧 𝑐 = 𝑒 𝑐𝑙𝑛 𝑧 (𝑐 𝑐𝑜𝑚𝑝𝑙𝑒𝑥, 𝑧 ≠ 0)
Since 𝑙𝑛 𝑧 is infinitely many-valued, 𝑧 𝑐 will, in general, be multivalued. The particular value
𝑧 𝑐 = 𝑒 𝑐𝑙𝑛 𝑧 is called the principal value of z c .
If 𝑐 = 𝑛 = 1,2,3, . . ., then 𝑧 𝑛 is single-valued and identical with the usual 𝑛th power of 𝑧.
If 𝑐 = −1, −2, . . ., the situation is similar
𝑛 1⁄ )𝑙𝑛 𝑧
If 𝑐 = 1/𝑛 , where 𝑛 = 2,3, … ,, then 𝑧 𝑐 = √𝑧 = 𝑒 ( 𝑛 , (𝑧 ≠ 0),

Example 6.23: find the general power of 𝑖 𝑖

𝜋 𝜋⁄ )=2𝑛𝜋
Solution: 𝑖 𝑖 = 𝑒 𝑖𝑙𝑛 𝑖 = exp(𝑖 ln 𝑖) = exp [𝑖 (2 𝑖 ± 2𝑛𝜋𝑖)] = 𝑒 −( 2 .
−𝜋⁄
The principal value (𝑛 = 0) is 𝑒 2.

Example 6.24: find the general power of (1 + 𝑖)2−𝑖

Solution: by direct calculation and multiplying out in the exponent
1
(1 + 𝑖)2−𝑖 = 𝑒𝑥𝑝[(2 − 𝑖) ln(1 + 𝑖)] = 𝑒𝑥𝑝 [(2 − 𝑖) {ln √2 + 𝜋𝑖 ± 2𝑛𝜋𝑖}]
4
𝜋⁄ ±2𝑛𝜋 1 1
= 2𝑒 4 [sin (2 ln 2) + 𝑖 cos (2 ln 2)]

Badri A, Moges B. and Teklebrhan B. 230 AKU

 APPLIED MATHEMATICS III

Unit Summary: of chapter 6

1. A complex number 𝒛 is an order pair (𝑥, 𝑦) of real number x and y that is 𝑧 =

(𝑥, 𝑦)𝑥 is called the real part and y is called the imaginary part of z, written 𝑥 =
𝑅𝑒 𝑧 and 𝑦 = 𝐼𝑚 𝑧.
2. The 𝑥𝑦 plane in which the complex numbers are represented in this way is called
the complex plane
3. Complex Conjugate Numbers: The complex conjugate 𝑧̅ of a complex
number 𝑧 = 𝑥 + 𝑖𝑦 is defined by 𝑧̅ = 𝑥 − 𝑖𝑦
4. Polar Form of Complex Numbers: Polar coordinates 𝜃 defined by
𝑥 = 𝑟 cos 𝜃 , 𝑦 = 𝑟 sin 𝜃 then 𝑧 = 𝑥 + 𝑖𝑦 hence 𝑧 = 𝑟(𝑐𝑜𝑠𝜃 + 𝑖𝑠𝑖𝑛𝜃) is called
polar form 𝑟 is called the absolute value or modulus of 𝑧 and is denoted by |𝑧|.
Hence |𝑧| = 𝑟 = √𝑥 2 + 𝑦 2 = √𝑧𝑧̅
5. The 𝒏𝒕𝒉 𝒓𝒐𝒐𝒕 of any complex number 𝑧 can expressed as
𝜃0 +2𝑘𝜋
1⁄
𝑧 𝑛 = 𝜔𝑘+1 = 𝑒 𝑖( 𝑛
)
, 𝑘 = 0, ±1, ±2, ⋯ Where |𝑧| = 𝑟 and 𝜃0 =
𝑚⁄ 𝑚 𝜃0 +2𝑘𝜋
= ( √|𝑧|) 𝑒 𝑖𝑚( )
𝑛
𝐴𝑟𝑔(𝑧) or 𝑧 𝑛 𝑛 𝑚 = 1,2, ⋯ 𝑎𝑛𝑑 𝑘 = 0,1,2, ⋯ , 𝑛 − 1

6. A sequence of complex numbers {𝑧𝑛 }1∞ is said to have the limit 𝒛𝟎 or to converges
to 𝒛𝟎 , and we write lim 𝑧𝑛 = 𝑧0 Or equivalently, 𝑧𝑛 → 𝑧0 as 𝑛 → ∞ if for any
𝑛→∞

𝜀 > 0 there exists an integer N such that |𝑧𝑛 − 𝑧0 | < 𝜀 for all 𝑛 > 𝑁.
7. Let 𝑓 be a function defined in some neighborhood of 𝑧0 itself. We say that the limit
of 𝑓(𝑧) as 𝑧 approaches 𝑧0 is the number 𝑤0 and write lim 𝑓(𝑧) = 𝑤0 Or
𝑧→𝑧0

equivalently, 𝑓(𝑧) → 𝑤0 as 𝑧 → 𝑧0 if for any 𝜀 > 0 there exists a positive number

𝛿 such that |𝑓(𝑧) − 𝑤0 | < 𝜀 whenever 0 < |𝑧 − 𝑧0 | < 𝛿 .

Badri A, Moges B. and Teklebrhan B. 231 AKU

 APPLIED MATHEMATICS III

8. Let 𝑓 be a function defined in some neighborhood of 𝑧0 . Then 𝑓 is continuous at

𝑧0 if lim 𝑓(𝑧) = 𝑓(𝑧0 ) In other words, for 𝑓 to be continuous at 𝑧0 , it must have
𝑧→𝑧0

a limiting value at 𝑧0 and this limiting value must be 𝑓(𝑧0 ). A function 𝑓 is said to
be continuous on a set 𝑆 if it is continuous at each point of 𝑆.
9. The derivative of a complex function 𝑤 = 𝑓(𝑧) at a fixed point 𝑧0 is written
𝑓′(𝑧0 ) and is defined by
𝑓(𝑧0 + ∆𝑧) − 𝑓(𝑧0 )
𝑓′(𝑧0 ) = lim
∆𝑧→0 ∆𝑧

Provided this limit exists. Then 𝑓 is said to be differentiable at 𝑧0 .

