FACULTY OF SCIENCE

DEPARTMENT OF MATHEMATICS AND COMPUTATIONAL SCIENCES

Module title : ENGINEERING MATHEMATICS 2

Module code : MTE201

Lecturer(s) : T Mazikana, M. Zhangazha, S Murambiwa

October 18, 2021

 MODULE OUTLINE

Module description

The module consists of five sections which are

1. Complex variables,

2. Ordinary di↵erential equations,

3. Fourier series and Laplace transforms,

4. Vector calculus,

5. Probability and Statistics.

The module aims to provide students with the mathematical knowledge and skills that are needed
to support subsequent engineering studies.

Aims

The module is intended to enable students to:

1. Demonstrate an understanding of the basic concepts and ideas of engineering mathematics

and how its ideas may be used in, or are connected to various fields of engineering,

2. Analyse and solve problems of both theoretical and practical importance.

Module Content

0.1 Complex Variables

1. Complex Numbers

1
 (a) Conjugate Complex Numbers,
(b) Modulus of Complex numbers,
(c) Polar Form of Complex Numbers,
(d) Roots of Complex Numbers

2. Continuity,Complex Di↵erentiation and Cauchy-Riemann Equations

(a) Limits and Continuity of complex functions,

(b) Complex di↵erentiation,
(c) Cauchy-Riemann Equations,
(d) Harmonic Functions

3. Complex Integration and Cauchy’s Theorem

(a) Contours,
(b) Contour Integrals,
(c) Simply and Multiply connected regions,

4. Cauchy’s Integral Formulae and Related Theorems

(a) Cauchy’s Integral Formulae,

(b) Maximum and Minimum Modulus Principle

5. Infinite Series: Taylor’s and Laurent Series

(a) Taylor and Maclaurin Series,

(b) Laurent Series

6. Residues of a Function

(a) Methods for Computing a residue,

(b) Applying Residues to evaluate definite integrals,
(c) Evaluation of improper real integrals

0.2 Ordinary Di↵erential Equations

1. Basic techniques for the solution of first order ODE’s.

(a) Separation of variables technique.

(b) Integrating factor method.
(c) Exact first order di↵erential equations.
(d) Bernoulli equation.

2
 2. Basic techniques for the solution of second order ODE’s.

(a) Methods of undetermined coefficients.

(b) Reduction of order.
(c) Method of variation of parameters.

3. Series solutions of linear ODE’s.

(a) Solutions near ordinary points.

(b) Singular points.
(c) Euler equations.

0.3 Fourier Series and Laplace Transforms

1. Introduction to Fourier Series

(a) Orthogonality.
(b) Euler Formula for the Coefficients.
(c) Bessels Inequality and Parsevals Identity.
(d) Fourier Series of Even and Odd Functions.

2. The Fourier Transform

(a) Fourier transforms as integrals.

(b) Properties of the Fourier transform.
(c) Fourier Transform of a Convolution.
(d) Fourier Sine and Cosine Transforms.

3. Application to Partial Di↵erential Equations

(a) Heat Equation for an Infinite Rod.

(b) Half Range Problems.

4. Laplace transform

(a) Laplace transform of the elementary functions.

(b) Properties of Laplace transform.
(c) Inverse Laplace transform.
(d) Laplace transform methods for solving initial value problem (IVP).

3
0.4 Vector Calculus

Vector calculus has been described as the language of mechanics and electromagnetism because of
its rigorous applications to problems in engineering and physics.

1. Vector Algebra

(a) Introduction of vectors: Magnitude, components, Direction cosines.

(b) Scalar, vector and triple product.
(c) Application to work.
(d) Vector identities

2. Di↵erentiation

(a) The vector valued function of a single variable.

(b) Space curves.
(c) Arc length.

3. Scalar and vector fields.

(a) Gradient, divergence and curl.

(b) The del notation.
(c) Conservative vector fields and the existence of scalar potential.
(d) Vector identities.

4. Orthogonal coordinates.

(a) General and orthogonal curvilinear coordinate system.

(b) cylindrical, spherical and paraboloidal coordinates.

5. Integration

(a) Line, surface and volume integrals.

(b) Conservative and solenoidal vector and existence of potential function in simply con-
nected domains.
(c) Oriented surfaces.

6. Integral theorems

(a) The divergence and Stoke’s theorems.

(b) Green’s formulas.

4
0.5 Probability and Statistics

1. Introduction

(a) What is statistics?

(b) Important definitions
(c) Applications of statistics
(d) Explanatory statistics

2. Continuous random variables

2.1. Continuous random variables in one-dimensional

2.2. Mean and variance for continuous random variables
2.3. Median and Mode for continuous random variables.
2.4. Continuous random variables in 2D.
2.5. Marginal probability distribution
2.6. Conditional probability for continuous random variables.
2.7. Independence in continuous random variables.
2.8. Properties of variance for continuous random variables.
2.9. The Normal distribution

3. Hypothesis testing

3.1 Introduction
3.2. Comparing a single mean to a specified value, when the population variance is known
and the dataset is not given.
2 2
3.3. The di↵erence between two means when population variances 1 and 2 are known
3.4. Hypothesis testing when the population variance is unknown, the dataset is given and
n < 30.
2 2
3.5. Di↵erence between two population mean when the population variances 1 and 2 are
not known and (n1 1) + (n2 1) = n1 + n2 2 < 30.
3.6 Paired Comparison

4. Confidence intervals

4.1 Introduction
4.2. Comparing a single mean to a specified value, when the population variance is known
and the dataset is not given.
2 2
4.3. The di↵erence between two means when population variances 1 and 2 are known
4.4. Hypothesis testing when the population variance is unknown, the dataset is given and
n < 30.

5
 2 2
4.5. Di↵erence between two population mean when the population variances 1 and 2 are
not known and (n1 1) + (n2 1) = n1 + n2 2 < 30.
4.6 Paired Comparison

5. Regression analysis

5.1 Introduction
5.2. Least squares estimation of 0 and 1

5.3. Analysis of variance approach to regression analysis

5.4. The coefficient of determination
5.5 Inferences concerning the regression coefficients

Selected Resources(references)

1. E Kreyzig, ADVANCED ENGINEERING MATHEMATICS.

2. Murray R. Spiegel, SCHAUM’S OUTLINE OF COMPLEX VARIABLE.

3. SCHAUM’S OUTLINE DIFFERENTIAL EQUATIONS.

6
Chapter 1

Complex Numbers

We know that God exists because Mathematics is consistent and we know that the devil exists because we
cannot prove the consistency.

—Andre Weil

Introduction

Complex are numbers of the form z = x + iy, where x, y 2 R and i2 = 1. It is denoted z 2 C.

We call x the real part of z denoted Re(z) and y the imaginary part of z denoted Im(z), that is,Re
(z) = x and Im (z) = y.
Let z1 = x1 + iy1 and z2 = x2 + iy2 where xi , yi 2 R, i = 1, 2. Then z1 = z2 if and only if x1 = x2
and y1 = y2 . p
Given z 2 C, z = x + iy we define the modulus of z, denoted by |z| = x2 + y 2 .

1.1 Conjugate Complex Numbers

Let z 2 C, where z = x + iy. Then the conjugate of z denoted z̄ is given by z̄ = x iy.

1.1.1 Properties

1. z + z̄ = 2Re(z)

2. z z̄ = 2Im(z)

7
 3. z̄¯ = z

4. If z1 , z2 , ..., zn 2 C, then z1 + z2 + ... + zn = z̄1 + z̄2 + z̄3 + ... + z̄n and z1 .z2 ...zn = z̄1 z̄2 ...z̄n

5. z is real if and only if z = z̄

6. z is purely imaginary if and only if z = z̄

7. |z|2 = z.z̄

1.2 Modulus of Complex numbers

1. |z̄| = |z|

2. If z1 , z2 , ..., zn 2 C, then |z1 z2 ...zn | = |z1 ||z2 |...|zn |

3. Re(z)  |z|, Im(z)  |z|

4. |z1 + z2 |  |z1 | + |z2 | or |z1 + z2 + ... + zn |  |z1 | + |z2 | + ... + |zn | (Triangle Inequality)

5. |z1 z2 | ||z1 | |z2 ||

Example 1.2.1. Show that if z1 + z2 and z1 z2 are both real then either z1 and z2 are both real or
z1 = z̄2 .

Solution
Let z1 + z2 = r1 and z1 z2 = r2 where r1 , r2 2 R. It is sufficient to prove the results when z1 and z2
are not both zero e.g z2 6= 0. Then

r2 r2 z̄2 r2
z1 = = = r3 z̄2 where r3 2 R.
z2 z2 z̄2 |z2 |

Thus r3 z̄2 + z2 = r1 .
Since Im(z̄2 ) = Im(z2 ) and Im(r1 ) = 0, we have r3 Im(z2 ) + Im(z2 ) = 0
(1 r3 )Im(z2 ) = 0. Therefore either r3 = 1 and hence z1 = z̄2 or Im(z) = 0 from which it follows
that z2 is real and consequently, z1 is also real.

1.3 Graphical Representation of Complex Numbers

Since a complex number x+iy can be considered as an ordered pair of real numbers, we can represent
such numbers by points in an xy plane called the complex plane or Argand diagram. The complex

8
number 3+4i can be considered as an ordered pair (3, 4). To each complex number there corresponds
one and only one point in the plane, and conversely to each point in the plane there corresponds
one and only one complex number. Because of this we often refer to the complex number z = x + iy
as the point z = (x, y). Sometimes, we refer to the x and y axes as the real and imaginary axes,
respectively, and to the complex plane as the z plane. The distance
p between two points, z1 = x1 +iy1
and z2 = x2 + iy2 , in the complex plane is given by |z1 z2 | = (x1 x2 )2 ) + (y1 y2 )2 ).

1.4 Polar Form of Complex Numbers

Let P be a point in the complex plane corresponding to the complex number x + iy or (x, y).

Then we see that

x = r cos ✓, y = r sin ✓

p
where r = x2 + y 2 = |x + iy| is called the modulus of z = x + iy (denoted by mod z or |z|), and ✓
is the angle that line OP makes with the positive x axis and is called the argument of z = x + iy
(denoted by arg z).
It follows that

z = x + iy = r(cos ✓ + i sin ✓)

9
which is called the polar form of the complex number, and r and ✓ are called polar coordinates. It
is sometimes convenient to write the abbreviation cis✓ for cos ✓ + i sin ✓.
For any complex number z 6= 0 there corresponds only one value of ✓ in 0  ✓  2⇡. However, any
other interval of length 2⇡, for example ⇡  ✓  ⇡, can be used. Any particular choice, decided
upon in advance, is called the principal range, and the value of ✓ is called its principal value.

1.4.1 Multiplication and Division in Polar Form

Let

z1 = r1 (cos ✓1 + i sin ✓1 ) and z2 = r2 (cos ✓2 + i sin ✓2 )

Then

z1 z2 = r1 r2 [(cos ✓1 cos ✓2 sin ✓1 sin ✓2 ) + i(cos ✓1 sin ✓2 + sin ✓1 cos ✓2 )]

Applying identities we get

z1 z2 = r1 r2 [cos(✓1 + ✓2 ) + i(sin(✓1 + ✓2 )] (1.1)

Taking arguments of the above shows that the argument of a product equals the sum of the
arguments of the factors

arg(z1 z2 ) = arg z1 + arg z2

Similarly we have

z1 r1
= [cos(✓1 ✓2 ) + i(sin(✓1 ✓2 )]
z2 r2

and similarly taking arguments of the above shows that the argument of a quotient equals the
di↵erence of the arguments of the factors

z1
arg = arg z1 arg z2
z2

10
1.4.2 De Moivre’s and Euler formulae

De Moivre’s theorem

A generalization of (1.1) leads to

z n = [r(cos ✓ + i sin ✓)]n = rn (cos n✓ + i sin n✓)

which is often called De Moivres theorem.

For |z| = r = 1, the formula becomes

(cos ✓ + i sin ✓)n = cos n✓ + i sin n✓

We can use this to express cos n✓ and sin n✓ in terms of powers of cos ✓ and sin ✓. For instance, for
n = 2 we have on the left cos2 ✓ + 2i sin ✓ cos ✓ sin2 ✓. Taking the real and imaginary parts on both
sides of (1.1) with n = 2 gives the familiar formulas

cos 2✓ = cos2 sin2 ✓ and sin 2✓ = 2 sin ✓ cos ✓

Euler’s Formula

The formula
ei✓ = cos ✓ + i sin ✓ (1.2)
is called Euler’s formula. It is more convenient, however, simply to take (1.2) as a definition of ei✓ .
In general, we define

ez = ex+iy = ex eiy = ex (cos y + i sin y)

In the special case where y = 0 this reduces to ex .

Note that the De Moivre’s theorem reduces to (ei✓ )n = ei✓n .

11
1.4.3 Roots of Complex Numbers

1
A number ! is called an nth root of a complex number z if ! n = z, and we write ! = z n . From De
Moivres theorem we can show that if n is a positive integer,
1 1
z n =[r(cos ✓ + i sin ✓)] n
 ✓ ◆ ✓ ◆
1 ✓ + 2k⇡ ✓ + 2k⇡
=r n cos + i sin k = 0, 1, 2, ..., n 1.
n n
1
from which it follows that there are n di↵erent values for z n , i.e., n di↵erent nth roots of z, provided
z 6= 0

1.4.4 The nth Roots of Unity

The solutions of the equation z n = 1 where n is a positive integer are called the nth roots of unity
and are given by

2k⇡ 2k⇡ 2k⇡i

z = cos + i sin = e n for k = 0, 1, 2..., n 1.
n n

2⇡i 2⇡i
If we let ! = cos + sin , the n roots are 1, !, ! 2 , ..., ! n 1 . Geometrically, they represent the
n n
n vertices of a regular polygon of n sides inscribed in a circle of radius one with center at the origin.
This circle has the equation |z| = 1 and is often called the unit circle.

12
Chapter 2

Continuity,Complex Di↵erentiation and Cauchy-

Riemann Equations

Definition 2.0.1. Suppose, to each value that a complex variable z can assume, there corresponds
one or more values of a complex variable w. We then say that w is a function of z and write
w = f (z) or w = G(z), etc. The variable z is sometimes called an independent variable, while w
is called a dependent variable. The value of a function at z = a a is often written f (a). Thus, if
f (z) = z 2 , then f (2i) = (2i)2 = 4.
Definition 2.0.2. If only one value of w corresponds to each value of z, we say that w is a single-
valued function of z or that f (z) is single-valued. If more than one value of w corresponds to each
value of z, we say that w is a multiple valued or many-valued function of z.

2.1 Limits and Continuity of complex functions

Definition 2.1.1. A function f (z) is said to have the limit l as z approaches a point z0 , written

lim f (z) = l
z!z0

if for any positive number ✏ (however small), we can find some positive number (usually depending
on ✏) such that |f (z) l| < ✏ whenever 0 < |z z0 | < .

In other terms we can define continuity as: If f is defined in a neighbourhood of z0 (except perhaps
at z0 itself) and if the values of f are ”close” to l for all z ”close” to z0 .
Definition 2.1.2. Let f (z) be a defined and single-valued function. The function f (z) is said to
be continuous at z = z0 if lim f (z) = f (z0 ). Note that this implies three conditions that must be
z!z0
met in order that f (z) be continuous at z = z0 .

13
 1. lim f (z) = l must exist.
z!z0

2. f (z0 ) must exist, i.e., f (z) is defined at z0 .

3. l = f (z0 ).

2.2 Di↵erentiation

Definition 2.2.1. If f (z) is a single-valued function defined on a domain D, then f (z) is di↵eren-
tiable at a point z 2 D if

0 f (z + z) f (z)
f (z) = lim
z!0 z

0
exists independent of the manner in which z ! 0. In this case f (z) is called the derivative of
f (z) at z.

0
Note: The limit f (z) exists provided the limit in independent of the manner in which z ! 0.

Class work:

1. Using the definition, find the derivative of f (z) = z 3 2z.

2. Show that f (z) = |z|2 is di↵erentiable at the origin and nowhere else.

Solution:

1. By definition, the derivative f (z) is

0 f (z + z) f (z)
f (z) = lim
z!0 z
3
(z + 2(z + z) (z 3 2z)
z)
= lim
z!0 z
z + 3z z + 3z( z)2 + ( z)3 2z
3 2
2 z z 3 + 2z
= lim
z!0 z
2 2
= lim 3z + 3z z + ( z) 2
z!0
2
= 3z 2

14
 2. If z 6= 0, then
f (z + z) f (z) |z + z|2 |z|2
lim = lim
z!0 z z!0 z
(z + z)(z + z) zz
= lim
z!0 z
zz + z z + z z + z z zz
= lim
z!0 z
z
= lim [z + z + z]
z!0 z
z
= z + z lim + lim z
z!0 z z!0
z
= z + z lim .
z!0 z
z x i y
Now consider lim = lim
z!0 z x!0 y!0 x+i y
z x
(i) Along the x-axis y = 0, lim = lim = 1,
z!0 z x!0 x
z i y
(ii) Along the y-axis x = 0, lim = lim = 1.
z!0 z y!0 y
z
Thus lim does not exist as limit depends upon the manner in which z ! 0.
z!0 z
f (0 +
z) f (0) | z|2 z. z
Now if z = 0, lim = lim = lim = 0.
z!0 z z!0 z z!0 z
0
Thus f (0) = 0 and so f (z) is di↵erentiable at the origin.

Exercise

1. Let f (z) = u(x, y) + iv(x, y), determine u(x, y) and v(x, y) when f (z) = z 2 + 3z + 4.
z2 + 4
2. Show that lim = 4i.
z!2i z 2i

2.2.1 Analytic Functions

Complex analysis is concerned with the theory and application of ”analytic functions”, that is,
functions that are di↵erentiable in some domain, so that we can do ”calculus in complex”.

Definition 2.2.2. A function is said to be analytic in a domain D if f (z) is defined and di↵erentiable
at all points of D. The function f (z) is said to be analytic at a point z = z0 in D if f (z) is analytic
in a neighbourhood of z0 .

Example:f (z) = |z|2 is not analytic at the origin.

15
Definition 2.2.3. A function f (z) analytic at each point of the complex plane is called an entire
function.

Example: Polynomials are entire functions.

2.3 Cauchy-Riemann Equations

To do complex analysis on any complex function, we require that function to be analytic on some
domain that is di↵erentiable in that domain.
The Cauchy-Riemann Equations provide a criterion (a test) for the analyticity of a complex function

f (z) = u(x, y) + iv(x, y)

Theorem 2.3.1. For the function f (z) = u(x, y) + iv(x, y) to be analytic in a domain D, it
is necessary and sufficient that there exists in this domain continuous partial derivatives of the
functions u(x, y) and v(x, y) satisfying the Cauchy-Riemann conditions.

ux = v y , uy = vx .

0
If the Cauchy-Riemann conditions are fulfilled, then the derivative f (z) can be represented in one
0
of the following forms f (z) = ux + ivx = vy iuy = ux iuy = vy + ivx .

Example:

1. f (z) = z 2 is analytic for all z.

f (z) = z 2 = x2 y 2 + 2ixy, thus u = x2 y 2 and v = 2xy. Hence ux = 2x = vx and
uy = 2y = vx and thus the Cauchy-Riemann conditions are satisfied. Therefore f (z) = z 2
is analytic for all z.
2. f (z) = z̄ is not analytic.
For f (z) = z̄ = x iy we have u = x, v = y.Thus ux = 1 6= vy = 1 and see that the first
Cauchy Riemann equation is not satisfied. We conclude that is not analytic.

2.4 Harmonic Functions

The great importance of complex analysis in engineering mathematics results mainly from the fact
that both the real part and the imaginary part of an analytic function satisfy Laplace’s equation,

16
which is the most important P DE of physics. It occurs in gravitation, electrostatics, fluid flow,
heat conduction, and other applications

Definition 2.4.1. Let u(x, y) and v(x, y) be functions of complex variables in a region R. Then
u(x, y) and v(x, y) are harmonic functions in R if their second partial derivatives with respect to x
and y exist and are continuous in R and

uxx + uyy = 0 and vxx + vyy = 0.

Theorem 2.4.1. If f (z) = u(x, y) + iv(x, y) is analytic in the domain D, then u(x, y) and v(x, y)
are harmonic functions in this domain D.

If two harmonic functions u and v satisfy the Cauchy Riemann equations in a domain D, they are the
real and imaginary parts of an analytic function f (z) in D. Then v is said to be a harmonic conjugate
function of u in D.

Classwork : Prove that the function u = 2x(1 y) is harmonic. Find the conjugate of u, that is,
find a function v such that f (z) = u + iv is analytic. Express f (z) in terms of z.

17
Tutorial 2.4.1. Attempt all questions

1. Show that for all z1 , z2 2 C, (i) |z1 + z2 |  |z1 | + |z2 | (ii) |z1 z2 | ||z1 | |z2 ||.

2. Prove that if z1 + z2 and z1 z̄2 are both real, then either z1 and z2 are both real or z1 = z2 .
z 5
3. If |z| = 2, show that | |  7.
z2 2z 3
4. Use De Moivre’s Theorem to find expressions for cos 5✓ and sin 5✓ as polynomials in cos ✓ and
sin ✓.

5. Find all the cube roots of i.

6. Sketch and describe the following sets

(i)|z 2 + i|  1, (ii) |2z + 3| > 4 (iii) Imz = 1, (iv) Imz 1, (v) |z 4| |z|.

7. Show that f (z) = z 2 is continuous everywhere. Show that f (z) = z̄ is continuous everywhere
but nowhere di↵erentiable.

8. Test for analyticity of the following functions

z 1 z2
(i) f (z) = (z̄ + 3i)2 + 6 (ii) f (z) = 2 (iii) f (z) = |z||z + 1| (iv) f (z) = .
z +4 z + z̄
9. Show that the function f (z) = z̄ is not analytic.

10. (a) Prove that u = e x (x sin y y cos y) is harmonic.

(b) Find v such that f (z) = u + iv is analytic, that is, the conjugate function of u.
(c) Find f (z) in terms of z.

11. Find the analytic function f (z) = w if its imaginary part is v(x, y) = 3x+2xy and if f (i) = 2.

12. Find the values of the real constants p and q for which f (z) = u + iv satisfies the Cauchy-
Riemann equations where u(x, y) = px4 6x2 y 2 + y 4 5x, v(x, y) = 4x3 y 4xy 3 + qy.
Given that p and q have these values, obtain an explicit formula for the di↵erentiable function
f (z) in terms of z and find its derivative.

13. Show that u = 3xy 2 x3 is harmonic in the entire complex plane. Find the conjugate harmonic
function v. Show that the analytic function is f (z) = z 3 + ic.

18
Chapter 3

Complex Integration and Cauchy’s Theo-

rem

3.1 Contours

Integrals of complex-valued functions of complex numbers are defined on curves in the complex
plane.
Definition 3.1.1. Arc
A set of points z = (x, y) in the complex plane is called an arc if x = x(t), y = y(t), a  t  b
where x(t) and y(t) are continuous real-valued function of a real parameter t.

Thus an arc z = x(t) + iy(t) and z = z(t) is a continuous mapping of parameter t.

Definition 3.1.2. Simple arc or Jordan arc
An arc is called a simple arc or Jordan arc if it does not cross itself.
Definition 3.1.3. Jordan curve
If an arc is simple except for the end of the parameter on [a, b], that is, z(a) = z(b), then the arc is
called simple closed curve or Jordan curve.
Example 3.1.1. .
A circle centred at z0 with radius r: |z z0 | = r is a simple closed curve. Written z z0 = rei✓
which implies z = z0 + rei✓ , 0  ✓  2⇡.
Unit circle in negative direction z = ei✓ , 0  ✓  2⇡

The arc z = ei2✓ , 0  ✓  2⇡ is a circle traversed twice in the positive direction. This not a simple
closed curve.
The arc z = ei2✓ , 0  ✓  ⇡ is a unit circle and is a simple closed curve.
Definition 3.1.4. Di↵erentiable arc
An arc z = x(t) + iy(t) a  t  b is called a di↵erentiable arc if x0 (t) and y 0 (t) exist and are
continuous on the parameter interval [a, b].

19
 p
Now z 0 (t) = x0 (t) + iy 0 (t) and |z 0 (t)| = (x0 (t))2 + (y 0 (t))2 is integrable over [a, b] and length of arc
Z b
is given by L = |z 0 (t)|d(t).
a

Definition 3.1.5. Smooth Arc

An arc is said to be smooth if z 0 (t) = x0 (t) + iy 0 (t) is continuous on [a, b] and z 0 (t) is non-zero on
(a, b)

Definition 3.1.6. Contour

A contour is an arc consisting of a finite number of smooth arcs joined end to end.

A simple closed contour is such that the initial and final value of z(t) are equal.

3.2 Contour Integrals

When integrating a complex function f (z) along a contour C between two points z1 and z2 in the
complex plane, we can’t always use simple intervals as we do in single variable calculus . To perform
the integration we describe the contour as some function z = z(t) on an interval a  t  b, where
z(a) = z1 and z(b) = z2 . Substituting this into the integral we get the complex line integral.

Definition 3.2.1. The contour integral of f (z) along C is defined as

Z Z b
f (z)dz = f (z(t)).z 0 (t)dt.
C a

Since C is a contour and z 0 (t) is piece-wise continuous and existence of integral is assured.

3.2.1 Integration along a straight line

When integrating along a straight line it is essential to first parametrize the line. When parametris-
ing a straight line it is useful to draw the line on an Argand diagram to help. We then find the
parametric function z(t). A general form of parametrisation we can use for straight lines is

z(t) = (1 t)z1 + tz2 with 0  t  1. (3.1)

This works for any straight line because when t = 0, z(t) = z1 and, as t increases from 0 to 1, z(t)
follows a linear path until it reaches z2 .
Z
Example 3.2.1. Let C be the line from z1 = 1 to z2 = 1 i. Calculate I = |z|2 dz.
C

20
Solution
With any problem of this kind it is always useful to take note of the start and end points of the con-
tour C. Firstly substitute the the endpoints z1 and z2 into equation (3.1) to get the parametrisation.
This gives

z(t) = (1 t)1 + t( 1 i) = 1 2t it with 0  t  1.

dz
Note: =( 2 i), and thus dz = ( 2 i)dt.
dt

Now substitute z(t) and dz = ( 2 i)dt into the integrand to get

Z 1
I= |1 2t it|2 ( 2 i)dt.
0

( 2 i) is a constant and can be taking out of the integral, then further simplification yields

Z 1 Z 1
2 2
I=( 2 i) ((1 2t) + t )dt = (2 + i) (1 4t + 5t2 )dt.
0 0

Now you can integrate the function to obtain

 1
5 4 2i
I =( 2 i) t 2t2 + t3 =
3 0 3 3
Z
4 2i
I = |z|2 dt = .
C 3 3

3.2.2 Integration along arcs

When integrating along a full circle, or an arc of a circle, in the anti-clockwise direction, we use the
following parametrisation
z(✓) = z0 + Rei✓ with ↵  ✓  (3.2)
Here z0 is the centre of the circle, R is the radius of the circle, and ↵ and are the angles from z0
at z1 and z2 . So for a full circle you would use ✓ 2 (0, 2⇡). Another important thing to consider
is the direction of the contour, as ✓ increases in the above equation the contour moves round
anti-clockwise, to make it move in the clockwise direction use

21
 i✓
z(✓) = z0 + Re with ↵  ✓ 

Example 3.2.2. Let C be the top half of a circle

Z of radius 2 centred on point z0 = (1 + i) oriented
in the anti-clockwise direction, evaluate I = Re(z)dz
C

Solution
Since we are going in the anti-clockwise direction we use parameterisation in equation (3.2) to get

z(✓) = 1 + i + 2ei✓ with 0  ✓  ⇡

dz
Note: = 2iei✓ , and thus dz = 2iei✓ d✓
d✓

Now substitute z(✓) and dz = 2iei✓ d✓ into the integrand to get

Z ⇡
I= Re(2ei✓ + 1 + i)2iei✓ d✓.
0

Simplify the integrand to get

Z ⇡ Z ⇡
i✓
I= (2cos✓ + 1)2ie d✓ = 4iei✓ cos✓ + 2iei✓ d✓
0 0

The trick here is to split the 4iei✓ cos✓ term into its complex parts, so that

i✓
e + ei✓
4iei✓ cos✓ = 4iei✓ ( ) = 2iei✓ (e i✓
+ ei✓ ) = 2i + 2ie2i✓ .
2

Putting this back into the integral we get

Z ⇡
I= 2i + 2ie2i✓ + 2iei✓ d
⇥0 ⇤⇡
I = 2i✓ + e2i✓ + 2ei✓ 0 = 4 + 2⇡i

22
Key points to follow when integrating

1. Sketch out the Complex plane, label endpoints and draw the contour, include a directional
arrow.

2. Find a parametrisation for the contour using one of the above methods or your own intuition.
Make sure the endpoints satisfy z(a) = z1 and z(b) = z2 .

3. Substitute the parametrisation into the integral, remember now you are integrating with
respect to t (or ✓).

4. Once you have done this you can simplify and integrate as normal.

Exercise.
Z
⇡ ⇡
1. Find zdz where C is given by z = 2ei✓ , ✓
C 2 2
2. Let f (z) = (z z0 )m where m is the integer and z0 a constant. Integrate f (z) counter-clockwise
around the circle C of radius ⇢ with center at z0 .

3.2.3 Properties of Contour Integrals

Z Z
1. f (z)dz = f (z)dz where C is C traversed in opposite direction.
C C
Z Z Z
2. f (z)dz = f (z)dz + f (z)dz where C = C1 + C2
C C1 C2
Z Z Z
3. [f (z) + g(z)]dz = f (z)dz + g(z)dz
C C C
Z Z
4. f (z)dz = h f (z)dz where h is a constant
h C
Z
5. | f (z)dz|  M L where L is the length of the contour, and |f (z)|  M .
C

Proof.
We prove (5) below:

23
 Z Z b
| f (z)dz| =| f (z(t))z 0 (t)dt|
C Z ab
 |f (z(t))z 0 (t)|dt
Za b
= |f (z(t))||z 0 (t)|dt
Za
 M |z 0 (t)|dt where |f (z)|  M on C
Z b
=M |z 0 (t)|dt
a
= ML

Z 2+i
dz
Exercise: Show that | |  2, where the path is the straight line from i to 2 + i.
i z2

3.3 Simply and Multiply connected regions

Definition 3.3.1. A region R is simply connected if any simple closed curve in R can be shrunk
to a point without leaving R.
A region R is multiply connected if R is not simply connected.

In other words a domain R is called simply connected if every simple closed curve (closed curve
without self-intersections) encloses only points of R

Example 3.3.1. .
1) |z| < 2 is simply connected.
2) 1 < |z| < 2 is multiply connected.