10. Analytic function is a function 𝑤 = 𝑓(𝑧) is said to be analytic at a point 𝑧0 if it is
defined at the point and 𝑓 ′ (𝑧) exists at every point in some neighborhood of 𝑧0 .
That is Analytic Function is always Differentiability but Differentiability may or
may not be Analytic Function
11. Cauchy-Riemann Equation: Let us consider the function 𝑤 = 𝑓(𝑧) = 𝑢(𝑥, 𝑦) +
𝑖𝑣(𝑥, 𝑦) defined in the neighborhood of some point 𝑧. if it has satisfied the
𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑣
following conditions = , = − 𝜕𝑦,
𝜕𝑥 𝜕𝑦 𝜕𝑦

12. Cauchy-Riemann equations in polar coordinates

𝜕𝑢 𝜕𝑣 𝜕𝑢 𝜕𝑣
𝑓 ′ (𝑧) = (cos 𝜃 − 𝑖 sin 𝜃) ( 𝜕𝑟 + 𝑖 𝜕𝑟 ) = 𝑒 −𝑖𝜃 (𝜕𝑟 + 𝑖 𝜕𝑟 ) and the Cauchy-
𝜕𝑢 1 𝜕𝑣 1 𝜕𝑢 𝜕𝑣
Riemann equations assume the function = 𝑟 𝜕𝜃, = − 𝜕𝑟
𝜕𝑟 𝑟 𝜕𝜃

13. Laplace’s equation: If 𝑓(𝑧) = 𝑢(𝑥, 𝑦) + 𝑖𝑣(𝑥, 𝑦) is analytic in a domain D, then

both 𝑢 and 𝑣 satisfy Laplace’s equation.∇2 u = uxx + uyy = 0 and ∇2 v = vxx +
vyy = 0 in D and have continuous second partial derivatives in D.
14. Complex exponential function is defined by 𝑒 𝑧 = 𝑒 𝑥+𝑖𝑦 = 𝑒 𝑥 (cos 𝑦 + 𝑖 sin 𝑦).
15. complex trigonometric functions is defined by
1
cos 𝑧 = 2 (𝑒 𝑖𝑧 + 𝑒 −𝑖𝑧 ) = cos 𝑥 cosh 𝑦 − 𝑖 sin 𝑥 sinh 𝑦
1
sin 𝑧 = 2𝑖 (𝑒 𝑖𝑧 − 𝑒 −𝑖𝑧 ) = sin 𝑥 cosh 𝑦 + 𝑖 cos 𝑥 sinh 𝑦

16. The complex hyperbolic cosine and sine are defined by the formulas
1 1
cosh 𝑧 = 2 (𝑒 𝑧 + 𝑒 −𝑧 ) = cos 𝑖𝑧, sinh 𝑦 = 2 (𝑒 𝑦 − 𝑒 −𝑦 ) = −𝑖 sin 𝑖𝑧

Badri A, Moges B. and Teklebrhan B. 232 AKU

 APPLIED MATHEMATICS III

17. The natural logarithm is defined by

ln 𝑧 = ln|𝑧| + 𝑖 𝐴𝑟𝑔 𝑧 ± 𝑖 2𝑛𝜋,, Where 𝑧 ≠ 0 𝑎𝑛𝑑 𝑛 = 1,2,3, … and
𝐴𝑟𝑔 𝑧 is principal value of arg z. That is restricted by – 𝜋 < Arg 𝑧 ≤ 𝜋.
18. General Powers are defined by 𝑧 𝑐 = 𝑒 𝑐𝑙𝑛 𝑧 (𝑐 𝑐𝑜𝑚𝑝𝑙𝑒𝑥, 𝑧 ≠ 0)

Miscellaneous Exercises

1. Verify that each of the two numbers 𝑧 = 1 ± 𝑖 satisfies the equation 𝑧 2 − 2𝑧 + 2 = 0

2. Divide 15 + 23𝑖 𝑏𝑦 − 3 + 7𝑖.
3. Find , in form 𝑥 + 𝑦𝑖
a. (2 + 3𝑖)2
b. (1 − 𝑖)10
c. √𝑖
d. 𝑒 𝜋𝑖/2
4. Represent in polar form, with the principal argument.
a) −4 − 4𝑖
b) −15𝑖
c) 12 + 𝑖
d) 0.6+0.8i
5. Find the principal argument 𝐴𝑟𝑔 𝑧 when
𝑖
I. 𝑧 = −2 − 2𝑖

II. 𝑧 = (√3 − 𝑖)6

6. Find a roots and graph all values of :
a. √81
b. √−32𝑖
3
c. √1
4
d. √−1
7. State the Cauchy Riemann equation and proof.
8. Find 𝑓(𝑧) = 𝑢(𝑥, 𝑦) + 𝑖𝑣(𝑥, 𝑦) as 𝑢 or 𝑣 are given

Badri A, Moges B. and Teklebrhan B. 233 AKU

 APPLIED MATHEMATICS III

a) 𝑢 = 𝑥𝑦
b) 𝑣 = −𝑒 −2𝑥 sin 2𝑦
c) 𝑣 = 𝑦/(𝑥 2 + 𝑦 2 )
d) 𝑢 = cos 3𝑥 cosh 3𝑦
9. Find all values of 𝑧 such that
I. 𝑒 𝑧 = −2
II. 𝑒 𝑧 = 1 + √3𝑖
III. exp(2𝑧 − 1) = 1
10. Find the value of:
a. cos(3 − 𝑖)
b. 𝑡𝑎𝑛 𝑖
c. sinh(1 + 𝜋𝑖)
d. cosh(𝜋 + 𝜋𝑖)
e. ln(0.6 + 0.8𝑖)
11. Show that
a. exp(2 ± 3𝜋𝑖) = −𝑒 2
b. exp(𝑧 + 𝜋𝑖) = − exp 𝑧
𝜋𝑖 𝑒
c. exp (2 + 4 ) = √2 (1 + i)

12. Show that |exp(𝑧 2 )| ≤ exp|z|2

13. Show that
𝜋
a. log(−𝑒𝑖) = 1 − 2 𝑖
1 π
b. log(1 − 𝑖) = 2 ln 2 − 4 i

Badri A, Moges B. and Teklebrhan B. 234 AKU

 APPLIED MATHEMATICS III

References:

 Ahlfors , L. V., Complex Analysis. 3rd ed. New York: McGraw-Hill, 1979.
 Bieberbach, L., Conformal Mapping. Providence, RI: American Mathematical Society,
2000.
 Henrici, P., Applied and Computational Complex Analysis. 3 vols. New York: Wiley,
1993.
 Hille, E., Analytic Function Theory. 2 vols. 2nd ed. Providence, RI: American
Mathematical Society, Reprint V1 1983, V2 2005.
 Knopp, K., Elements of the Theory of Functions. New York: Dover, 1952.
 Knuth, D. E., the Art of Computer Programming. 3vols. 3rd ed. Reading, MA: Addison-
Wesley, 1997–2009.
 Kreyszig, E., Introductory Functional Analysis with Applications. New York: Wiley,
1989.
 Kreyszig, E., on methods of Fourier analysis in multigrid theory. Lecture Notes in Pure
and Applied Mathematics 157. New York: Dekker, 1994, pp. 225–242.
 Kreyszig, E., Basic ideas in modern numerical analysis and their origins. Proceedings of
the Annual Conference of the Canadian Society for the History and Philosophy of
Mathematics. 1997, pp. 34–45.
 Kreyszig, E., and J. Todd, QR in two dimensions. Elemente der Mathematik 31 (1976),
pp. 109–114.
 Mortensen, M. E., Geometric Modeling. 2nd ed. New York: Wiley, 1997.
 Morton, K. W., and D. F. Mayers, Numerical Solution of Partial Differential Equations:
An Introduction. New York: Cambridge University Press, 1994.

Badri A, Moges B. and Teklebrhan B. 235 AKU

 APPLIED MATHEMATICS III

 Ortega, J. M., Introduction to Parallel and Vector Solution of Linear Systems. New
York: Plenum Press, 1988.
 Overton, M. L., Numerical Computing with IEEE Floating Point Arithmetic.
Philadelphia: SIAM, 2004.
 Press, W. H. et al., Numerical Recipes in C: The Art of Scientific Computing. 2nd ed.
New York: Cambridge University Press, 1992.