Theorem 3.3.1. Cauchy’s Theorem

If f (z) is analytic in a simply connected
Z domain D and if f 0 (z) is continuous at each point within
and on a closed contour C, then f (z)dz = 0.
C

Example
Z 3.3.2. . Z
1) ez dz = 0, and z n dz = 0 for any closed path because these functions are analytic for
all Zz.
dz
2) = 0 where C is the unit circle. This is because f (z) is analytic inside C and the only
z2 + 4
points where f (z) is not analytic (i.e at z = ±2) lies outside C.

Take note of the following important points.

24
 Z Z 2⇡
it
1. z̄dz = e ieit dt = 2⇡ where C is a unit circle. This does not contradict Cauchy’s
C 0
theorem because f (z) = z̄ is not analytic.
Z
1 1
2. 2
dz = 0 where C is a unit circle. We can see that f (z) = 2 is not analytic at z = 0, so
C z z
Cauchy’s theorem fails. Thus this result does not follow from Cauchy’s theorem, but follows
procedures of section 3.2 above.

It was first shown by Goursat that it is unnecessary to assume the continuity of f 0 (z) and that
Cauchy’s Theorem holds if we only assume that f 0 (z) exists at all points within and on C.

Theorem 3.3.2.ZCauchy-Goursat Theorem If f (z) analytic at all points within and on a closed
contour C, then f (z)dz = 0.
C

Exercise. Z
1
1) Evaluate I = 2
, where C is a unit circle traversed counter-clockwise. 2) Evaluate
Z C z 5z + 6
1
I= 2 (z
, where C : |z| = 4
C z 2)(z 4)

3.3.1 Consequences of Cauchy’s Theorem

Z b
1. If f (z) is analytic in a simply connected region R, then f (z)dz is independent of the path
a
in R joining two points a and b in R.

2. Let f (z) be analytic in a simply connected region R, a be a fixed point in R and z be any
Z b
point in R. Then F (z) = f (u)du is analytic in R and F 0 (z) = f (z).
a
Z
3. If f (z) is analytic in a region bounded by two simple closed curves C1 and C2 , then f (z)dz =
Z C1

f (z)dz where C1 and C2 are both traversed in the positive direction.

C2 Z
Thus for simply connected R with boundary C, f (z)dz can be expressed in terms of con-
C Z
venient contours e.g circles on which the integral is easy to evaluate. Hence f (z)dz =
Z C

f (z)dz where is a circle wholly contained in C.

4. Further, if f (z) is analytic in a region bounded by non-overlapping simple closed curves C,

C1 , C2 , ..., Cn where C1 , C2 , ..., Cn are inside C, then

25
Z Z Z Z
f (z)dz = f (z)dz + f (z)dz + ... + f (z)dz.
C C1 C2 Cn

26
Chapter 4

Cauchy’s Integral Formulae and Related

Theorems

4.1 Cauchy’s Integral Formulae

Definition 4.1.1. Cauchy’s Integral Formulae

If f (z) is analytic inside and on the boundary C of a simply-connected region D and a is an interior
point of C, then

Z
1 f (z)
f (a) = dz.
2⇡i C z a
2
z +1
Example 4.1.1. Integrate g(z) = counter-clockwise around
z2 1
i) |z 1| = 1
ii) |z + 1| = 1
iii) |z i| = 1

Solution
f (z) is not analytic at 1 and 1. These are the points we have to watch for.

i) The circle |z 1| = 1 encloses the point z0 = 1 where g(z) is not analytic. Hence from
Cauchy’s Integral Formulae we have to write

z2 + 1 z2 + 1 1
g(z) = = , thus
z2 1 z+1 z 1
z2 + 1
f (z) =
z+1

27
 Z  2
z2 + 1 z +1
From the Cauchy’s Integral Formulae 2
dz = 2⇡if (1) = = 2⇡i
C z 1 z+1 z=1

ii) The function g(z) is as before, but f (z) changes because we must take z0 = 1 (instead of
1). Hence we must write

z2 + 1 z2 + 1 1
g(z) = = , thus
z2 1 z 1 z+1
z2 + 1
f (z) =
z 1
Z  2
z2 + 1 z +1
From the Cauchy’s Integral Formulae 2
dz = 2⇡if ( 1) = = 2⇡i
C z 1 z 1 z= 1
Z
z2 + 1
iii) 2
dz = 0. Why?
C z 1
I 2
ez
Example 4.1.2. Evaluate dz, i = 1, 2, 3
Ci z2 6z
where

1. C1 : |z 2| = 1;

2. C2 : |z 2| = 2

3. C3 : |z 2| = 5

Solution

1. In the domain bounded by the circle C1 , the integrand is an analytic function, hence by the
Cauchy-Goursat Theorem
I 2
ez
dz = 0.
C1 z2 6z

2. In the domain bounded by the circle C3 , the function is not analytic on z0 = 0, thus we can
write

I 2 I 2
ez ez 1
dz = dz, thus
C2 z2 6z C2 (z 6) z
z2
e
f (z) =
(z 6)

28
 Applying Cauchy’s Integral formula we have

I I ez
z2
e z 6 dz
dz =
C2 z2 6z C2 z
" 2
#
ez
=2⇡i
z 6
z=0
⇡i
=
3
3. In the domain bounded by the circle C3 , the function is not analytic on z0 = 0 and at z0 = 6.
Applying Cauchy’s Integral formula we have

I 2 I ez
2
I 2
ez
ez z 6 z
dz = dz + dz
C3 z2 6z 1
z 2
z 6
2
# " " 2#
ez ez
= 2⇡i + 2⇡i
z 6 z
z=0 z=6
e36 1
=( )⇡i.
3

4.1.1 Derivatives of Analytic Functions

If f (z) is analytic inside and on the boundary C of a simply-connected region D, then the value of
the derivative of f (z) at a point in C is given by

Z
01 f (z)
f (a) = dz
2⇡i C (z a)2

and in general

Z
(n) n! f (z)
f = .
2⇡i C (z a)n+1

If f (z) is analytic in a region D, then f 0 (z), f 00 (z), ..., f (n) (z) are analytic in D.
Example 4.1.3. .
i) for any contour enclosing the point ⇡i (counterclockwise)
I
cos(z)
= 2⇡i(cos z)0 |z=⇡i = 2⇡i sin 2⇡i.
(z ⇡i)2

29
ii) for any contour enclosing the point i we obtain by counterclockwise integration
I 4
z 3z 2 + 6 000

3
dz = ⇡i(z 4 3z 2 + 6) |z= i = 18⇡i
(z + i)

Theorem 4.1.1. Morera’s Theorem Z

If a function f (z) is continuous in a simply connected domain R and if f (z)dz = 0 around every
C
simple closed curve C in R, then f (z) is analytic in R.

Theorem 4.1.2. Cauchy’s Inequality

Let f (z) be analytic inside and on a circle CR of radius R centred at z0 . If |f (z)|  M for all
z 2 CR , then derivatives of f at z0 satisfy

n!M
|f (n) (z0 |  (n = 1, 2, 3, ..., ).
Rn

Proof
Let CR be positivelyI oriented. By the extended Cauchy’s Integral Theorem we have
n! f (⇠)
|f (n) (z0 )| = | d⇠| ⇠ 2 CR
2⇡i CZR (⇠ z0 )n + 1
n!
| || iRd✓|
2⇡i CR M
ZRn+1
n! M iR 2⇡
= d✓
2⇡i Rn+1 0
n! M i
= 2⇡
2⇡i Rn
n!
= n
R

Theorem 4.1.3. Liouville’s Theorem

Every bounded entire function is constant.

Example 4.1.4. .
Let f (z) be entire and let |f (z)| 1 on the whole complex plane. Prove that f is a constant.

Solution
1 1 1
Since |f (z)| 1 > 0 on the whole plane,is bounded and  1 for all z. Therefore is
f f (z) f
1 1
bounded on C by Liouville’s Theorem = c, a constant, that is, f (z) = = k, a constant.
f (z) c
Hence f is a constant.

30
4.2 Maximum and Minimum Modulus Principle

Theorem 4.2.1. Maximum Modulus Theorem

Suppose f (z) is analytic within and on a closed contour C and f (z) is not a constant. Then |f (z)|
attains its maximum value on the boundary of C and not at any interior point.

Theorem 4.2.2. Minimum Modulus Theorem

Suppose f (z) is analytic within and on a closed contour C and f (z) 6= 0 inside C. Then |f (z)|
attains its minimum value on the boundary of C and not at any interior point.

4.2.1 Singular Points

A point at which f (z) fails to be analytic is called a singular point or singularity of f (z). Various
types of singularities exist but we will only define zeros and poles.

Definition 4.2.1. Poles

g(z)
is a pole of order n of the function f (z) if f (z) = where g(z) is analytic and non-zero.
(z )n
If n = 1, is called a simple pole.

Example 4.2.1. .
1
a) f (z) has a pole of order 3 at z = 2.
(z 2)3
3z 2
b) f (z) = has a pole of order 2 at z = 1, and simple poles at z = 1 and
(z 12 )(z 1)()z 4
z = 4.

Definition 4.2.2. Zeros

↵ is a zero of order n of a function f (z) if f (z) = (z ↵)n g(z) where g(z) is analytic and non-zero
within the neighbourhood of ↵.
Thus if g(z) = (z ↵)n f (z), where f (z0 ) 6= 0 and n is a positive integer, then ↵ is called a zero of
order n of g(z). If n = 1, ↵ is called a simple zero.

Exercise
z8 + z4 + 2
1) Locate and name all the singularities of f (z) = .
(z 1)3 (3z + 2)2
2) Determine where f (z) is analytic.

Theorem 4.2.3. Rouche’s Theorem

If f (z) and g(z) are analytic functions within and on a closed contour C and |g(z)| < |f (z)| on C,
then f (z) and f (z) + g(z) have the same number of zeros inside C.

Example 4.2.2. Find the number of roots inside the following contours of the equation z 7 5z 3 +
12 = 0.

31
(a) C1 : |z| = 1
(b) C2 : |z| = 2
(c) C3 : 1  |z|  2. Solution

(a) Consider the circle C1 . Let f (z) = 12, g(z) = z 7 5z 3 . On C1 we have

|g(z)| = |z 7 5z 3 |  |z 7 | + |5z 3 |  6 < 12 = |f (z)|

Hence, by Rouche’s theorem, f (z) + g(z) = z 7 5z 3 + 12 has the same number of zeros inside
|z| = 1 as f (z) = 12, i.e., there are no zeros inside C1 .

(b) Consider the circle C2 : |z| = 2. Let f (z) = z 7 , g(z) = 12 5z 3 . On C2 we have

|g(z)| = |12 5z 3 |  |12| + |5z 3 |  60 < 27 = |f (z)|

Hence, by Rouche’s theorem, f (z) + g(z) = z 7 5z 3 + 12 has the same number of zeros inside
|z| = 2 as f (z) = z 7 , i.e., all the zeros are inside C2 .

(c) In (b) we have shown that there are 7 zeroes inside C2 and in (a) we have shown that there
are no zeros inside C1 , Hence all the roots lie inside |z| = 2 but outside |z| = 1. Therefore
there are 7 roots in C3

Exercise
Determine the number of zeroes of the polynomial 2z 5 6z 2 + z + 1 in 1  z  2.

32
Tutorial 4.2.1. Attempt all questions

1. Evaluate the following integrals:

Z i
(a) (z 1)dz on a straight line from 1 to i.
1
Z i+1
(b) (z 1)dz on the parabola y = x2 .
Z0
(c) xdz on the circle |z| = 1 described in the counter-clockwise direction.
C
Z i
(d) (x2 +iy 2 )dz on the right half of the unit circle |z| = 1 described in the counter-clockwise
i
direction.
Z
dz
(e) 2
where C is the circle |z| = 2.
C z 1
Z
(f ) ez dz where C is the perimeter of the square with vertices at points z = 0, z = 1,z =
C
1 + i, z = i traversed in that order.

2. Evaluate
Z the followingZintegrals by using the
Z Cauchy Integral Zformula.
cos z ez z2 1 sin( ⇡z
2
)
(a) dz (b) 2
dz (c) 2
dz (d) 2
dz
|z|=3 z |z|=1 z + 2z |z|=3 z + 1 |z 1|=2 z + 2z 3
Z
eiz
(e) 2
dz.
|z i|=1 z + 1

3. Find the number of roots of z 9 + 4z 2 = 0 in where

(a) : |z| = 1 (b) : |z| < 2 (c) : 1  |z|  2.

4. For the function f (z) = z 5 + 5z 3 + 2z find the number of zeros in the circle |z| < 1, in the
annulus 1  |z|  2 and in the annulus 2  |z| < 3.

33
Chapter 5

Infinite Series: Taylor’s and Laurent Se-

ries

Definition 5.0.3. An infinite sequence of complex numbers {zn } = {z1 , z2 , ..., zn , ...} converges to
z or has a limit z if for all ✏ > 0 there exists n0 2 N such that |zn z| < ✏ whenever n > n0 .

Definition 5.0.4. An infinite series ⌃zn = z1 + z2 + z3 + ... + zn + ... of complex numbers converges
to the sum S if the sequence of partial sum {Sn } where Sn = ⌃nk=1 zk converges to S. A sequence
converges absolutely if ⌃|zn | converges.

5.0.2 Power Series

Power series play an important role in complex analysis. They are the most important series in
complex analysis because their sums are analytic functions and every analytic function can be
represented by power series.

Definition 5.0.5. A power series in powers of is a series of the form

P1
n=0 an (z z0 )n = a0 + a1 (z z0 ) + a2 (z z0 )2 + ...

where z is a complex variable,a1 , a2 , ... are complex (or real) constants, called the coefficients of the
series, and z0 is a complex (or real) constant, called the center of the series.
If z0 = 0 we obtain as a particular case a power series in powers of z:

P1
n=0 an z n = a0 + a1 z + a2 z 2 + ...

34
5.1 Taylor and Maclaurin Series

Theorem 5.1.1. Taylor’s Theorem If f (z) is analytic inside a circle centre a with radius R,
that is, |z a| < R for all z 2 |z a| < R, then

P1 f (n) (a)
f (z) = n=0 an (z a)n where an =
n!

A Maclaurin series is a Taylor series with center z0 = 0.

1
Note that the Maclaurin expansion of is the geometric series
1 z

1 P1 n 2
= n=0 z = 1 + z + z + ... and converges for |z| < 1
1 z
Example 5.1.1. .
a) Let f (z) = ln(1 + z). Expand f (z) in Taylor series about z = 0 and determine the region of
convergence.
1
b) Expand f (z) = in Taylor series about z = 0 and determine the region of convergence.
1+z

5.2 Laurent Series

A function f (z) that is single-valued and analytic in the annulus r < |z z0 | < R (including the
cases where r = 0 and R = +1) can be expanded in the annulus in the Laurent series

f (z) = ⌃1
n= 1 an (z z0 )n = ⌃n=1 1 an (z z 0 ) n + ⌃1
n=0 an (z z0 )n

Z
1 f (z)
where the expansion coefficients are an = dz n = 0, ±1, ±2, .... Here is an
2⇡i (z z0 )n+1
arbitrary circle centred at z0 and lying entirely inside the annulus. The series
c n
⌃n=1 1 (z z0 )n = ⌃1 n=1 is called the principal part of the Laurent series, while the series
(z z0 )n
⌃1n=0 cn (z z0 )n is called the regular part of the Laurent series.

Since it is hard to compute the coefficients using the formula, we will find the coefficients
I using other
1 f (z)
tricks or familiar series. To make it easier we will avoid to use the formula an = dz
2⇡i (z 1)n+1
in this course. When working Laurent problem we have to follow a couple of easy steps listed below

35
1. Whenever possible, use a known Taylor series as a help (i.e be familiar with the series
1
1 z
, sin z, ez etc).

2. Start by concentrating on the annular region where it is centred, is it bounded or unbounded?.

Then determine the constants ak and plug them in the formulae ⌃n=1 1 an (z z0 )n +⌃1 n=0 an (z
z0 )n , where z0 is the center of the series.

1
1. Expand f (z) = is a Laurent series valid for 0 < |z 1| < 2.
(z 1)2 (z + 1)2

Solution
1
The function f (z) = is analytic in the annulus 0 < |z 1| < 2.
(z 1)2 (z
+ 1)2
The annular region tells us that the series is centred at 1, so that means our Laurent series
should be in terms of powers of z 1. We notice that f (z) already has z 1 in the denominator,
1
so we have to only expand .
(z + 1)2
1 1
For now lets ignore the power of 2
and expand valid for 0 < |z 1| < 2. Our
(z + 1) z+1
1
powers should be in terms of z 1, so we manipulate , thus
z+1
2 3
"1 ✓ ◆ # 1
1 1 16 1 7 1 X 1
n X ( 1)n
= = 6 ✓ ◆ 7= (z 1) n
= (z 1)n .
z+1 (z 1) + 2 24 z 1 5 2 n=0 2 2 n+1
1 n=0
2
1
Thus to get the expansion for , we di↵erentiate both sides and get
(z + 1)2
1
X ( 1)n
1
= n(z 1)n 1
(z + 1)2 n=0
2n+1

Therefore
X1
1 1 ( 1)n
f (z) = = n(z 1)n 1
(z 1)2 (z + 1)2 (z 1)2 n=0 2n+1
X1
( 1)n
= n(z 1)n 3

n=0
2n+1
1
X ( 1)n
= n+4
(n + 3)(z 1)n
n= 3
2

1
2. For the function f (z) = find the Laurent series expansion in
(z 1)(z 2)
(a) the region |z| < 1 (b) region 1 < |z| < 2 (c) the region |z| > 2

36
 1 1 1
NB =1 z + z2 ... + ( 1)n z n and f (z) = .
1+z z 2 z 1

Solution
1 1 1
a) Expansion in the circle |z| < 1: f (z) = z + . But
2 (1 2 ) 1 z

1 1 1 1 z k zk 1
( )= ⌃ ( ) = ⌃1 and = ⌃1
k=0 zk .
2 1 z2 2 k=0 2 k=0
2 k+1 1 z

1 1 1 3 7
Thus f (z) = = ⌃1
k=0 (1 )z k = + z + z 2 + ....
(z 1)(z 2) 2k+1 2 4 8

(b) For 1 < |z| < 2 the series in a) (z 2) 1 remains convergent in the annulus since |z| < 2,
while the series representing (1 z) 1 is divergent for |z| > 1. For this reason we write f (z)
as follows
1 1 1 1 1 zk 1 1 1 1 z
f (z) = ( ) = ⌃ k=0 k+1 ⌃1k=0 k+1 = ... ....
2 1 z2 z (1 z1 ) 2 z z2 z 2 4

(c) For |z| > 2 the series in (b) representing (z 2) 1 we write

1 1 1 1 1 1 2 k 1 1 1 1 2k 1 2k 1
f (z) = = ⌃ ( ) ⌃ = ⌃ ⌃1 = ⌃1 .
z (1 z2 ) z (1 z1 ) z k=0 z z k=0 z k k=0 k+1
z k=0
z k+1 k=0
z k+1

37
Tutorial 5.2.1. Attempt all questions

1
1. Expand f (z) = in Taylor series about z = 0 and determine the region of convergence.
1 z
1
2. Expand f (z) = in Taylor series about z = 0 and determine the region of convergence.
1+z
3. Use Taylors Theorem to show that 0 < |z| < 4
1 1 P1 z n
= + n=0 n+2 .
4z z 2 4z 4
4. Use Taylors Theorem to show that 0 < |z 1| < 4
z 1 P (z 1)n
= 3 1n=0 .
(z 1)(z 3) 2(z 1) 2n+1
1
5. For the function f (z) = , find the Laurent series expansion in
(1 z)(2 + z)
(a) the region |z| < 1
(b) the region 1 < |z| < 2
(c) the region |z| > 2
2z 3
6. Expand in Laurent series the function f (z) = in the neighbourhood of the singular
z2 3z + 2
points of f (z).

38
Chapter 6

Residues of a Function

Definition 6.0.1. A point z0 is said to be an isolated singular point or simply an isolated singularity
of a function if there is a neighbourhood of this point in which the function is analytic everywhere
except at z = z0

6.0.1 Types of singularities

1. Removable singularities An isolated singular point of a function f (z) is said to be removable

if at this point f (z) has a finite limit, that is, lim f (z) = A(finite).
z!z0
1 cosz
Example f (z) = at z = 0.
z2
2. Poles A point z0 is said to a pole of a function f (z) if there exist a positive integer n such
z
that lim (z z0 )n f (z) = A 6= 0, e.g. f (z) = at z = 1
z!z0 z+1
3. Essential Singuality A point z0 is said to be an essential singularity of a function f (z) if
1
there is it is a singular that is neither a removable singularity nor a pole, e.g. f (z) = e z at
z = 0.

Definition 6.0.2. The residue of an analytic function f (z) Z at an isolated singularity z0 is defined
1
as a complex number equal to the value of the integral f (⌘)d⌘ taken in the positive direction
2⇡i
around any closed curve lying in the domain Z of analyticity of the function f (z) and containing a
1
unique singularity z0 denoted Resz0 f = f (⌘)d⌘.
2⇡i

Note: Comparing the residue of a function with the coefficient of the Laurent series in powers of
(z z0 ) we see that the following equality holds:

39
 Z
1
Residuez0 f = Resz0 f = f (⌘)d⌘ = c 1 .
2⇡i

6.1 Methods for Computing a residue

(a) Let the function f (z) have a first order pole at the point z0 , that is, let it be represented in the
c 1
form of the series f (z) = + ⌃1 k=0 ck (z z0 )k .
z z0
Then multiplying both sides of this equality by z z0 and passing to limit as z ! z0 , we get,

c 1 = Resz0 f = lim f (z).

z!z0

'(z)
(b) If the function f (z) = where '(z0 ) 6= 0, (z0 ) = 0, 0 (z0 ) 6= 0, then applying formula in
(z)
a) and L’Hospital’s rule, we find
(z z0 )'(z) '(z0 ) '(z0
c 1 = Resz0 = lim = , that is, Resz0 f = .
z!z0 (z) (z) (z0 ) (z0 )
lim
z!z0 z z0
(c) Let the function f (z) have a pole of order m z at the point z0 and represent it in the form
c k c 1
f (z) = ⌃2k=m k
+ + ⌃1k=0 ck (z z0 )k multiplying both sides by (z z0 )m , di↵erentiating
(z z0 ) z z0
m 1 times and passing to limit as z ! z0 , we get
1 dm 1 [(z z0 )m f (z)]
Resz0 f = c1 = lim .
(m 1)! z!z0 dz m 1
1
Example 6.1.1. Find Resz0 f where f (z) = , z0 = i.
(1 + z 2 )2

6.2 Applying Residues to evaluate definite integrals

Theorem 6.2.1. Let f be analytic everywhere on and inside a simple closed contour C excepts as
aZ finite number of poles a1 , a2 , ..., an inside C. Then
f (z)dz = 2⇡i⌃nk=1 Resak f , where ⌃Reszk f is the sum of the residues of f (z) at its poles within
C
C.

Proof
Exercise(Schaum’s Outline page 210)
Z
dz
Example 6.2.1. Evaluate being taken in the positive direction along the circle |z| = 1.
C zcosz

40
Solution
1 3⇡ 5⇡ 7⇡
The function f (z) = is not analytic at 0, ± , ± , ± , ..., but only z = 0 lies inside the
z cos z 2 2 2
circle |z| = 1. Thus from the Residue theorem
Z 
dz 1
=2⇡i Resz=0
C z cos z z cos z
h z i
=2⇡i lim
z!0 z cos z
1
=2⇡i lim
z!0 cosz

=2⇡i

6.3 Evalaution of improper real integrals

We will apply the residue theorem to compute real integrals of the form

Z 2⇡ Z 1 Z 1 Z 1
(a) f (cos ✓, sin ✓) (b) f (x)dx (c) f (x) cos ↵xdx or f (x) sin ↵xdx
0 1 1 1

In all the forms of integrals above, we are going to follow the following steps/ideas

1. Choose a closed contour in C and a related integrand.

2. Use the Residue theorem for the contour integral.
3. Pick out the value of our desired integral

Z 2⇡
6.4 Evaluation of integrals of the form R(cos ✓, sin ✓)d✓
0

Z 2⇡
Consider the integral f (cos ✓, sin ✓)d✓. We are going to let C be the unit circle z(✓) = ei✓ .
0
Using the identity ei✓ = cos ✓ + i sin ✓ and e i✓ = cos ✓ + i sin ✓ and making the substitution z = ei✓ ,
1 1 1 1
we obtain the relationships cos ✓ = (z + ), sin✓ = (z ), dz = izd✓.
2 z 2i z
From
Z the obtained equalities we conclude that the desired integral is transformed to to the form
f (z)dz, where f (z) is a rational function, having only poles, and the interval can be evaluated
|z|=1
with the aid of the residue theorem.

41
 Z 2⇡
d✓
Example 6.4.1. Evaluate I = .
0
5
+ sin ✓
4

Solution
We are going to let C be the unit circle z(✓) = ei✓ with 0  ✓  2⇡.
Now
1 i✓
sin ✓ = (e e i✓ ) and dz = iei✓ d✓
2i
1 dz
sin ✓ = (z z 1 ) ) d✓ =
2i iz
Thus after
Z substituting sin ✓ and d✓ the real integral equals to the contour integral
dz
I = ✓ ◆
C 5 z z 1
iz +
Z 4 2i
4
= dz
Z 2z 2 + 5zi 2
4
= dz
C (z + 2i)(2z + i)

We now use the Residue theorem to find the contour integral.

i i
The simple poles are at z = 2i and z = , but only the pole z = is enclosed by the contour
2 2
i
C, so we only find the residue at z = . Thus
2


i 4 4
Resz= i = limi z ( ) =
2 z= 2
2 (z + 2i)(2z + i) 3i

Therefore

Z
4
I= dz =2⇡iResz= i
C (z + 2i)(2z + i) 2

4
=2⇡i( )
3i
8
=
3⇡

42
 Z 1
6.5 Integrals of the form f (x)dx
1

Recall that

Z 1 Z 0 Z 1 Z 0 Z R
f (x)dx = f (x)dx + f (x)dx = lim f (x)dx + lim f (x)dx
1 1 0 R!1 R R!1 0

If both limits exist, the improper integral converges.

The Cauchy principal value of an integral is defined as following

Z 1 Z R
f (x)dx = lim f (x)dx
1 R!1 R

Note that if an improper integral converges in the usual sense then it does converge in the Cauchy
principal value sense. However, if an integral converges
Z in the Cauchy principal value sense it may
1
not converge in the usual sense. For example sin x does not exist since sin x is oscillating, but
Z 1 1

the principle vale of sin x = 0 because sin x is an odd function and the area from 1 to 0 is
1
opposite of that of 0 to 1.
Z 1
2x2 1
Example 6.5.1. dx
0 x4 + 5x2 + 4

Solution
Let C be the contour shown below made of two parts C1 (the line segment) and C2 (the semicircular
arc).

Note that the line segment C1 corresponds to that of our integral and since we have to choose a
closed contour we choose C2 resulting in a semi circle arc.
Thus

43
 Z ✓Z R Z ◆
2z 2 1 2z 2 1 2z 2 1
lim dz = lim dz + dz
R!1 C z 4 + 5z 2 + 4 R!1 4 2
R z + 5z + 4
4 2
C2 z + 5z + 4
Z 1 Z
2x2 1 2z 2 1
=P V dx + lim dz
1 x4 + 5x2 + 4 R!1 C z 4 + 5x2 + 4
2

Z
2z 2 1
We now evaluate dz using Residue theorem. We can see that
C z 4 + 5z 2 + 4

2z 2 1 2z 2 1
f (z) = =
(z 2 + 1)(z 2 + 4) (z i)(z + i)(z 2i)(z + 2i)

Thus we have poles at z = ±i and ±2i. We can see that only poles i and 2i are enclosed by the arc
C, so we find the residues at i and 2i.

2i2 1 3 1
Resi = = =
(2i)( i)(3i) 2 ⇥ 3i 2i
2(2i)2 1 8 1 9 3
Res2i = = = =
(i)(3i)4i) 12i 12i 4i
1 3 1
Sum of residues = + = .
2i 4i 4i
Thus

Z
2z 2 1
dz =2⇡i[Sum of residues]
C z 4 + 5z 2 + 4
✓ ◆
1
=2⇡i
4i
⇡
=
2

Now since our contour integral was represented as

Z ✓Z R Z ◆
2z 2 1 2z 2 1 2z 2 1
lim dz = lim dz + dz
R!1 C z 4 + 5z 2 + 4 R!1 4 2
R z + 5z + 4
4 2
C2 z + 5z + 4
Z 1 Z
2x2 1 2z 2 1
=P V dx + lim dz
1 x4 + 5x2 + 4 R!1 C z 4 + 5x2 + 4
2

44
Thus we have

Z 1 Z
⇡ 2x2 1 2z 2 1
= PV dx + dz
2 1 x4 + 5x2 + 4 C2 z 4 + 5x2 + 4

Z
2z 2 1
We now show that lim dz = 0 on C2 where |z| = R.
R!1 C z 4 + 5x2 + 4
2
Since C2 is a semi circular arc, we are going to bound the integral using the M L inequality.

Note that |2z 2 1|  2|z|2 + 1 = 2R2 + 1. So |z 4 + 5z 2 + 4| = |(z 2 + 1)(z 2 + 4)| = |z 2 + 1||z 2 + 4|

||z 2 | Z |1||||z|24| = (R2 1)(R2 4). Z
(2R2 + 1)⇡R
So | f (z)dz|  ! 0 as R ! 1. Since lim | f (z)dz| ! 0 we have
CR (R2 1)(R2 4) R!1 CR
Z Z 1 Z R
⇡
lim f (z)dz = 0. Therefore f (x)dx = lim f (x)dx = .
R!1 C 1 R!1 R 2
R Z 1 2 Z 1 2
2x 1 1 2x 1 1 ⇡ ⇡
Since f (x) is even 4 + 5x2 + 4
dx = 4 + 5x2 + 4
dx = . = .
0 x 2 1 x 2 2 4

In general we have the following

Lemma 6.5.0.1. If the function f (z) is analytic in the half plane Im z > 0, except for the points
M
z1 , z2 , ..., zn , and if for all |z| > R there takes place the estimate |f (z)| < (M and > 0
|z|1+
constants
Z ), then the following relationship holds true:
lim f (⌘)d⌘ = 0 where CR is a semi-circle |z| = Rof radius R, lying in the upper half-plane.
R!1 CR

Z 1 Z 1
6.6 Integrals of the form f (x)cos xdx or f (x)sin xdx
1 1

In computing these integrals we have to use the following key ideas.

key ideas

Remember Euler’s formula ei↵x = cos ↵z + i sin ↵z, so

Z Z Z
i↵z
f (z)e = f (z) cos ↵zdz + i f (z) sin ↵zdz
C C C

45
 Z
Thus it is easier to evaluate f (z)ei↵z .
C

We will use Jordan’s Lemma to evaluated these forms of integrals.