CHAPTER 7
COPLEX INTEGRALS
Introduction

Chapter 7 laid the groundwork for the study of complex analysis, covered complex numbers in
the complex plane, limits, and differentiation, and introduced the most important concept of
analyticity. A complex function is analytic in some domain if it is differentiable in that domain.
Complex analysis deals with such functions and their applications. The Cauchy–Riemann
equations and also analytic functions satisfy Laplace’s equation. Furthermore, the Cauchy
integral formula shows the surprising result that analytic functions have derivatives of all orders.
Hence, in this respect, complex analytic functions behave much more simply than real-valued
functions of real variables, which may have derivatives, only up to a certain order. Complex
integration is attractive for several reasons. Some basic properties of analytic functions are
difficult to prove by other methods. This includes the existence of derivatives of all orders just
discussed. A main practical reason for the importance of integration in the complex plane is that
such integration can evaluate certain real integrals that appear in applications and that are not
accessible by real integral calculus.

Unit Objectives:

On the completion of this unit, students should be able to:

 Understand the definition of a complex integrals

 Understand the definition of line integral in the complex plane;

Badri A, Moges B. and Teklebrhan B. 236 AKU

 APPLIED MATHEMATICS III

 Understand the Cauchy’s integral theorem and formula;

 Understand the relation between Cauchy’s integral theorem and Cauchy’s integral formul

7.1. Line Integral in the complex plane

Overview:

In this section, we are going to deal with the definition and notation of the complex integrals

by considering various theorems.

Section Objectives:

At the end of this subtopic, students will be able to:

 define a line integral;

 represent a line integral using the notation and examples ;
 Evaluating the line integral.

As in calculus, in complex analysis we distinguish between definite integrals and indefinite

integrals or anti derivatives. Here an indefinite integral is a function whose derivative equals a
given analytic function in a region. By inverting known differentiation formulas we may find
many types of indefinite integrals.
Complex definite integrals are called (complex) line integrals. They are written

∫ f(z)dz.
C

Badri A, Moges B. and Teklebrhan B. 237 AKU

 APPLIED MATHEMATICS III

Here the integrand 𝑓(𝑧) is integrated over a given curve 𝐶 or a portion of it (an 𝑎𝑟𝑐, but we
shall say “curve” in either case, for simplicity). This curve 𝐶 in the complex plane is called the
path of integration. We may represent 𝐶 by a parametric representation
𝑧(𝑡) = 𝑥(𝑡) = 𝑖𝑦(𝑡) (𝑎 ≤ 𝑡 ≤ 𝑏) … … … … … … … … … … … … … … … … … … … … … … . . (1)
The sense of increasing 𝑡 is called the positive sense on 𝐶, and we say that 𝐶 is oriented
by (1).
We assume 𝐶 to be a smooth curve, that is, 𝐶 has a continuous and nonzero derivative
𝑑𝑧
𝑧 ′ (𝑡) = = 𝑥′(𝑡) + 𝑖𝑦′(𝑡)
𝑑𝑡
at each point. Geometrically this means that 𝐶 has everywhere a continuously turning tangent, as
follows directly from the definition
𝑧(𝑡 + ∆𝑡 − 𝑧(𝑡))
𝑧 ′ (𝑡) = lim
∆𝑡→0 ∆𝑡
Here we use a dot since a prime ' denotes the derivative with respect to 𝑧.
Definition of the Complex Line Integral
This is similar to the method in calculus. Let 𝐶 be a smooth curve in the complex plane given by
(1), and let 𝑓(𝑧) be a continuous function given (at least) at each point of 𝐶. We now subdivide
(we “partition”) the interval 𝑎 ≤ 𝑡 ≤ 𝑏 in (1) by points
t 0 (= a), t1 , … , t n−1 , t n (= b)
Where t 0 < t1 < t 2 … < t n . To this subdivision there corresponds a subdivision of 𝐶 by points
z0 , z1 , z2 … , zn−1 , zn (= z)

Fig.7.1. Tangent vector 𝑧′(𝑡) of a curve C in the Fig. 7.2.Complex line integral

Badri A, Moges B. and Teklebrhan B. 238 AKU

 APPLIED MATHEMATICS III

Complex plane given by 𝑧(𝑡) the arrowhead on the curve indicates the positive sense (sense of
increasing 𝑡)
Wherezj = z(t j ). On each portion of subdivision of 𝐶 we choose an arbitrary point, say, a point
ζ1 between z0 and z1 (that is, ζ1 = z(t)where 𝑡 satisfiest 0 ≤ t ≤ t1 ), a point ζ2 between z1
and z2 etc. Then we form the sum
𝑠𝑛 = ∑𝑛 𝑓(ζm )∆zm where ∆zm = zm − zm−1 …………(2)
We do this for each 𝑛 = 2,3, ⋯ in a completely independent manner, but so that the greatest
|∆𝑡𝑚 | = |t m − t m−1 | approaches zero as 𝑛 → ∞. This implies that the greatest |∆𝑧𝑚 |also
approaches zero. Indeed, it cannot exceed the length of the arc of 𝐶 from zm−1 to zm and the
latter goes to zero since the arc length of the smooth curve C is a continuous function of 𝑡. The
limit of the sequence of complex numbers s2 , s3 … thus obtained is called the line integral (or
simply the integral) of 𝑓(𝑧) over the path of integration C with the orientation given by (1).
This line integral is denoted by

∫ 𝑓(𝑧)𝑑𝑧 𝑜𝑟 ∮ 𝑓(𝑧)𝑑𝑧 … … … … … … … … … … … … … … … … … … … . . (3)

𝐶 𝐶

If 𝐶 is a closed path (one whose terminal point 𝑍 coincides with its initial point z0 , as
for a circle or for a curve shaped ).
General Assumption: - All paths of integration for complex line integrals are assumed to
be piecewise smooth, that is, they consist of finitely many smooth curves joined end to end.

Basic Properties Directly Implied by the Definition

1. Linearity: - Integration is a linear operation, that is, we can integrate sums term by term
and can take out constant factors from under the integral sign. This means that if the
integrals of 𝑓1 and 𝑓2 over a path 𝐶 exist, so does the integral of 𝑘1 𝑓1 + 𝑘2 𝑓2 over the
same path and

∫[𝑘1 𝑓1 (z) + 𝑘2 𝑓2 (𝑧)]𝑑𝑧 = 𝑘1 ∫ 𝑓1 (z)𝑑𝑧 + 𝑘2 ∫ 𝑓2 (z)𝑑𝑧 … … … … … … . . (4)

𝐶 𝐶 𝐶

2. Sense reversal in integrating over the same path, from z0 to 𝑍 (left) and from Z to z0
(right), introduces a minus sign as shown

Badri A, Moges B. and Teklebrhan B. 239 AKU

 APPLIED MATHEMATICS III

z z0

∫ 𝑓(𝑧)𝑑𝑧 = − ∫ 𝑓(𝑧)𝑑𝑧 … … … … … … … … … … … … … … … … … (5)

z0 z

3. Partitioning of path (see Fig. 7.3.)

∫ 𝑓(𝑧)𝑑𝑧 = ∫ 𝑓(𝑧)𝑑𝑧 + ∫ 𝑓(𝑧)𝑑𝑧 … … … … … … … … … … … … … … … . (6)

𝐶 𝐶1 𝐶2

Fig.7.3. Partitioning of path [formula (6)]