Theorem 6.6.1. Jordan’s Lemma If f (z) is analytic in the upper half-plane (0 < arg z < ⇡)
Z number of singular points and f (z) ! 0 as |z| ! 1 in the half plane,
except perhaps for a finite
then for > 0, lim f (z)ei z dz = 0 where CR is a semi-circle in the upper half-plane with
R!1 CR
radius R and the centre the origin.

M
In general, if |f (z)|  where k > 0 and M is a constant, then Jordan’s Lemma is satisfied.
Rk
Z 1
xsin x
Example 6.6.1. Find dx
1 x2 + 2x + 2

Solution
Let C be the contour shown below made of two parts C1 (the line segment) and C2 (the semicircular
arc).

Thus

Z ✓Z R Z ◆
zeiz z(cos z + i sin z) zeiz
lim dz = lim dz + dz
R!1 C z 2 + 2z + 2 R!1 R z 2 + 2z + 2 2
C2 z + 2z + 2
Z 1 Z 1 Z
x cos x x sin x zeiz
=P V dx + P V i dx + lim dz
1 x2 + 2x + 2 2
1 x + 2x + 2 R!1 C z 2 + 2z + 2
2

Z
zeiz
We now evaluate 2
dz using the Residue theorem.
C z + 2z + 2
z z
We can see that consider the functions F (z) = f (z)eiz where f (z) = = .
z2 + 2z + 2 (z ↵)(z )
The poles are at z = 1 ± i.
Let ↵ = 1 + i and = 1 i. In the upper half plane we have one pole z = 1 + i, so we only
find the residue at z = 1 + i.

46
 ( 1 + i)ei(1+i)
Resz=↵ = .
2i

Thus
Z X
zeiz
dz =2⇡i Residues
C z 2 + 2z + 2
( 1 + i)e i 1
=2⇡i = ⇡( 1 + i)e i e 1
2i
⇡ ⇡
= ( 1 + i)e 1 = ( 1 + i)(cos 1 isin 1)
e e
⇡
= [(sin 1 cos 1) + i(cos 1 + sin 1)]
e
So

⇡
[(sin 1
cos 1) + i(cos 1 + sin 1)] =
Z 1
e Z 1 Z
x cos x x sin x zeiz
PV dx + P V i dx + lim dz
1 x2 + 2x + 2 2
1 x + 2x + 2 R!1 C z 2 + 2z + 2
2

Z
zeiz
By Jordan’s Lemma lim dz = 0. Thus
R!1 C2 z 2 + 2z + 2

Z 1 Z 1
⇡ x cos x x sin x
[(sin 1 cos 1) + i(cos 1 + sin 1)] = P V 2
dx + P V i dx
e 1 x + 2x + 2 1 x2 + 2x + 2

Now equating imaginary parts we have

Z 1
x sin x ⇡
dx = [(cos 1 + sin 1)
1 x2 + 2x + 2 e

47
Tutorial 6.6.1. Attempt all questions

1. Find the residues at the singular points of the following functions

z2 z + 1 ez z 2 + 4z + 1
(a) f (z) = (b) f (z) = 2 (c) f (z) =
(z 1)(z 4)(z + 3) z (z + ⇡i)4 z 4 3z 2 4
z3 4 eiz
(d)f (z) = 4 (e) f (z) = 4
z + 2z 2 + 1 z + 2z 2 + 1
aiz
e
(f ) f (z) = 2 , a 6= b
z + (a + b2 )z + a2 b2
2

Z 1
dx ⇡
2. Prove that if a > 0 4 4
= p .
0 x +a 2 2a3
Z 1 2
x +1
3. Evaluate the integral I = dx.
0 x4 + 1
Z 1
x2 dx 7
4. Show that 2 2 2
= .
1 (x + 1) (x + 2x + 2) 50⇡
Z 1
x2 ⇡
5. Show that 2 + 1)2
dx = .
0 (x 4
Z 1
sin ↵x
6. Evaluate the integral I = dx.
0 x
Z 1
(x 1) cos 5x
7. Evaluate 2
dx.
1 x 2x + 5
Z 1
x sin x
8. Evaluate dx.
0 (x + 1)(x2 + 4)
2

Z 1
x sin x ⇡
9. Show that 2
dx = e 2 .
0 x +4 2
Z 2⇡
d✓
10. (a) Evaluate
Z0 2⇡5 + 3 sin ✓
cos 3✓ e3✓ + e 3✓ z3 + z 3
(b) Show that d✓ Hint: cos 3✓ = = .
0 5 4 cos ✓ 2 2

48
Chapter 7

Introduction to Di↵erential Equations

A drunkard may not know which number is larger, 2/3 or 3/5, but he knows that 2 bottles of vodka for 3
people is better than 3 bottles of vodka for 5 people.

—Edward Frenkel

Many of the basic laws of the physical sciences, biological and social sciences are formulated in terms
of mathematical equations involving certain known and unknown quantities and their derivatives.
Such equations are called di↵erential equations.

The term general solution will be used for a family of solutions of the di↵erential equation
containing one or more arbitrary constants.

A particular solution will be any solution that is free of arbitrary constants.

A di↵erential equation involving only ordinary derivatives (derivatives of a function of one vari-
able) is called an ordinary di↵erential equation.
Theorem 7.0.2. The order of a di↵erential equation is defined as the order of the highest derivative
appearing in the equation.
Example 7.0.2. The following are di↵erential equations with indicated orders.

dy
1. = ay (first order).
dx
2. x00 (t) 3x0 (t) + x(t) = cos t (second order).
3
3. (y (4) ) 5 2y 00 = cos x (fourth order).

A di↵erential equation is said to be linear if it has the form

dn y dn 1 y dy
p0 (x) + p 1 (x) + · · · + pn 1 (x) + pn (x)y = f (x)
dxn dxn 1 dx
49
where p0 (x), p1 (x), · · · , pn (x) are functions of x only.

A function that satisfies the di↵erential equation when substituted is called a solution of the
equation. Sometimes referred to as the integral of the equation and its graph is called an integral
curve or solution curve.

7.1 First Order Di↵erential Equations

There are various ways of solving first order ordinary di↵erential equations and these include

1. Separation of variables technique

2. Integrating factor method

3. Exact first order D.Es

7.1.1 Separation of variables technique

Simplest type of first order di↵erential equation and is of the form

dy
= f (x). (7.1)
dx
This equation is a first order autonomous equation and first order autonomous equations can always
be solved by the separation of variables technique. The technique involve the following procedures

1. Isolate the derivative of the unknown function on one side of the equation.

2. Divide by the quantity on the other side of the equation.

3. Integrate both sides of the resulting equation making use of the Fundamental Theorem of
Calculus and the initial condition.

Solving equation (1.1) by integrating both sides gives a solution of the form
Z
y(x) = f (x) dx = F (x) + c.

dy 1
Example 7.1.1. Solve =
dx
Z 1 + x2
dx 1
We can see that y(x) = = tan x + c.
1 + x2

50
Note that equations of the form

dy f (x)
= or f (x)dx = g(y)dy
dx g(y)

are also separable.

dy
Example 7.1.2. Solve = x(1 + y 2 ).
dx

Solution: Dividing by 1 + y 2 and integrating with respect to x, we get

Z ✓ ◆ Z
1 dy
dx = x dx
y + y 2 dx
Z Z
dy
= x dx
1 + y2
1
) tan 1 y = x2 + c where c is a constant.
2 ✓ ◆
1 2
Therefore y = tan x +c .
2

Homogeneous and non-homogeneous ODE

A first-order linear equation has the form

dy
+ a(x)y = f (x).
dx
If f (x) is the zero function, the linear di↵erential equation is said to be homogeneous, otherwise
we say it is non-homogeneous.

dy
7.1.2 Solution of the homogeneous Equation + a(x)y = 0.
dx

Separating variables, we have Z Z

dy
= a(x)dx + c.
y
Integrating, we have Z
ln |y| = a(x)dx + c

and R
a(x)dx
y = c1 e .

51
Example 7.1.3. Solve the homogeneous di↵erential equation

y 0 + 3y = 0.

Solution: Rewriting the equation as y 0 = 3y, and separating variables, we have

Z Z
dy
= 3 dx + c
y
ln |y| = 3x + c
y = c1 e 3x .

7.1.3 Integrating factor method

Another method of solving the general first order linear non homogeneous autonomous and non
autonomous equations is the integrating factor method. The following steps are to be followed
in using the integrating factor technique to solve first order non homogeneous, non autonomous
equations.

1. Rearrange the given equation as necessary to be in the form

d
x(t) + a(t)x(t) = b(t).
dt
The coefficient of the derivative term must always be 1.
R
a(t)dt
2. Compute the integrating factor e .

3. Multiply both sides of the equation obtained in step 1 by the integrating factor. The result
is guaranteed to be of the form
d R R
[x(t)e a(t)dt ] = b(t)e a(t)dt
dt

4. Integrate both sides and solve for the unknown function. Make use of the initial conditions
and the fundamental theorem of calculus.
Example 7.1.4. Solve the equation y 0 = y + x2 .

Solution: Rewriting the equation as

y0 y = x2 , (7.2)
R
dx
we see that a(x) = 1, thus the integrating factor is e = e x . Multiplying both sides of (7.2) by
e x , we get
e x (y 0 y) = x2 e x

or
d x
ye = x2 e x .
dx
52
Integrating both sides, we get
Z
x
ye = x2 e x
dx + c

= c (x2 + 2x + 2)e x
y = ce x (x2 + 2x + 2).
✓ ◆
dy
Example 7.1.5. Solve x + 4y = x6 .
dx

Solution: Dividing by x, we have ✓ ◆

dy 4
+ y = x5 .
dx x
Z Z
4 4 R 4
Hence a(x) = and so a(x)dx = dx = 4 ln x = ln x4 . Thus e a(x)dx = eln x = x4 .
x x
Multiplying by x4 , we get
✓ ◆
dy d
x4 + 4x3 y = x9 , or x4 y = x9 .
dx dx
x10
Integrating, we have x4 y = + c. Hence
10
x6 c
+ 4. y=
10 x
Exercise 7.1.1. Find the general solution of the following equations.
(a) 3xy 0 y = ln x + 1 (b) (sin x)y 0 + (cos x)y = 0 (c) (3x2 + 1)y 0 2xy = 6x
p dy
(d) (x2 + 1)y 0 + xy = (1 2x) x2 + 1 (e) x sin x + (sin x + x cos x)y = xex
dx
dy 2 dy
(f ) sin x cos x + y = tan x (g) (1 + sin x) + 2(cos x)y = tan x.
dx dx

7.1.4 The Bernoulli Equation

Definition 7.1.1. A Bernoulli equation is the equation of the form

dy
+ a(x)y = f (x)y n , n 6= 0, 1. (7.3)
dx

When n = 0, 1 the Bernoulli equation becomes a linear equation.

Set z = y 1 n
. Then z 0 = (1 n)y n 0
y . So if we multiply both sides of equation (7.3) by (1 n)y n
,
we obtain
n 0
(1 n)y y + (1 n)a(x)y 1 n
= (1 n)f (x)
or
dz
+ (1 n)a(x)z = (1 n)f (x).
dx
This equation is now linear and can be solved as before.

53
Example 7.1.6. Solve
dy y 5 2 3
= xy .
dx x 2

Solution: Here n = 3 so let z = y 2 . Let z 0 = 2y 3 y 0 and multiply both sides of the equation by
2y 3 to obtain
2
2y 3 y 0 + y 2
= 5x2
x
or
2z
z0 + = 5x2 . (7.4)
x
R dx 2
The integrating factor for this linear equation is e2 x = e2 ln x = eln x = x2 . Multiplying both sides
of equation (7.4) by x2 , we have x2 z 0 + 2xz = 5x4 or (x2 z)0 = 5x4 . Thus
Z
x z = 5 x4 dx + c = x5 + c.
2

Hence
1
2
y = z = x3 + cx 2
or y = (x3 + cx 2 ) 2 .

7.1.5 Exact Equations

Given the di↵erential equation

M (x, y)dx + N (x, y)dy = 0. (7.5)
Here we use partial derivatives to solve the ordinary di↵erential equations of this form.
From calculus the total di↵erential dg of a function of two variables g(x, y) is defined by
@g @g
dg = dx + dy.
@x @y
If we can find a function g(x, y) such that
@g @g
= M, = N,
@x @y
then (7.5) becomes dg(x, y) = 0. In this case M dx + N dy is said to be an exact di↵erential and
(7.5) is called an exact di↵erential equation.

To determine if a di↵erential equation is exact we use the cross-derivative test.

Theorem 7.1.1. The equation

M (x, y)dx + N (x, y)dy = 0

is exact if and only if

@M @N
= .
@y @x

54
If M dx + N dy is exact, then we can solve the di↵erential equation (7.5) by finding the function g
above.

Example 7.1.7. Solve the equation

(1 sin x tan y)dx + (cos x sec2 y)dy = 0.

Solution: Letting M (x, y) = 1 sin x tan y and N (x, y) = cos x sec2 y, we have

@M @N
= sin x sec2 y = ,
@y @x
@g
so the equation is exact. We now seek a function g(x, y) of two variables, such that = M and
@x
@g
= N , then
@y
Z Z
g(x, y) = M dx = (1 sin x tan y)dx
= x + cos x tan y + h(y).

The constant of integration h(y) is an arbitrary function of y since we must introduce the most
general term that varnish under partial di↵erentiation with respect to x. But
@g
cos x sec2 y = N (x, y) = = cos x sec2 y + h0 (y).
@y

This means h0 (y) = 0 so h(y) = k, a constant. The general solution is

g(x, y) = x + cos x tan y + k

7.2 Initial Value Problems (IVPs)

We can solve a first order di↵erential equation

dy
= f (x, y)
dx
subject to the condition y = y0 when x = x0 or y(x0 ) = y0 .

This is an example of an initial value problem and the condition y(x0 ) = y0 is called an initial
condition and x0 is called the initial point.
dy
Example 7.2.1. (a) = 2y 3x, y(0) = 2. (x = 0 is the initial point)
dx
(b) x00 (t) + 5x0 (t) + sin x(t) = 0, x(1) = 0, x0 (1) = 7. (t = 1 is the initial point)

55
7.3 Boundary Value Problems (BVPs)

A boundary value problem consists of a di↵erential equation and a collection of values that must
be satisfied by the solution of the di↵erential equation or its derivatives.
d2 y
Example 7.3.1. (a) + 5xy = cos x, y(0) = 0, y 0 (1) = 2.
dx2
dy
(b) + 5xy = 0, y(0) = y(1) = 2.
dx

56
Chapter 8

Second order di↵erential equations

A second order d.e is linear if it can be written in the form

y 00 + p(x)y 0 + q(x)y = r(x).

The equation is linear in the unknown function y and its derivatives, whereas p(x), q(x)[and] r(x)
may be any given functions of x.

Superposition principle
If y1 (x) and y2 (x) are any two solutions of a linear homogeneous di↵erential equation

y 00 + p(x)y 0 + q(x)y = 0,

then c1 y( x) + c2 y2 (x) is also a solution to the di↵erential equation for any constants c1 and c2 .

Proof. Exercise.

Linear independence and dependence of solutions

Suppose y1 and y2 are defined on [a, b] then they are linearly dependent if there exists c1 and c2
(c1 , c2 6= 0) such that
c1 y1 + c2 y2 = 0
otherwise they are linearly independent.