Existence of the Complex Line Integral
Our assumptions that are 𝑓(𝑧) is continuous and 𝐶 is piecewise smooth imply the existence of
the line integral (3). This can be seen as follows.
As in the preceding chapter let us write 𝑓(𝑧) = 𝑢(𝑥, 𝑦) + 𝑖𝑣(𝑥, 𝑦). we also set
𝜁𝑚 = 𝜁𝑚 + 𝑖𝜂𝑚 𝑎𝑛𝑑 ∆𝑧𝑚 = ∆𝑥𝑚 + 𝑖∆𝑦𝑚 .
Then (2) may be written
𝑛

𝑠𝑛 = ∑(𝑢 + 𝑖𝑣)(∆𝑥𝑚 + 𝑖∆𝑦𝑚 ) … … … … … … … … … … . . (7)

Where 𝑢 = 𝑢(𝜁𝑚 , 𝜂𝑚 ), 𝑣 = 𝑣(𝜁𝑚 , 𝜂𝑚 ) and we sum over 𝑚 from 1 to 𝑛. Performing the

multiplication, we may now split up 𝑠𝑛 into four sums:

𝑠𝑛 = ∑ 𝑢∆𝑥𝑚 − ∑ 𝑣∆𝑦𝑚 + 𝑖 [∑ 𝑢∆𝑦𝑚 − ∑ 𝑣∆𝑥𝑚 ]

These sums are real. Since 𝑓 is continuous, 𝑢 and 𝑣 are continuous. Hence, if we let n approach
infinity in the aforementioned way, then the greatest ∆𝑥𝑚 and ∆𝑦𝑚 will approach zero and each
sum on the right becomes a real line integral:

lim 𝑠𝑛 = ∫ 𝑓(𝑧)𝑑𝑧
𝑛→∞
𝐶

Badri A, Moges B. and Teklebrhan B. 240 AKU

 APPLIED MATHEMATICS III

= ∫ 𝑢𝑑𝑥 − ∫ 𝑣𝑑𝑦 + 𝑖 [∫ 𝑢𝑑𝑦 + ∫ 𝑣𝑑𝑥 ] … … … … … … . . (8)

𝐶 𝐶 𝐶 𝐶

This shows that under our assumptions on 𝑓 and 𝐶 the line integral (3) exists and its value is
independent of the choice of subdivisions and intermediate points 𝜁𝑚 .
Theorem 1: (Indefinite Integration of Analytic Functions)
Let 𝑓(𝑧) be analytic in a simply connected domain D. Then there exists an indefinite integral of
𝑓(𝑧) in the domain D, that is, an analytic function 𝐹(𝑧)such that 𝐹 ′ (𝑧) = 𝑓(𝑧) in D, and for all
paths in 𝐷 joining two points z0 and z1 in D. We have
z1

∫ 𝑓(𝑧)𝑑𝑧 = 𝐹(z1 ) − 𝐹( z0 ) [𝐹 ′ (𝑧) = 𝑓(𝑧)] … … … … … … … … … . (9)

z0

(Note that we can write z0 and z1 instead of 𝐶, since we get the same value for all those 𝐶 from
to z0 𝑡𝑜 z1 .)

This theorem will be proved in the next section.

Simple connectedness is quite essential in Theorem 1, since analytic functions are our main
concern, and since differentiation formulas will often help in finding 𝐹(𝑧)for a given
𝐹 ′ (𝑧) = 𝑓(𝑧) the present method is of great practical interest.
If 𝑓(𝑧) is entire we can take for 𝐷 the complex plane (which is certainly simply connected)
1+𝑖
1+i 2 1 1 2 2
Example 7.1: 1) ∫0 𝑧 𝑑𝑧 = 3 𝑧 2 ⎥0 = 3 (1 + 𝑖)2 = − 3 + 3 𝑖
𝜋𝑖
πi
2) ∫−πi cos 𝑧 𝑑𝑧 = sin 𝑧 ⎥−𝜋𝑖 = 2 sin 𝜋𝑖 = 2𝑖 sinh 𝜋 = 23.097𝑖

Theorem 2: (Integration by the Use of the Path)

Let C be a piecewise smooth path, represented by 𝑧 = 𝑧(𝑡), where𝑎 ≤ 𝑡 ≤ 𝑏. Let 𝑓(𝑧)be a
continuous function on C. Then
𝑏
𝑑𝑧
∫ 𝑓(𝑧)𝑑𝑧 = ∫ 𝑓[𝑧(𝑡)]𝑧′(𝑡)𝑑𝑡 (𝑧 ′ = ) … … … … … … … … … … … … … . . (10)
𝑑𝑡
𝐶 𝑎

Badri A, Moges B. and Teklebrhan B. 241 AKU

 APPLIED MATHEMATICS III

PROOF: The left side of (10) is given by (8) in terms of real line integrals and we show that the
right side of (10) also equals (8). We have 𝑧 = 𝑥 + 𝑖𝑦, hence 𝑧 ′ = 𝑥 ′ + 𝑖𝑦. We simply write 𝑢
for 𝑢[𝑥(𝑡, 𝑦(𝑡))]and 𝑣 for 𝑣[𝑥(𝑡, 𝑦(𝑡))] . We also have 𝑑𝑥 = 𝑥 ′ 𝑑𝑡 and 𝑑𝑦 = 𝑦 ′ 𝑑𝑡.
Consequently, in (10)
𝑏 𝑏

∫ 𝑓[𝑧(𝑡)]𝑧 ′ (𝑡)𝑑𝑡 = ∫(𝑢 + 𝑖𝑣)(𝑥 ′ + 𝑖𝑦 ′ )𝑑𝑡

𝑎 𝑎

= ∫[𝑢𝑑𝑥 − 𝑣𝑑𝑦 + 𝑖(𝑢𝑑𝑦 + 𝑣𝑑𝑥)]

𝐶

= ∫[𝑢𝑑𝑥 − 𝑣𝑑𝑦] 𝑖 = ∫[𝑢𝑑𝑦 + 𝑣𝑑𝑥]

𝐶 𝐶

Dependence on path: Now comes a very important fact. If we integrate a given function
𝑓(𝑧)from a point z0 to a point z1 along different paths, the integrals will in general have
different values. In other words, a complex line integral depends not only on the endpoints of
the path but in general also on the path itself. The next example gives a first impression
of this, and a systematic discussion follows in the next section.
Example 7.2:: Integral of a Non analytic Function. Dependence on Path
Integrate 𝑓(𝑧) = 𝑅𝑒𝑧 = 𝑥 from 0 to 1 + 2𝑖
(a) Along 𝐶 ∗ in Fig. 7.4,
(b) along C consisting of C1 and C2 .
Solution: (a) 𝐶 ∗ can be represented by 𝑧(𝑡) = 𝑡 + 2𝑖𝑡 (0 ≤ 𝑡 ≤ 1). Hence 𝑧 ′ (𝑡) = 1 + 2𝑖 and
𝑓[𝑧(𝑡)] = 𝑥(𝑡 = 𝑡) on 𝐶 ∗ . We now calculate
1
∫𝐶 ∗ 𝑅𝑒 𝑧𝑑𝑧 = ∫0 𝑡(1 + 2𝑖)𝑑𝑡
1 1
= 2 (1 + 2𝑖) = 2 + 𝑖

Fig. 7.4

Badri A, Moges B. and Teklebrhan B. 242 AKU

 APPLIED MATHEMATICS III

b) we now have
C1 : z(t) = t, 𝑧 ′ (𝑡) = 1, 𝑓(𝑧(𝑡)) = 𝑥(𝑡) = 𝑡 (0 ≤ 𝑡 ≤ 1)
C2 : z(t) = t + it, 𝑧 ′ (𝑡) = 𝑖, 𝑓(𝑧(𝑡)) = 𝑥(𝑡) = 1 (0 ≤ 𝑡 ≤ 2).
using (6) we calculate
1 2 1
∫𝐶 𝑅𝑒 𝑧 𝑑𝑧 = ∫𝐶 𝑅𝑒 𝑧 𝑑𝑧 + ∫𝐶 𝑅𝑒 𝑧 𝑑𝑧 = ∫0 𝑡𝑑𝑡 + ∫0 1. 𝑖𝑑𝑡 2 + 2𝑖.
1 2

Note that this result differs from the result in (a).