1. y1 and y2 are linearly independent if c1 y1 + c2 y2 = 0 ! c1 and = 0.

2. y1 and y2 are linearly dependent i↵ one is a constant multiple of the other. i.e c1 + c2 =
0, c1 , c2 6= 0 ! y1 = cc21 y2 , y1 is a constant multiple of y2 .

Definition 8.0.1. A basis of y 00 + p(x)y 0 + q(x)y = 0 on an interval [a, b] is a pair y1 , y2 of linearly

independent solutions of the di↵erential equation on the interval [a, b].

57
Example 8.0.2. Show that y1 = cos x and y2 = sin x are linearly independent solutions of the D.E.
y 00 + y = 0
Assuming everyone can show that the two are solutions of the D.E, cos x and sin x are not propor-
tional, i.e yy12 = cot x 6= constant. Hence they form a basis of the D.E for all x and a general
solution takes the form y = c1 cos x + c2 sin x..

We now look at a condition under which y1 and y2 are linearly independent.

The wronskian

Suppose the functions y1 and y2 are di↵erentiable on [a, b], the wronskian of y1 and y2 is the
determinant
y1 (x) y2 (x)
W (x) = = y1 y20 y10 y2
y10 (x) y20 (x)

8.1 Homogeneous second order equations

We discuss the method of solving second order linear equations with constant coefficients.
Example 8.1.1. Determine solutions to y 00 9y = 0
We can get some solutions here simply by inspection. We need functions whose second derivative is
9 times the original function. One of the first functions that I can think of that comes back to itself
after two derivatives is an exponential function and with proper exponents the 9 will get taken care
of as well. So, it looks like the following two functions are solutions.

y(t) = e3x and y(x) = e 3x

In fact if you think about it any function that is in the form y(x) = c1 e3x + c2 e 3x will be a solution
to the di↵erential equation. This example leads us to the principle of superposition.

In order to solve the linear homogeneous D.E. with constant coefficients, the general method is
as follows:

• Try a solution of the form emx . This produces the characteristic equation (a quadratic eqn in
this case) to be solved for m.

• Solve the characteristic equation.

• The general solution of the D.E is obtained from the roots of the characteristic equation.

• Make use of initial conditions to determine the unknown constants in the general solution.

58
 d2 x dx
Example 8.1.2. Solve 2
= 4 3x
dt dt
Let x(t) = e , then, x (t) = me , x00 (t) = m2 emt , substituting in the D.E we have
mt 0 mt

m2 emt + 4memt + 3emt = 0

(m2 + 4m + 3)emt = 0, ! (m2 + 4m + 3) = 0
m = 3 orm = 1
hence
x(t) = c1 e 3t + c2 e t .

As we have noted, the characteristic equation is quadratic and so will have two roots, m1 and m2 .
The roots will have three possible forms. These are

1. Real, distinct roots, m1 6= m2 .

2. Complex roots, m1,2 = a ± bi.

3. Double roots,m1 = m2 = m.

8.1.1 Case 1 Real, distinct roots, m1 6= m2 .

Example 8.1.3. Solve the following IVP. y 00 + 11y 0 + 24y = 0, y(0) = 0, y 0 (0) = 7
Solution
The characteristic equation is
m2 + 11y 0 + 24y = (m + 8)(m + 3) = 0
whose roots arem1 = 8 and m2 = 3 and so the general solution is
y(t) = c1 e 8t + c2 e 3t and its derivative is
y 0 (t) = 8c1 e 8t 3c2 e 3t
Now, plug in the initial conditions to get the following system of equations.

0 = y(0) = C1 + C2
(8.1)
7 = y 0 (0) = 8C1 3C2

giving C1 = 7/5 and C2 = 7/5.

The actual solution to the di↵erential equation is then
8t 3t
y(t) = 7/5e 7/5e

Exercise 8.1.1. Solve the following initial value problems

1. y 00 + 3y 0 10y = 0 y(0) = 4, y 0 (0) = 2.

2. 3y 00 + 2y 0 8y = 0 y(0) = 6, y 0 (0) = 18.

3. 4y 00 5y 0 = 0 y( 2) = 0, y 0 (2) = 7.

59
8.1.2 Case 2 Complex roots, m1,2 = a ± bi.

Consider the D.E

ay 00 + by 0 + cy = 0
whose characteristic equation is am2 + bm + c = 0, whose roots are complex in the form
m1,2 = ± µi
Now, recall that we arrived at the characteristic equation by assuming that all solutions to the
di↵erential equation will be of the form y(t) = emt .
Plugging our two roots into the general form of the solution gives the following solutions to the
di↵erential equation. y1 (t) = e +µi , y2 (t) = e µi
Recall Euler’s formula
ei✓ = cos ✓ + i sin ✓.
A nice variant of Euler’s formula that we also need is
i✓
e = cos( ✓) + i sin( ✓) = cos ✓ i sin ✓.
Now, split up our two solutions into exponentials that only have real exponents and exponentials
that only have imaginary exponents. Then use Euler?s formula, or its variant, to rewrite the second
exponential.
y1 (t) = e +µi = e t (cos(µt) + i sin(µt))
µi
y2 (t) = e = e t (cos(µt) i sin(µt))
This does not eliminate the complex nature of the solutions, but it does put the two solutions
into a form that we can eliminate the complex parts. Recall from superposition/linearity principle,
that the general solution can be written as a combination of the two solutions in other words
y(t) = C1 y1 (t) + C2 y2 (t) will also be a solution. Using this, note that if we add the two solutions
together we will arrive at
y1 + y2 = 2e t cos(µt).
This is a real solution and just to eliminate the 2 let’s divide everything by 2.
Thus U (t) = 1/2y1 (t)+1/2y2 (t) = e t cos(µt). Note that this is equivalent to taking C1 = C2 = 1/2.

Now we can arrive at a second solution in a similar manner. This time let’s subtract the two original
solutions to arrive at
y( t) y2 (t) = 2ie t sin µt
On the surface this doesn’t appear to fix the problem as the solution is still complex. However,
upon learning that the two constants, c1 and c2 can be complex numbers we can arrive at a real
solution by dividing this by 2i. This is equivalent to taking c1 = 2i1 and c2 = 2i1 .
Our second solution will be
1 1
V (t) = y1 (t) y2 (t) = e t sin(µt)
2i 2i
and we now have two solutions to the D.E.

So if the roots of the characteristic equation happen to be m1,2 = ± µi, the general solution to
the D.E is
y(t) = C1 e t cos µt + C2 e t sin(µt).

60
Example 8.1.4. Solve the I.V.P y 00 4y 0 + 9y = 0, y(0) = 0, y 0 (0 = 8) p
2
The characteristic equation for this D.E is m 4m + 9 = 0 whose roots are m1,2 = 2 ± i 5.
The general solution to the D.E is then
p p
y(t) = C1 e2t cos( 5t) + C2 e2t sin( 5t)
8 p
After applying initial conditions show that p e2t sin( 5t)
5
Exercise 8.1.2. Solve

1. y 00 8y 0 + 17y = 0, y(0) = 4, y 0 (0) = 1

2. 4y 00 + 24y 0 + 37y = 0, y(⇡) = 1, y 0 (⇡) = 0

8.1.3 Case 3 Repeated roots,m1 = m2 = m.

Consider the D.E

ay 00 + by 0 + cy = 0
whose characteristic equation is am2 + bm + c = 0, whose roots are complex in the form

m1 = m2 = m

. This leads to a problem however. Recall that the solutions are

y1 (t) = em1 t = emt , y2 (t) = em2 t = emt

These are the same solution and will NOT be ”good enough” to form a general solution. Lets get
out of our way a bit.

From the quadratic formula we know that the roots to the characteristic equation are
p
b ± b2 4ac
m1,2 =
2a
In this case, since we have double roots we must have b2 4ac = 0. This is the only way we can
get double roots and in this case the roots will be m1,2 = 2ab so the one solution that we have got
bt
is y1 (t) = e 2a .
To find a second solution we will use the fact that a constant times a solution to a linear homogeneous
di↵erential equation is also a solution. If this is true then maybe we will get lucky and the following
will also be a solution
bt
y2 (t) = v(t)y1 (t) = v(t)e 2a
with a proper choice of v(t). To determine if this in fact can be done, let us plug this back into the
di↵erential equation and see what we get. We will first need a couple of derivatives.
bt b bt
y2 (t) = v 0 e 2a ve 2a
2a
61
 bt b 0 bt b bt b2 bt
y200 = v 00 e 2a v e 2a e 2a + ve 2a
2a 2a 4a2
bt b 0 bt b2 bt
= v 00 e 2a ve 2a + ve 2a
a 4a2
Now plugging these into the D.E
b 0 bt
bt b2 bt bt b bt bt
a(v 00 e 2av e 2a + 2 ve 2a ) + b(v 0 e 2a ve 2a ) + c(ve 2a = 0
a 4a 2a
We can factor out an exponential out of all the terms. Collecting all the coefficients of v and its
derivatives.
bt b2 b2
e 2a (av 00 + ( b + b)v 0 + ( + c)v) = 0
4a 2a
bt b2
e 2a (av 00 + ( + c)v) = 0
4a
bt 1 2
e 2a (av 00 (b 4ac)v) = 0
4a
Now, because we are working with a double root we know that that the second term will be zero.
bt
Also exponentials are never zero. Therefore, y2 (t) = v(t)y1 (t) = v(t)e 2a will be a solution to the
di↵erential equation provided v(t) is a function that satisfies the following di↵erential equation.
av 00 = 0 or v 00 = 0
Note that we can drop a because it cannot be zero otherwise we would not have the second order
D.E. Z Z
0 00
v = v dt = c, v(t) = v 0 dt = ct + k.

Thus the two solutions are then

bt bt
y1 (t) = e 2a , y2 (t) = (ct + k)e 2a

Thus
bt bt bt bt
y(t) = C1 e 2a + C2 (ct + k)e 2a = (C1 + C2 k)e 2a + C2 cte 2a .
Notice that we rearranged things a little. Now, c, k, c1 , and c2 are all unknown constants so any
combination of them will also be unknown constants. In particular, c1 + c2 k and c2 c are unknown
constants so we will just rewrite them as follows.
bt bt
y(t) = c1 e 2a + c2 te 2a

So if the roots of the characteristic equation are m1 = m2 = m, then the general solution is
y(t) = c1 emt + c2 temt .
Example 8.1.5. Solve the I.V.P y 00 4y 0 + 4y = 0, y(0) = 12, y 0 (0) = 3
Show that the general solution is y(t) = c1 e + c2 te and hence y(t) = 12e2t
2t 2t
27te2t
Exercise 8.1.3. Solve

9
1. 16y 00 40y 0 + 25y = 0, y(0) = 3, y 0 (0) =
4
2. y 00 + 14y 0 + 49y = 0, y( 4) = 1, y 0 ( 4) = 5

62
8.1.4 Reduction of Order

Finding solutions to non-constant coefficient, second order D.Es of the form

p(t)y 00 + q(t)y 0 + r(t)y = 0.

In general, finding solutions to these kinds of di↵erential equations can be much more difficult than
finding solutions to constant coefficient di↵erential equations. Reduction of order requires that a
solution be known. Without this known solution wo will not be able to do reduction of order.

Example 8.1.6. Find the general solution to 2t2 y 00 + ty 0 3y = 0 given that t 1 is a solution.
Once we have a first solution we will then assume that a second solution will have the form

y2 (t) = v(t)y1 (t)

for a proper choice of v(t). To determine the proper choice, we plug the guess into the di↵erential
equation and get a new di↵erential equation that can be solved for v(t). So, let’s do that for this
problem. Here is the form of the second solution as well as the derivatives that we’ll need.

y2 (t) = t 1 v(t), y20 (t) = t 2v + t 1v0, y200 (t) = 2t 3 v 2t 2 v 0 + t 1 v 00

Plugging these into the D.E. gives

2t2 (2t 3 v 2t 2 v 0 + t 1 v 00 ) + t( t 2 v + t 1 v 0 ) 3(t 1 v(t) = 0

Rearranging and simplifying gives

2tv 00 + ( 4 + 1)v 0 + (4t 1

t 1
3t 1)v = 0

2tv 00 3v 0 = 0
Note that upon simplifying, the only terms remaining are those involving the derivatives of v. The
term involving v drops out. If you’ve done all of your work correctly this should always happen.
Sometimes, as in the repeated roots case, the first derivative term will also drop out. So in order
for y2 (t) = v(t)y1 (t) to be a solution then v must satisfy

2tv 00 3v 0 = 0 (8.2)

This appears to be a problem. In order to find a solution to a second order non-constant coefficient
di↵erential equation we need to solve a di↵erent second order non-constant coefficient di↵erential
equation. However, this isn’t the problem that it appears to be. Because the term involving the v
drops out we can actually solve (8.2) and we can do it with the knowledge that we already have at
this point. We will solve this by making the following change of variable.

w = v0, w0 = v 00

with this change of variable, (8.2) becomes

2tw0 3w = 0

63
and this is a linear, first order D.E. that we can solve. This also explains the name of this method.
We’ve managed to reduce a second order di↵erential equation down to a first order di↵erential
equation.
3
w(t) = Ct 2
Recall our change of variable
v0 = w
with this we can easily solve for v(t)
Z Z
3 2 5
v(t) = wdt = Ct 2 = Ct 2 + k
5
This is the most general possible v(t) that we can use to get a second solution. So, just as we did
in the repeated roots section, we can choose the constants to be anything we want so choose them to
clear out all the extraneous constants. In this case we can use
5
C= , k=0
2
Using these gives the following for v(t) and for the second solution.
5
v(t) = t 2
5 3
y2 (t) = t 1 (t 2 ) = t 2
The general solution will then be
3
1
y(t) = C1 t + C2 t 2 .

Exercise 8.1.4. Find the general solution to

t2 y 00 + 2ty 0 2y = 0

given that y1 (t) = t is a solution.

8.2 Non-homogeneous Second order linear equations

8.2.1 Method of undetermined coefficients

Consider
y 00 + p(x)y 0 + q(x)y = r(x)

1. First find the general solution of the corresponding homogeneous equation yc .

2. Find the solution of the non-homogeneous equation, call this yp .

3. The general solution of the original non-homogeneous equation is then y = yc + yp .

64
Example 8.2.1. Solve y 00 + 3y 0 + 2y = 1 + 6x
Solution
We first find the complimentary solution yc which is the general solution to the homogeneous D.E.

y 00 + 3y 0 + 2y = 0

The characteristic equation is

m2 + 3m + 2 = 0 giving us (m + 2)(m + 1) = 0 impliying m1 = 2, m2 = 1

which leads to the solution

2x x
yc = c1 e + c2 e
Now we want to find the particular solution yp . Since r(x) is a polynomial of order 1, choose

yp = A + Bx, yp0 = B, yp00 = 0

Substituting into the original equation we have

0 + 3B + 2(A + Bx) = 1 + 6x
3B + 2A = 1
2B = 6
B=1 (8.3)
9 1 = 2A
A= 4
yp = 4 + 3x
2x x
y = yc + yp = c1 e + c2 e 4 + 3x
Example 8.2.2. Solve y 00 + 6y 0 + 9y = 27x2 , y(0) = 2, y 0 (0) = 1
Solution
For the homogeneous equation, the characteristic equation is

m2 + 6m + 9 = 0, (m + 3)(m + 3) = 0 hence m = 3 twice

3x 3x
yc = c1 e + c2 xe
For the particular solution choose
yp = A + Bx + Cx2
yp0 = B + 2Cx (8.4)
yp00 = 2C
Substitute in the original equation we have
2C + 6(B + 2Cx) + 9(A + bx + Cx2 ) = 27x2
2C + 6B + 9A = 0
12C + 9B = 0
(8.5)
9C = 27, implying C = 3
B = 4 and A = 2
yp = 2 4x + 3x2
3x 3x
y(x) = c1 e + c2 xe +2 4x + 3x2 .
Now make use of initial conditions to find c1 and c2 .