Bounds for Integrals (ML-Inequality)

There will be a frequent need for estimating the absolute value of complex line integrals. The
basic formula is

|∫ 𝑓(𝑧)𝑑𝑧| ≤ 𝑀𝐿 (𝑀𝐿 − inequality) … … … … … … … … … … . (𝟏𝟏)

𝐶

L is the length of C and M a constant such that |𝑓(𝑧)| ≤ 𝑀 everywhere on C.

Example 7.3: Estimation of an Integral
Find an upper bound for the absolute value of the integral

∫ 𝑧 2 𝑑𝑧,
C

C the straight-line segment from 0 𝑡𝑜 1 + 𝑖, Fig 7.5

Fig 7.5

Solution: 𝐿 = √2 𝑎𝑛𝑑 |𝑓(𝑧)| = |𝑧 2 | ≤ 2 on C gives by (11)

|∫ 𝑧 2 𝑑𝑧| ≤ 2√2 = 2.8284

C

Badri A, Moges B. and Teklebrhan B. 243 AKU

 APPLIED MATHEMATICS III

Example 7.4: Evaluate ∫C 𝑧 2 𝑑𝑧, where C is the straight line joining the origin o to the point P (2,
1) in the complex plane.

Solution: the equation of the line op is 𝑥 = 2𝑦, 0≤𝑦≤1

Thus 𝑑𝑧 = 𝑑𝑥 + 𝑖𝑑𝑦 = 2𝑑𝑦 + 𝑖𝑑𝑦 = (2 + 𝑖)𝑑𝑦

Also, 𝑧 2 = (𝑥 + 𝑖𝑦)2 = 𝑥 2 − 𝑦 2 + 2𝑖𝑥𝑦 = 3𝑦 2 + 4𝑖𝑦 2

Hence

∫ 𝑧 2 𝑑𝑧 = ∫(3 + 4𝑖) 𝑦 2 (2 + 𝑖)𝑑𝑦

C 0

1 1
= (𝟐 + 𝟏𝟏) ∫0 𝑦 2 𝑑𝑦 = 3 (2 + 11𝑖)

1+𝑖
Example 7.5: Evaluate the integral ∫0 (𝑥 − 𝑦 + 𝑖𝑥 2 )𝑑𝑧
a) along the straight line from 𝑧 = 0 to 𝑧 = 1 + 𝑖
b) along the real axis from 𝑧 = 0 to 𝑧 = 1 and then along a line parallel to imaginary axis
from 𝑧 = 1 to 𝑧 = 𝑥 + 𝑖𝑦

P (1, 1)
Y=x

o
M (1, 0)
o

Fig. 7.6
Solution: a) the equation of the straight line 𝑜𝑝, refers the fig.7.6 is 𝑦 = 𝑥, thus along the line
𝑜𝑝, 𝑧 = 𝑥 + 𝑖𝑦 = 𝑥 + 𝑖𝑥 = (1 + 𝑖)𝑥, which gives 𝑑𝑧 = (1 + 𝑖)𝑑𝑥, 0 ≤ 𝑥 ≤ 1,
And hence

Badri A, Moges B. and Teklebrhan B. 244 AKU

 APPLIED MATHEMATICS III

1+𝑖 1

∫ (𝑥 − 𝑦 + 𝑖𝑥 2 )𝑑𝑧 = ∫(𝑥 − 𝑥 + 𝑖𝑥 2 )(1 + 𝑖)𝑑𝑥

0 0
1
= 𝑖(1 + 𝑖) ∫0 𝑥 2 𝑑𝑥

1
= − 3 (1 − 𝑖)

b) Along the path 𝑂𝑀, we have 𝑦 = 0 and thus 𝑧 = 𝑥 + 𝑖𝑦 = 𝑥 and hence 𝑑𝑧 = 𝑑𝑥, 0 ≤ 𝑥 ≤ 1
Also, along the path MP, we have 𝑥 = 1 and thus 𝑧 = 𝑥 + 𝑖𝑦 = 1 + 𝑖𝑦, and hence 𝑑𝑧 = 𝑖𝑑𝑦,
0 ≤ 𝑦 ≤ 1.
Therefore, the line integral
1+𝑖 1 1
2 2
∫ (𝑥 − 𝑦 + 𝑖𝑥 )𝑑𝑧 = ∫(𝑥 + 𝑖𝑥 )𝑑𝑥 + ∫(1 − 𝑦 + 𝑖)𝑖𝑑𝑦
0 0 0
1 1
𝑥2 𝑖𝑥 3 𝑦2
=[2 + ] + [(𝑖 − 1)𝑦 − 𝑖 ]
3 0 2 0
1 𝑖 𝑖 1 5
= 2 + 3 + (𝑖 − 1) − 2 = − 2 + 6 𝑖

Exercise: Evaluate ∮𝐶 |𝑧 2 |2 𝑑𝑧 around the square with vertices at (0, 0), (1, 0), (1, 1), (0, 1)

C (0, 1)
B (1, 1)

x
O (0, 0) A (1, 0)

7.2 Cauchy’s Integral Theorem and Formula

Overview:

In this section, we are going to deal with state and proof of Cauchy’s Integral Theorem and
Cauchy’s Integral Formula and given by examples.

Badri A, Moges B. and Teklebrhan B. 245 AKU

 APPLIED MATHEMATICS III

Section Objectives:

At the end of this subtopic, students will be able to:

 state Cauchy integral theorem and formula;

 represent an example of the theorem;

Let us continue our discussion of simple connectedness which we started here:

1. A simple closed path is a closed path that does not intersect or touch itself as shown in
Fig7.6. For example, a circle is simple, but a curve shaped like an 8 is not simple.

Simple Simple Not simple Not simple

Fig.7.6. closed paths

2. A simply connected domain D in the complex plane is a domain such that every simple
closed path in D encloses only points of D. Examples: The interior of a circle (“open disk”),
ellipse, or any simple closed curve. A domain that is not simply connected is called multiply
connected. Examples: An annulus, a disk without the center, for example, 0 < |𝑧| < 1. See also
Fig. 7.7.
More precisely, a bounded domain D (that is, a domain that lies entirely in some circle about
the origin) is called p-fold connected if its boundary consists of p closed

Simply simply doubly triply

Badri A, Moges B. and Teklebrhan B. 246 AKU

 APPLIED MATHEMATICS III

Connected connected connected connected

Fig.7.7. Simply and multiply connected domains
Note: connected sets without common points. These sets can be curves, segments,
or single points (such as𝑧 = 0 for 0 < |𝑧| < 1, for which 𝑝 = 2 ). Thus,
D has 𝑝 − 1 “holes,” where “hole” may also mean a segment or even a single point.
Hence an annulus is doubly connected (𝑝 = 2).