65
Example 8.2.3. Solve y 00 + 4y = 8e 2t , y(0) = 0, y 0 (0) = 2
Solution
For the complimentary solution, m = ±2i

yc = c1 cos(2t) + c2 sin(2t)

Choose
yp = M e 2t
, such that yp0 = 2M e 2t
, yp00 = 4M e 2t

Substitute in D.E we have

2t 2t 2t
4M e + 4M e = 8e , hence M = 1
2t
y = c1 cos(2t) + c2 sin(2t) + e
Applying initial conditions
3t
y(t) = 2 sin 2t cos 2t + e

Note that r(x) is a trig function e.g. y 00 4y = 13 cos 3x, then choose yp = A cos 3x + B sin 3x
If r(x) takes a combination of polynomials and trig functions, also choose yp as a combination of
those.

8.2.2 Method of variation of parameters

The technique of undetermined coefficients has severe limitations. It can only be used when the
coefficients p(x) and q(x) are constants, and even then, it only works if r(x) takes a particular
simple form. We now develop a more powerful method that always works regardless of the nature
of p, q, and r, provided the general solution of the corresponding homogeneous equation

y 00 + p(x)y 0 + q(x)y = 0 is already known.

We assume that in some way, the general solution of the corresponding homogeneous equation has
been found. Lets say
y(x) = c1 y1 (x) + c2 y2 (x) (8.6)
We then replace c1 and c2 by unknown functions v1 (x) and v2 (x) and attempt to determine v1 and
v2 in such a manner that
y = v1 y1 + v2 y2 (8.7)
will be a solution to
y 00 + p(x)y 0 (x) + q(x)y(x) = r(x). (8.8)
With two unknown functions (v1 and v2 ) to find, we need two equations relating these functions.
One of these is obtained by requiring that (8.7) be a solution to (8.8). We will later see what the
second equation should be. We begin by computing the derivative of (8.7) arranged as follows:

y 0 = (v1 y10 + v2 y20 ) + (v10 y1 + v20 y2 ) (8.9)

66
Another di↵erentiation will introduce second derivatives of unknowns v1 and v2 . We avoid the
second expression of (8.9) by letting
v10 y1 + v20 y2 = 0 (8.10)
This leads to
y 0 = v1 y10 + v2 y20 (8.11)
00
y = v1 y100 + v10 y10 + v2 y200 + v20 y20 (8.12)
Substituting (8.7), (8.11) and (8.12) into (8.8) and rearranging we get

v1 (y100 + py10 + qy1 ) + v2 (y200 + py20 + qy2 ) + v10 y10 + v20 y20 = r(x) (8.13)

Since y1 and y2 are solutions to the homogeneous equation, we have

v10 y10 + v20 y20 = r(x) (8.14)

Taking (8.10) and (8.14), we have two equations in 2 unknowns v10 and v20 i.e

v10 y1 + v20 y2 = 0
(8.15)
v10 y10 + v20 y20 = r(x)

✓ ◆ ✓ ◆✓ ◆
v10 1 y20 y2 0
=
v20 y1 y2 y10 y1 r(x)
y10 y20
This gives
y2 r(x) y1 r(x)
v10 = and v20 = (8.16)
W (y1 , y2 ) W (y1 , y2 )
The above formulae are legitimate, for the Wronskian is non-zero by the linear independence of y1
and y2 .
Next we integrate (8.16) to find v1 and v2
Z Z
y2 r(x) y1 r(x)
v1 = dx and v2 = dx
W W
Putting everything together we have
Z Z
y2 r(x) y1 r(x)
y = y1 dx + y2 dx
W W

is the particular solution to (8.8).

Example 8.2.4. Find the general solution to y 00 + y = tan x

Solution
The complimentary function is
yc = c1 cos x + c2 sin x

67
Set up yp = v1 cos x + v2 sin x

yp0 = v10 cos x v1 sin x + v20 sin x + v2 cos x

but v10 cos x + v20 sin x = 0
hence yp0 = v1 sin x + v2 cos x
yp00 = v10 sin x v1 cos x + v20 cos x v2 sin x
Substituting in y 00 + y = tan x we have

v10 sin x v1 cos x + v20 cos x v2 sin x + v1 cos x + v2 sin x = tan x

v10 sin x + v20 cos x = tan x

Thus we have ✓ ◆✓ ◆ ✓ ◆
cos x sin x v10 0
=
sin x cos x v20 tan x
cos x sin x
In this case the Wronskian W = =1
sin x cos x
2 cos2 x 1
v10 = sin x1 tan x = sin x
= = cos x sec x
0
cos x cos x (8.17)
v2 = cos x tan x = sin x

Integrating we have
v1 = sin x ln | sec x + tan x|
(8.18)
v2 = cos x
T husy(x) = c1 cos x + c2 sin x cos x ln | sec x + tan x|

Exercise 8.2.1. Solve

1. y 00 + 5y 0 + 6y = x2 + 2x
ex
2. y 00 2y 0 + y = x
, y(1) = 0, y 0 (1) = 0

68
Tutorial 8.2.1. Answer all questions

d2
1. Solve the equation dt2
A(t) = 6t + 8 with A(0) = 4 and A0 (0) = 9.
2
2. Suppose dtd 2 x(t) = 4x(t) + e t . What is the solution of the equation if the initial displacement
is 1 and the initial velocity is 0?

3. Find the general solution of the equation A00 (t) + A(t) = et .

4. Find the general solution of the equation A00 (t) + A(t) = sint
d2
5. Solve the equation dt2
x(t) + 9x(t) = 5 with initial conditions x(0) = 1 and x0 (0) = 0.
d2
6. Solve the equation dt2
x(t) + 9x(t) = t2 + t with initial conditions x(0) = 1 and x0 (0) = 0.
d
7. Solve the equation dt
x(t) = 4x(t) + t using the method of undetermined coefficients.
2
8. Solve dtd 2 x(t) = 4x(t) + sin2t. Hint: For the undetermined coefficients method try Atcos2t +
Btsin2t.

9. Using the method of variation of parameters, solve the following

ex
(a) y 00 + 5y 0 + 6y = x2 + 2x (b) y 00 2y 0 + y = x
, y(1) = 0, y 0 (1) = 1

10. Consider the following equation

2 0 2
y 00 y + 2 y = x sin x, x 2 [0, 1)
x x
Given that y1 (x) = x and y2 (x) = x2 are linearly independent solutions of the homogeneous
equation on (1, 1), find the particular solution using the method of variation of parameters.

69
Chapter 9

Series solutions

In this chapter we will be looking at non-constant coefficient di↵erential equations. While we won’t
cover all possibilities in this chapter we will be looking at two of the more common methods for
dealing with these kind of di↵erential equation.

The first method that we’ll be taking a look at, series solutions, will actually find a series repre-
sentation for the solution instead of the solution itself. You first saw something like this when you
looked at Taylor series in your Calculus class. As we will see however, these won’t work for every
di↵erential equation.

The second method that we’ll look at will only work for a special class of di↵erential equations.
This special case will cover some of the cases in which series solutions can’t be used.

9.1 Review: Power Series

Before looking at series solutions to a di↵erential equation we will first need to do a cursory review
of power series. A power series is a series in the form,
1
X
f (x) = an (x x0 ) n (9.1)
n=0

where, x0 and an are numbers. We can see from this that a power series is a function of x. The
function notation is not always included, but sometimes it is so we put it into the definition above.

Before proceeding with our review we should probably first recall just what series really are. Recall
that series are really just summations. One way to write our power series is then,
1
X
f (x) = an (x x0 ) n (9.2)
n=0
=a0 + a1 (x x0 ) + a2 (x x0 )2 + a3 (x x0 )3 + ...

70
Notice as well that if we needed to for some reason we could always write the power series as,
1
X
f (x) = an (x x0 ) n
n=0
=a0 + a1 (x x0 ) + a2 (x x0 )2 + a3 (x x0 )3 + ...
X1
=a0 + an (x x0 )n
n=1

All that we’re doing here is noticing that if we ignore the first term (corresponding to n = 0 ) the
remainder is just a series that starts at n = 1. When we do this we say that we’ve stripped out the
n = 0 , or first, term. We don’t need to stop at the first term either. If we strip out the first three
terms we’ll get,
1
X 1
X
an (x x0 )n = a0 + a1 (x x0 ) + a2 (x x0 ) + an (x x0 ) n
n=0 n=3

There are times when we’ll want to do this so make sure that you can do it.

Now, since power series are functions of x and we know that not every series will in fact exist,
it then makes sense to ask if a power series will exist for all x. This question is answered by looking
at the convergence of the power series. We say that a power series converges for x = c if the series,
1
X
an (c x0 ) n
n=0

converges. Recall that this series will converge if the limit of partial sums,
N
X
lim an (c x0 ) n
N !1
n=0

exists and is finite. In other words, a power series will converge for x=c if
1
X
an (c x0 ) n
n=0

is a finite number.

Note that a power series will always converge if x = x0 . In this case the power series will become
1
X
an (x x 0 ) n = a0
n=0

With this we now know that power series are guaranteed to exist for at least one value of x.

Convergence of power series is important because in order for a series solution to a di↵erential
equation to exist at a particular x it will need to be convergent at that x. If it’s not convergent at
a given x then the series solution won’t exist at that x. So, the convergence of power series is fairly
essential.

71
Next we need to do a quick review of some of the basics of manipulating series. We’ll start with
addition and subtraction.
There really isn’t a whole lot to addition and subtraction. All that we need to worry about is that
the two series start at the same place and both have the same exponent of the x x0 . If they do
then we can perform addition and/or subtraction as follows,
1
X 1
X 1
X
an (x x0 ) n ± bn (x x0 ) n = (an ± bn )(x x0 ) n
n=n0 n=n0 n=n0

In other words all we do is add or subtract the coefficients and we get the new series.
One of the rules that we’re going to have when we get around to finding series solutions to di↵erential
equations is that the only x that we want in a series is the x that sits in (x x0 )n .
This means that we will need to be able to deal with series of the form,
1
X
(x x0 ) c an (x x0 ) n
n=0

where c is some constant. These are actually quite easy to deal with.
1
X
c
(x x0 ) an (x x0 )n =(x x0 )c (a0 + a1 (x x0 ) + a2 (x x0 )2 + ...)
n=0
=a0 (x x0 )c + a1 (x x0 )1+c + a2 (x x0 )2+c + ...
1
X
= an (x x0 )n+c
n=0

So, all we need to do is to multiply the term in front into the series and add exponents. Also note
that in order to do this both the coefficient in front of the series and the term inside the series must
be in the form x x0 . If they are not the same we can’t do this, we will eventually see how to deal
with terms that aren’t in this form.

Next we need to talk about di↵erentiation of a power series. By looking at (3.2) it should be fairly
easy to see how we will di↵erentiate a power series. Since a series is just a giant summation all we
need to do is di↵erentiate the individual terms. The derivative of a power series will be,

f 0 (x) =a1 + 2a2 (x x0 ) + 3a3 (x x0 )2 + ...

1
X
= nan (x x0 ) n 1

n=1
1
X
= nan (x x0 ) n 1

n=0

So, all we need to do is just di↵erentiate the term inside the series and we’re done. Notice as well
that there are in fact two forms of the derivative. Since the n = 0 term of the derivative is zero it
won’t change the value of the series and so we can include it or not as we need to. In our work we
will usually want the derivative to start at n = 1, however there will be the occasion problem were
it would be more convenient to start it at n = 0.

72
Following how we found the first derivative it should make sense that the second derivative is,
1
X
f 00 (x) = n(n 1)an (x x0 ) n 2

n=2
1
X
= n(n 1)an (x x0 ) n 2

n=1
1
X
= n(n 1)an (x x0 ) n 2

n=0

In this case since the n = 0 and n = 1 terms are both zero we can start at any of three possible
starting points as determined by the problem that we’re working.

Next we need to talk about index shifts. As we will see eventually we are going to want our power
series written in terms of (x x0 ) and they often won’t, initially at least, be in that form. To get
them into the form we need we will need to perform an index shift.
Index shifts themselves really aren’t concerned with the exponent on the x term, they instead are
concerned with where the series starts as the following example shows.
Example 9.1.1. Write the following as a series that starts at n=0 instead of n=3.
1
X
n2 an 1 (x + 4)n+2
n=3

Solution
An index shift is a fairly simple manipulation to perform. First we will notice that if we define
i = n 3 then when n=3 we will have i=0. So what we?ll do is rewrite the series in terms of i
instead of n. We can do this by noting that n = i + 3. So, everywhere we see an n in the actual
series term we will replace it with an i + 3. Doing this gives,
1
X 1
X
n2 an 1 (x + 4)n+2 = (i + 3)2 ai+3 1 (x + 4)i+3+2
n=3 i=0
1
X
= (i + 3)2 ai+2 (x + 4)i+5
i=0

The upper limit won’t change in this process since infinity minus three is still infinity.
The final step is to realize that the letter we use for the index doesn’t matter and so we can just
switch back to n’s. 1 1
X X
2 n+2
n an 1 (x + 4) = (n 2)2 an 3 (x + 4)n
n=3 n=5
Now, we usually don’t go through this process to do an index shift. All we do is notice that we
dropped the starting point in the series by 3 and everywhere else we saw an n in the series we
increased it by 3. In other words, all the n’s in the series move in the opposite direction that we
moved the starting point.
Example 9.1.2. Write the following as a series that starts at n = 5 instead of n = 3.
1
X
n2 an 1 (x + 4)n+2
n=3

73
Solution
To start the series to start at n = 5 all we need to do is notice that this means we will increase the
starting point by 2 and so all the other n’s will need to decrease by 2. Doing this for the series in
the previous example would give,
1
X 1
X
n2 an 1 (x + 4)n+2 = (n 2)2 an 3 (x + 4)n
n=3 n=5

Now, as we noted when we started this discussion about index shift the whole point is to get our
series into terms of (x x0 )n . We can see in the previous example that we did exactly that with an
index shift. The original exponent on the (x + 4) was n + 2. To get this down to an n we needed to
decrease the exponent by 2. This can be done with an index that increases the starting point by 2.
Exercise 9.1.1. Write each of the following as a single series in terms of (x x0 ) n .

P P1
1. (x + 2)2 1 n=3 nan (x + 2)
n 4
n=1 nan (x + 2)
n+1

P
2. x 1 n=0 (n 5)2 bn+1 (x 3)n+3

There is one final fact that we need take care of before moving on. Before giving this fact for power
series let’s notice that the only way for

a + bx + cx2 = 0

to be zero for all x is to have a = b = c = 0.

We’ve got a similar fact for power series.

FACT
If 1
X
an (x x0 ) n = 0
n=0

for all x, then

an = 0, n = 0, 1, 2, 3, ...
This fact will be key to our work with di↵erential equations so don’t forget it.

Review: Taylor Series

We are not going to be doing a whole lot with Taylor series once we get out of the review, but
they are a nice way to get us back into the swing of dealing with power series. Remembering how
Taylor series work will be a very convenient way to get comfortable with power series before we
start looking at di↵erential equations.