7.2.1 Cauchy’s Integral Theorem

Theorem: (Cauchy’s Integral Theorem)

If 𝑓(𝑧) is analytic and 𝑓′(𝑧) is continuous in a simply connected domain in D, then for every piecewise
closed curve C in D the contour integral

∫ 𝑧 2 𝑑𝑧 = 0 … … … … … … … … … … … … … … … … … … … . (∗)
C

Proof: Writing 𝑓(𝑧) = 𝑢 + 𝑖𝑣 and 𝑑𝑧 = 𝑑𝑥 + 𝑖𝑑𝑦,

We have

∮ 𝑓(𝑧)𝑑𝑧 = ∮(𝑢 + 𝑖𝑣)(𝑑𝑥 + 𝑖𝑑𝑦)

𝐶 𝐶

= ∮𝐶(𝑢𝑑𝑥 − 𝑣𝑑𝑦) + 𝑖 ∮𝐶(𝑣𝑑𝑥 + 𝑢𝑑𝑦)

𝜕𝑢 𝜕𝑢 𝜕𝑣 𝜕𝑣
Since 𝑓′(𝑧) is continuous, therefore 𝜕𝑥 , 𝜕𝑦 , 𝜕𝑥 , 𝜕𝑦

Are also continuous in D, and hence in the region enclosed by C. thus Green’s theorem becomes
𝜕𝑢 𝜕𝑢 𝜕𝑢 𝜕𝑣
∮ 𝑓(𝑧)𝑑𝑧 = − ∬( + )𝑑𝑥𝑑𝑦 + 𝑖 ∬( − )𝑑𝑥𝑑𝑦 … … … … … … (∗∗)
𝜕𝑥 𝜕𝑦 𝜕𝑥 𝜕𝑦
𝐶 𝐸 𝐸

Badri A, Moges B. and Teklebrhan B. 247 AKU

 APPLIED MATHEMATICS III

Where E is the region bounded by the closed curve C, since 𝑓(𝑧)is analytic, u and v satisfied the
C-R.E.
And thus the integrands of the two double integrals on the right side of (**) are identically zero
and hence we obtain

∮ 𝑓(𝑧)𝑑𝑧 = 0
𝐶

Note: Analytic of 𝑓(𝑧) is any sufficient but not necessary condition of ∮𝐶 𝑓(𝑧)𝑑𝑧 = 0.
Example 7.6: Evaluate the following integrals by applying Cauchy’s integral theorem, in each
applicable
a) ∮𝐶 cos 𝑧 𝑑𝑧
b) ∮𝐶 sec 𝑧 𝑑𝑧
𝑑𝑧
c) ∮𝐶 𝑧 2 −5𝑧+6

d) ∮𝐶 𝑧̅𝑑𝑧 where C is the unit circle |𝑧| = 1

Solution: a) The integrand 𝑓(𝑧) = cos 𝑧 is analytic for all 𝑧 and also 𝑓 ′ (𝑧) = sin 𝑧 is continuous
every where, and hence on and inside C also, thus by Cauchy’s theorem ∮𝐶 cos 𝑧 𝑑𝑧 = 0
1 𝜋 3𝜋
b. The integrand 𝑓(𝑧) = sec 𝑧 = cos 𝑧 is not analytic at the points 𝑧 = ± 2 , ± , ⋯ but
2

all these points lie outside the unit circle |𝑧| = 1 hence 𝑓(𝑧) is analytic and 𝑓 ′ (𝑧) is
continuous in and on C. thus by Cauchy’s theorem ∮𝐶 sec 𝑧 𝑑𝑧 = 0
1 1
c. The integrand 𝑓(𝑧) = 𝑧 2 −5𝑧+6 = (𝑧−2)(𝑧−3) is analytic every where except at 𝑧 =

2 𝑎𝑛𝑑 𝑧 = 3, the points which lie outside the unit circle |𝑧| = 1 and hence 𝑓(𝑧) is
𝑑𝑧
analytic and 𝑓 ′ (𝑧) is continuous in and on C. thus by Cauchy’s theorem ∮𝐶 𝑧 2 −5𝑧+6 = 0.

d. The integrand 𝑓(𝑧) = 𝑧̅ is analytic and hence the Cauchy’s theorem is not applicable.
In fact, about 𝐶: |𝑧| = 1 we have

2𝜋 2𝜋

∮ 𝑧̅𝑑𝑧 = ∫ 𝑒 −𝑖𝜃 𝑒 𝑖𝜃 𝑑𝜃 = 𝑖 ∫ 𝑑𝜃 = 2𝜋𝑖.

𝐶 0 0

Independence of path

Badri A, Moges B. and Teklebrhan B. 248 AKU

 APPLIED MATHEMATICS III

An integral of 𝑓(𝑧) is independent of path in a domain D, if for very 𝑧1 , 𝑧2 in D the value of

𝑧2
∫𝑧 𝑓(𝑧)𝑑𝑧 dependence only on the end points 𝑧1 𝑎𝑛𝑑 𝑧2 and not on the choice of the path C of
1

the path C joining 𝑧1 𝑡𝑜 𝑧2 .

Theorem: (Independence of path) If 𝑓(𝑧) is analytic in a simple connected domain D, then
∫𝐶 𝑓(𝑧)𝑑𝑧 is independent of the path for every piecewise smooth curve C lying entirely with in
D.
Theorem: (Extension of the Cauchy’s integral theorem) If 𝑓(𝑧) is analytic on and between
two closed paths 𝐶1 and 𝐶2 , then

∫ 𝑓(𝑧)𝑑𝑧 = ∫ 𝑓(𝑧)𝑑𝑧
𝐶1 𝐶2

Theorem: If 𝑓(𝑧) is analytic on between the region included in the closed curves 𝐶1 , 𝐶2 , 𝐶3 etc.
then

∫ 𝑓(𝑧)𝑑𝑧 = ∫ 𝑓(𝑧)𝑑𝑧 + ∫ 𝑓(𝑧)𝑑𝑧 + ∫ 𝑓(𝑧)𝑑𝑧 + ⋯

𝐶 𝐶1 𝐶2 𝐶3

𝐶1

𝐶2
𝐶3
𝐶4

Theorem: (Fundamental theorem of calculus):-

If 𝑓(𝑧) is analytic in a simple connected domain D, and if 𝛼 and 𝛽 are any points in D, then
𝛽

∫ 𝑓(𝑧)𝑑𝑧 = 𝐹( 𝛽) − 𝐹(𝛼), 𝑤ℎ𝑒𝑟𝑒 𝐹 ′ (𝑧) = 𝑓(𝑧)

𝛼

Example 7.7: Evaluate 𝐼 = ∫𝐶 sin 𝑧 𝑑𝑧, where C is composed of the circular arc 𝐶1 and straight
line segment 𝐶2 that connects the point 𝑧 = 0 and 𝑧 = 𝑖𝜋

Badri A, Moges B. and Teklebrhan B. 249 AKU

 APPLIED MATHEMATICS III

Solution: the integrand 𝑓(𝑧) = sin 𝑧 is an entire function. Hence, its integral is independent of
path so we may write
y

𝐶2

𝐶1
𝑖𝜋

∫ sin 𝑧 𝑑𝑧 = ∫ sin 𝑧 𝑑𝑧 = [− cos 𝑧]𝑖𝜋

0 = 1 − cos 𝑖𝜋 = 1 − cosh 𝜋
𝐶 0

1 3 2𝑖
Exercise: 1) ∫0 𝑧 2 𝑒 𝑧 𝑑𝑧 2) ∫0 sinh 𝑧 𝑑𝑧

7.2.2. Cauchy’s Integral Formula

Theorem: (Cauchy’s Integral Formula)

Let 𝑓(𝑧) be analytic in a simply connected domain D. Then for any point in D, then for any point
𝑧𝑜 in D and any simple closed path C in D that encloses 𝑧𝑜

1 𝑓(𝑧)
𝑓(𝑧𝑜 ) = 2𝜋𝑖 ∮𝐶 𝑧−𝑧 𝑑𝑧
𝑜

The integration being taken counter clockwise

𝑧 2 +1
Example 7.8: Evaluate the integral ∮𝐶 𝑧 2−1 𝑑𝑧, 𝐶: |𝑧 − 1| = 1
𝑧 2 +1 𝑧 2 +1⁄𝑧+1
Solution: writing the integrand as = we observe e that 𝑓(𝑧) = 𝑧 2 + 1⁄𝑧 + 1 is
𝑧 2 −1 𝑧−1

analytic on and inside C, and here 𝑧𝑜 = 1, as shown the fig.

Hence by Cauchy’s integral formula

(1, 0)
Badri A, Moges B. and Teklebrhan B. 250 AKU
 APPLIED MATHEMATICS III

𝑧2 + 1
∮ 2 𝑑𝑧 = 2𝜋𝑖𝑓(1) = 2𝜋𝑖
𝑧 −1
𝐶

𝑧 2 +1
Example 7.9: Evaluate the integral ∮𝐶 𝑧(2𝑧−1) 𝑑𝑧 𝐶: |𝑧| = 1

𝑧 2 +1
Solution: Let 𝐼 = ∮𝐶 𝑧(2𝑧−1) 𝑑𝑧

𝑧 2 +1 1
The integrand is not analytic at the point 𝑧 = 0 and 𝑧 = 2 both of which lie inside C,
𝑧(2𝑧−1)

writing it as

𝑧2 + 1 1 1
= (𝑧 2 + 1) [ − ]
𝑧(2𝑧 − 1) 1
(𝑧 − 2) 𝑧

Therefore, using the Cauchy’s integral formula

𝑧2 + 1 𝑧2 + 1 𝑧2 + 1
𝐼=∮ 𝑑𝑧 = ∮ 𝑑𝑧 − ∮ 𝑑𝑧
𝑧(2𝑧 − 1) 1 𝑧
𝐶 𝐶 𝑧 − 𝐶
2

5𝜋𝑖 𝜋𝑖
= 2𝜋𝑖[𝑧 2 + 1]𝑧=1 − 2𝜋𝑖[𝑧 2 + 1]𝑧=0 = − 2𝜋𝑖 =
2 2 2

7.3. The Derivatives of Analytic Functions

Overview:

In this section, we are going to deal with state and proof of Derivatives of an Analytic Function
or Generalized Cauchy’s integral formula and given by examples.

Section Objectives:

At the end of this subtopic, students will be able to:

 State and proof Derivatives of an Analytic Function or Generalized Cauchy’s

integral formula;
 Represent an example of the theorem;

Badri A, Moges B. and Teklebrhan B. 251 AKU

 APPLIED MATHEMATICS III

As mentioned, a surprising fact is that complex analytic functions have derivatives of all orders.
This differs completely from real calculus. Even if a real function is once differentiable we
cannot conclude that it is twice differentiable nor that any of its higher derivatives exist. This
makes the behavior of complex analytic functions simpler than real functions in this aspect. To
prove the surprising fact we use Cauchy’s integral formula.

Theorem: (Derivatives of an Analytic Function or Generalized Cauchy’s integral formula)

If 𝑓(𝑧) is analytic in a domain D. Then it has derivatives of all orders in D, which are then also
analytic in D, and the values of these derivatives at a point 𝑧𝑜 in D are given by the formulas

𝑛! 𝑓(𝑧)
𝑓 (𝑛) (𝑧𝑜 ) = ∮ 𝑑𝑧 , 𝑛 = 1,2,3, …
2𝜋𝑖 (𝑧 − 𝑧𝑜 )𝑛+1
𝐶

Where C is any simple closed path in D and taken in counter-clockwise sense.

Proof: the Cauchy’s integral formula is

1 𝑓(𝑧)
𝑓(𝑧𝑜 ) = ∮ 𝑑𝑧
2𝜋𝑖 𝑧 − 𝑧𝑜
𝐶

Differentiating it under the integral sign with respect to 𝑧𝑜 we obtain

1! 𝑓(𝑧)
𝑓′(𝑧𝑜 ) = ∮ 𝑑𝑧
2𝜋𝑖 (𝑧 − 𝑧𝑜 )2
𝐶

Similarly,
2! 𝑓(𝑧)
𝑓′′(𝑧𝑜 ) = ∮ 𝑑𝑧
2𝜋𝑖 (𝑧 − 𝑧𝑜 )3
𝐶

And in general
𝑛! 𝑓(𝑧)
𝑓 (𝑛) (𝑧𝑜 ) = ∮ 𝑑𝑧
2𝜋𝑖 (𝑧 − 𝑧𝑜 )𝑛+1
𝐶

Example 7.10: Evaluate the integral

𝑒𝑧
∮ 𝑑𝑧 , 𝐶: |𝑧| = 1,
𝑧3
𝐶

Taken in counter clockwise sense

Badri A, Moges B. and Teklebrhan B. 252 AKU

 APPLIED MATHEMATICS III

ez
Solution: Let I = ∮C z3 dz. here 𝑓(𝑧) = 𝑒 𝑧 is analytic in the region bounded by the simple close
1
curve |𝑧| = 1, the singular point 𝑧 = 0 of 𝑧 3 lies inside |𝑧| = 1, hence applying the generalized

Cauchy’s integral formula.

ez 2πi d2 z
I=∮ dz = [e ]z=0 = 𝜋𝑖
z3 2! dz 2
C

Example 7.11: Evaluate the integral

𝑧+1
∮ 𝑑𝑧 , 𝐶: |𝑧 − 3| = 2
𝑧(𝑧 − 2)(𝑧 − 4)3
𝐶

In the counter clockwise sense

𝑧+1
Solution: Let I = ∮𝐶 𝑧(𝑧−2)(𝑧−4)3 𝑑𝑧. the integrand has singularities at 𝑧 = 0, 2 𝑎𝑛𝑑 4 out of these

𝑧 = 2 𝑎𝑛𝑑 4 lies inside C. consider two non intersecting closed contours 𝐶1 𝑎𝑛𝑑 𝐶2, As shown
fig. lying completely with in C, respectively about the point 𝑧 = 2 𝑎𝑛𝑑 𝑧 = 4.

𝑧+1
I=∮ 𝑑𝑧
𝑧(𝑧 − 2)(𝑧 − 4)3
𝐶

𝑧+1 𝑑𝑧 𝑧+1 𝑑𝑧
= ∮[ 3
] + ∮[ ]
𝑧(𝑧 − 4) (𝑧 − 2) 𝑧(𝑧 − 2) (𝑧 − 4)3
𝐶1 𝐶2

= I1 + I 2 , 𝑠𝑎𝑦
Now, using the Cauchy’s integral formula
𝑧+1 𝑑𝑧 𝑧+1 3𝜋𝑖
I1 = ∮ [ ] = 2𝜋𝑖 [ ] = −
𝑧(𝑧 − 4)3 (𝑧 − 2) 𝑧(𝑧 − 4)3 𝑧=2 8
𝐶1

Similarly, using the Cauchy’s integral formula

𝑧+1 𝑑𝑧 2𝜋𝑖 d2 𝑧 + 1 23𝜋𝑖
I2 = ∮[ ] 3
= 2
[ ] =
𝑧(𝑧 − 2) (𝑧 − 4) 2! dz 𝑧(𝑧 − 2) 𝑧=4 64
𝐶2

3𝜋𝑖 23𝜋𝑖 𝜋𝑖
Therefore, I = I1 + I 2 = − + =−
8 64 64

Badri A, Moges B. and Teklebrhan B. 253 AKU

 APPLIED MATHEMATICS III

Unit Summary:

- The complex line integral of a function𝑓(𝑧) taken over a path C is denoted by

∫𝐶 𝑓(𝑧)𝑑𝑧 or, if C is closed, also by ∮𝐶 𝑓(𝑧) (1)

I𝑓(𝑧) is analytic in a simply connected domain D, then we can evaluate (1)

as in calculus by indefinite integration and substitution of limits, that is,

∫ 𝑓(𝑧)𝑑𝑧 = 𝐹(𝑧1 ) − 𝐹(𝑧0 ) [𝐹 ′ (𝑧) = 𝑓(𝑧)]

𝐶

- A general method of integration, not restricted to analytic functions, uses the equation
𝑏 𝑑𝑧
𝑧 = 𝑧(𝑡) of C, where 𝑎 ≤ 𝑡 ≤ 𝑏, ∫𝐶 𝑓(𝑧)𝑑𝑧 = ∫𝑎 𝑓(𝑧(𝑡))𝑧 ′ (𝑡)𝑑𝑡 (𝑧 ′ = )
𝑑𝑡

- Cauchy’s integral theorem is the most important theorem in this chapter. It states that if
𝑓(𝑧)is analytic in a simply connected domain D, then for every closed path C in D,
∮𝐶 𝑓(𝑧)𝑑𝑧 = 0
Under the same assumptions and for any 𝑧0 in D and closed path C in D containing 𝑧0 in
1 𝑓(𝑧)
its interior we also have Cauchy’s integral formula 𝑓(𝑧0 ) = 2𝜋𝑖 ∮𝐶 𝑧−𝑧 𝑑𝑧.
0

Badri A, Moges B. and Teklebrhan B. 254 AKU

 APPLIED MATHEMATICS III

- Generalized Cauchy’s integral formula these assumptions 𝑓(𝑧) have derivatives of all
orders in D that are themselves analytic functions in D. 𝑓 (𝑛) (𝑧0 ) =
𝑛! 𝑓(𝑧)
∮ 𝑑𝑧 (𝑛 = 1, 2, … )
2𝜋𝑖 𝐶 (𝑧−𝑧0 )𝑛+1

This implies Morera’s theorem (the converse of Cauchy’s integral theorem) and
Cauchy’s inequality which in turn implies Liouville’s theorem that an entire function that
is bounded in the whole complex plane must be constant.

Miscellaneous Exercises

4𝑧 2 +𝑧+5 𝑥 𝑦
1. If 𝐹(𝑎) = ∮𝐶 𝑑𝑧, 𝑤ℎ𝑒𝑟𝑒 𝐶: (2) + ( 3) = 1,
𝑧−𝑎

Taken in counter clockwise sense, then fid 𝐹(3, 5), 𝐹(𝑖), 𝐹 ′ (−1), 𝐹 ′′ (−𝑖)

sin 𝑧
2. ∫𝐶 𝑑𝑧 integrate clockwise around the unit circle
𝑧4
𝑒𝑧
3. ∫𝐶 𝑧 𝑛 𝑑𝑧 integrate clockwise around the unit circle
𝑧6
4. ∫𝐶 (2𝑧−1)6 𝑑𝑧 integrate clockwise around the unit circle
𝑑𝑧
5. ∫𝐶 (𝑧−2𝑖)2 (𝑧−𝑖/2)2 𝑑𝑧 integrate clockwise around the unit circle
(1+𝑧) sin 𝑧
6. ∮𝐶 𝑑𝑧 , 𝐶: |𝑧 − 𝑖| = 2 Counterclockwise.
(2𝑧−1)2

exp(𝑧 2 )
7. ∮𝐶 𝑧(𝑧−2𝑖)2 𝑑𝑧 , 𝐶: |𝑧 − 3𝑖| = 2 Clockwise.
ln(z+3)
8. ∮𝐶 (𝑧−2)(𝑧+1)2 𝑑𝑧 , C the boundary of the square with vertices±1.5, ±1.5𝑖

counterclockwise.

Badri A, Moges B. and Teklebrhan B. 255 AKU

 APPLIED MATHEMATICS III

References:

 Ahlfors, L. V.: "Complex Analysis," 3d ed., McGraw-Hill Higher Education, Burr

Ridge, IL, 1979.
 Bak, J., and D. J. Newman: "Complex Analysis," 2d ed., Springer-Verlag, New York,
1997.
 Conway, J. B.: "Functions of One Complex Variable," 2d ed., 6th Printing, Springer-
Verlag, New York, 1997.
 Fisher, S.D.: "Complex Variables:' 2d ed., Dover Publications, Inc., Mineola, NY, 1990.
 Flanigan, F. J.: "Complex Variables: Harmonic and Analytic Functions,” Dover
Publications, Inc., and Mineola, NY. 1983.
 Hille, E.: "Analytic Function Theory," Vols. 1 and 2, 2d ed., Chelsea Publishing Co., New
York, 1973.
 Kaplan, W.: "Advanced Calculus," 5th ed., Addison-Wesley Higher Mathematics,
Boston, MA, 2003.
 Krantz, S. G.: "Complex Analysis: The Geometric Viewpoint," Carns Mathematical
Monograph Series, the Mathematical Association of America, Washington, DC, 1990.

Badri A, Moges B. and Teklebrhan B. 256 AKU

 APPLIED MATHEMATICS III

 Krzyz, J. G.: "Problems in Complex Variable Theory," Elsevier Science, New York, 1972.
 Lang, S.: "Complex Analysis," 3d ed., Springer-Verlag, New York, 1993.
 Markushevich, A. 1.: "Theory of Functions of a Complex Variable," 3 vols. in one, 2d
ed., American Mathematical Society, Providence, RI, 1977.
 Mathews, J. H., and R. W. Howell: "Complex Analysis for Mathematics and
Engineering," 4th ed., Jones and Bartlett Publishers, Sudbury, MA, 2001.
 Rubenfeld, L.A.: "A First Course in Applied Complex Variables," John Wiley & Sons, Inc.,
New York, 1985.
 Silverman, R. A.: "Complex Analysis with Applications," Dover Publications, Inc., ineola,
NY, 1984.
 Whittaker, E. T., and G. N. Watson: "A Course of Modern Analysis," 4th ed., Cambridge
University Press, New York, 1996.

Badri A, Moges B. and Teklebrhan B. 257 AKU