Taylor Series
If f (x) is an infinitely di↵erential function then the Taylor Series of f (x) about x = x0 is,
1
X f (n) (x0 )
f (x) = (x x0 ) n
n=0
n!

74
Recall that
f (0) (x) = f (x) f (n) (x) = nth derivative of f (x)

Definition 9.1.1. A function, f (x), is called analytic at x = a if the Taylor series for f (x) about
x = a has a positive radius of convergence and converges to f (x).

We need to give one final note before proceeding into the next section. We started this section out
by saying that we weren’t going to be doing much with Taylor series after this section. While that
is correct it is only correct because we are going to be keeping the problems fairly simple. For more
complicated problems we would also be using quite a few Taylor series.

9.1.1 Series Solutions to Di↵erential Equations

Before we get into finding series solutions to di↵erential equations we need to determine when we
can find series solutions to di↵erential equations. So, let’s start with the di↵erential equation,

p(x)y 00 + q(x)y 0 + r(x)y = 0 (9.3)

This time we really do mean non constant coefficients. To this point we’ve only dealt with constant
coefficients. However, with series solutions we can now have non constant coefficient di↵erential
equations. Also, in order to make the problems a little nicer we will be dealing only with polynomial
coefficients.
Now, we say that x = x0 is an ordinary point if provided both

q(x) r(x)
and
p(x) p(x)

are analytic at x = x0 . That is to say that these two quantities have Taylor series around x = x0 .
We are going to be only dealing with coefficients that are polynomials so this will be equivalent to
saying that
p(x0 ) 6= 0
for most of the problems.
If a point is not an ordinary point we call it a singular point.
The basic idea to finding a series solution to a di↵erential equation is to assume that we can write
the solution as a power series in the form,
1
X
y(x) = an (x x0 ) n (9.4)
n=0

and then try to determine what the an’s need to be. We will only be able to do this if the point
x = x0 , is an ordinary point. We will usually say that (9.4) is a series solution around x = x0 .
Let’s start with a very basic example of this. In fact it will be so basic that we will have constant
coefficients. This will allow us to check that we get the correct solution.

75
Example 9.1.3. Determine a series solution for the following di↵erential equation about x0 = 0.

y 00 + y = 0

Solution
Notice that in this case p(x) = 1 and so every point is an ordinary point. We will be looking for a
solution in the form,
X1
y(x) = an x n
n=0

We will need to plug this into our di↵erential equation so we’ll need to find a couple of derivatives.
1
X 1
X
0 n 1 00
y (x) = nan x y (x) = n(n 1)an xn 2

n=1 n=2

Recall from the power series review section on power series that we can start these at n = 0 if we
need to, however it’s almost always best to start them where we have here. If it turns out that it
would have been easier to start them at n = 0 we can easily fix that up when the time comes around.
So, plug these into our di↵erential equation. Doing this gives,
1
X 1
X
n 2
n(n 1)an x + an x n = 0
n=2 n=0

The next step is to combine everything into a single series. To do this requires that we get both
series starting at the same point and that the exponent on the x be the same in both series.
We will always start this by getting the exponent on the x to be the same. It is usually best to get
the exponent to be an n. The second series already has the proper exponent and the first series will
need to be shifted down by 2 in order to get the exponent up to an n. If you don’t recall how to do
this take a quick look at the first review section where we did several of these types of problems.
Shifting the first power series gives us,
1
X 1
X
(n + 2)(n + 1)an+2 xn + an x n = 0
n=0 n=0

Notice that in the process of the shift we also got both series starting at the same place. This won’t
always happen, but when it does we’ll take it. We can now add up the two series. This gives,
1
X
[(n + 2)(n + 1)an+2 + an ]xn = 0
n=0

Now recalling the fact from the power series review section we know that if we have a power series
that is zero for all x (as this is) then all the coefficients must have been zero to start with. This
gives us the following,

(n + 2)(n + 1)an+2 + an = 0, n = 0, 1, 2, ...

This is called the recurrence relation and notice that we included the values of n for which it must
be true. We will always want to include the values of n for which the recurrence relation is true

76
since they won’t always start at n = 0 as it did in this case.
Now let’s recall what we were after in the first place. We wanted to find a series solution to the
di↵erential equation. In order to do this we needed to determine the values of the an’s. We are
almost to the point where we can do that. The recurrence relation has two di↵erent an’s in it so we
can’t just solve this for an and get a formula that will work for all n. We can however, use this to
determine what all but two of the an’s are.
To do this we first solve the recurrence relation for the an that has the largest subscript. Doing this
gives,
an
an+2 = n = 0, 1, 2, ...
(n + 2)(n + 1)
Now, at this point we just need to start plugging in some value of n and see what happens,
a0 a1
n =0 a2 = , n=1 a3 =
(2)(1) (3)(2)
a0 a1
n =2 a4 = , n = 3 a5 =
(4)(3)(2)(1) (5)(4)(3)(2)
a0 a1
n =4 a6 = , n = 5 a7 =
(6)(5)(4)(3)(2)(1) (7)(6)(5)(4)(3)(2)
. .
. .
. .
( 1)k a0 ( 1)k a1
a2k = , k = 1, 2, ... a2k+1 = , k = 1, 2, ...
(2k)! (2k + 1)!

Notice that at each step we always plugged back in the previous answer so that when the subscript
was even we could always write the an in terms of a0 and when the coefficient was odd we could
always write the an in terms of a1 . Also notice that, in this case, we were able to find a general
formula for an’s with even coefficients and an’s with odd coefficients. This won’t always be possible
to do.
There’s one more thing to notice here. The formulas that we developed were only for k = 1, 2, ?
however, in this case again, the will also work for k=0. Again, this is something that won’t always
work, but does here.
Do not get excited about the fact that we don’t know what a0 and a1 are. As you will see, we actually
need these to be in the problem to get the correct solution.
Now that we’ve got formulas for the an’s let’s get a solution. The first thing that we’ll do is write
out the solution with a couple of the an’s plugged in.
1
X
y(x) = an x n (9.5)
n=0
=a0 + a1 x + a2 x2 + a3 x3 + ... + a2k x2k + a2k+1 x2k+1 + ...
a0 2 a1 3 ( 1)k a0 2k ( 1)k+1 a1 2k+1
=a0 + a1 x x x + ... + x + x + ...
2! 3! (2k)! (2k + 1)!

The next step is to collect all the terms with the same coefficient in them and then factor out that

77
coefficient.

x2 ( 1)k a0 2k x3 ( 1)k+1 a1 2k+1

y(x) =a0 {1 ... + x } + a1 {x = + ... + x + ...} (9.6)
2! (2k)! 3! (2k + 1)!
X1
( 1)k a0 2k x3 ( 1)k+1 a1 2k+1
=a0 x } + a1 {x = + a1 x
n=0
(2k)! 3! (2k + 1)!

Before working another problem let’s take a look at the solution to the previous example. First, we
started out by saying that we wanted a series solution of the form,
1
X
y(x) = an x n
n=0

and we didn’t get that. We got a solution that contained two di↵erent power series. Also, each
of the solutions had an unknown constant in them. This is not a problem. In fact, it’s what we
want to have happen. From our work with second order constant coefficient di↵erential equations
we know that the solution to the di↵erential equation in the last example is,

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

Solutions to second order di↵erential equations consist of two separate functions each with an
unknown constant in front of them that are found by applying any initial conditions. So, the form
of our solution in the last example is exactly what we want to get. Also recall that the following
Taylor series,
X1 X1
( 1)n x2n ( 1)n x2n+1
cos(x) = sin(x) =
n=0
(2n)! n=0
(2n + 1)!
Recalling these we very quickly see that what we got from the series solution method was exactly
the solution we got from first principles, with the exception that the functions were the Taylor series
for the actual functions instead of the actual functions themselves.
Now let’s work an example with nonconstant coefficients since that is where series solutions are
most useful.

78
Tutorial 9.1.1. Attempt all questions

1. Find a series solution around x0 = 0 for the following di↵erential equation.

y 00 xy = 0

2. Find the first four terms in each portion of the series solution around x0 = 2 for the following
di↵erential equation.
y 00 xy = 0

3. Find the first four terms in each portion of the series solution around x0 = 0 for the following
di↵erential equation.
(x2 + 1)y 00 4xy 0 + 6x = 0

79
9.2 Euler Equations

In this section we want to look for solutions to

00
ax2 y + bxy 0 + cy = 0 (9.7)

around x0 = 0. These type of di↵erential equations are called Euler Equations.

Recall from the previous section that a point is an ordinary point if the quotients,
bx b c
2
= and
ax ax ax2
have Taylor series around x0 = 0. However, because of the x in the denominator neither of these
will have a Taylor series around x0 = 0 and so x0 = 0 is a singular point. So, the method from the
previous section won’t work since it required an ordinary point.
However, it is possible to get solutions to this di↵erential equation that aren’t series solutions.
Let’s start o↵ by assuming that x > 0 (the reason for this will be apparent after we work the first
example) and that all solutions are of the form,

y(x) = xr (9.8)

Now plug this into the di↵erential equation to get,

ax2 (r)(r 1)xr 2 + bx(r)xr 1 + cxr =0

ar(r 1)xr + b(r)xr + cxr =0
(ar(r 1) + b(r) + c)xr =0

Now, we assumed that x > 0 and so this will only be zero if,

ar(r 1) + b(r) + c = 0 (9.9)

So solutions will be of the form (3.8) provided r is a solution to (3.9). This equation is a quadratic
in r and so we will have three cases to look at : Real Distinct Roots, Double Roots, and Complex
Roots

Real Distinct Roots

There really isn’t a whole lot to do in this case. We’ll get two solutions that will form a fundamental
set of solutions and so our general solution will be,

y(x) = c1 xr1 + c2 xr2

Example 9.2.1. Solve the following di↵erential equation

00 0
2x2 y + 3xy 15y = 0 (9.10)

80
Solution
We first need to find the roots to (3.9).

2r(r 1) + 3r 15 =0
5
2r2 + r 15 =(2r 5)(r + 3) = 0 ) r1 = , r2 = 3
2
The general solution is then,
5
3
y(x) = c1 x 2 + c2 x

Double Roots

This case will lead to the same problem that we’ve had every other time we’ve run into double roots
(or double eigenvalues). We only get a single solution and will need a second solution. In this case
it can be shown that the second solution will be,

y2 (x) = xr ln x

and so the general solution in this case is,

y(x) = c1 xr + c2 xr ln x = xr (c1 + c2 ln x)

Example 9.2.2. Find the general solution to the following di↵erential equation.
00 0
x2 y 7xy + 16y = 0

Solution
We can see that the roots are

r(r 1) 7r + 16 =0
(r 4)2 =0 ) r=4

So the general solution is then,

y(x) = c1 x4 + c2 x4 ln x

Complex Roots

In this case we’ll be assuming that our roots are of the form,

r1,2 = ± µi

If we take the first root we’ll get the following solution.

+µi
x

81
This is a problem since we don’t want complex solutions, we only want real solutions. We can
eliminate this by recalling that,
r
xr = eln x = er ln x
Plugging the root into this gives,
+µi
x =e( +µi) ln x

ln x µi ln x
=e e
ln x
=e (cos(µ ln x) + i sin(µ ln x))
=x cos(µ ln x) + ix sin(µ ln x)
Note that we had to use Euler formula as well to get to the final step. Now, as we’ve done every
other time we’ve seen solution like this we can take the real part and the imaginary part and use
those for our two solutions.

So, in the case of complex roots the general solution will be,
y(x) = c1 x cos(µ ln x) + c2 x sin(µ ln x) = x (c1 cos(µ ln x) + c2 sin(µ ln x)
Example 9.2.3. Find the solution to the following di↵erential equation.
00 0
x2 y + 3xy + 4y = 0
Solution
The roots are
r(r 1) + 3r + 4 =0
p
r2 + 2r + 4 =0 ) r1,2 = 1± 3i
The general solution is then,
1
p 1
p
y(x) = c1 x cos( 3 ln x) + c2 x sin( 3 ln x)

9.3 Method of Frobenius

To solve di↵erential equations of the form

00 0
x2 y + xp(x)y + q(x)y = 0 (9.11)
about a regular singular point where p(x) and q(x) are low-degree polynomials, we employ the
following theorem due to Frobenius.
Theorem 9.3.1. If x = x0 is a regular singular point of equation (3.11), then there exists at least
one solution of the form
1
X 1
X
y = (x x0 ) r cn (x x0 ) n = cn (x x0 )n+r , (9.12)
n=0 n=0

82
where the number c0 6= 0 and r is a constant to be determined. Note that p(x) and q(x) are analytic
at the origin, and therefore each has a convergent Taylor series there.

Observe that if p(x) and q(x) are constant polynomials, then (3.11) reduces to a Euler equation.

To evaluate r and an in Theorem (3), one proceeds as in the power series method describe in the
previous section. It follows that
1
X
y 0 (x) = (n + r)cn xn+r 1

n=0
X1
y 00 (x) = (n + r)(n + r 1)cn xn+r 2

n=0

Furthermore, we suppose that p(x) and q(x) have the expansions

p(x) =p0 + p1 x + p2 x2 + ... + pn xn + ...

q(x) =q0 + q1 x + q2 x2 + ... + qn xn + ...
0 00
Substituting these expressions for y, y , y , p(x), and q(x) in (3.11) and gathering like terms, we
find that
00 0
0 =x2 y + xp(x)y + q(x)y
1
X
= (n + r)(n + r 1)cn xn+r + (p0 + p1 x + p2 x2 + ... + pn xn + ...)
n=0
1
X 1
X
⇥ (n + r)cn xn+r + (q0 + q1 x + q2 x2 + ... + qn xn + ...) cn xn+r
n=0 n=0
=(r(r 1) + p0 r + q0 )c0 + b1 x + b2 x2 + ... (9.13)

where the general term bn depends on n and all earlier coefficients for each n 1. The most
important conclusion to draw from (3.13) comes from the fact that each coefficient of the general
power series expansion must be zero, so that since c0 6= 0,

r(r 1) + p0 r + q0 = 0 (9.14)

Equation (3.14) is called the indicial equation for the Method of Frobenius. Note that this equation
is quadratic in r; its two roots are the values of r that are used in (3.12)

9.3.1 Cases of Indicial Roots

When using the method of Frobenius, we usually distinguish three cases corresponding to the nature
of the indicial roots.

83
CASE 1 Roots Not Di↵ering By an Integer

If r1 and r2 are distinct and do not di↵er by an integer, then there exist two linearly independent
solutions of equation (3.11) of the form
1
X
y1 = cn xn+r1 , c0 6= 0
n=0
1
X
y2 = bn xn+r2 , b0 6= 0
n=0

When the roots of the indicial equation di↵er by a positive integer, we may or may not be able to
find two solutions. If not, then one solution corresponding to the smaller root contains a logarithmic
term. When the indicial roots are equal, a second solution always contains a logarithm.

CASE 2 Roots di↵ering by a Positive Integer

If r1 r2 = N , where N is a positive integer, then there exist two linearly independent solutions of
(3.11) of the form
1
X
y1 = cn xn+r1 , c0 6= 0
n=0
1
X
y2 = Cy1 (x) ln x + bn xn+r2 , b0 6= 0,
n=0

where C is a constant that could be zero.

CASE 3 Equal Indicial Roots

If r1 = r2 , there always exist two linearly independent solutions of (3.11) of the form
1
X
y1 = cn xn+r1 , c0 6= 0
n=0
1
X
y2 = y1 (x) ln x + bn xn+r1 .
n=1

Example 9.3.1. Find a Frobenius series solution for the Bessel-Cli↵ord equation
00 0
x2 y + (1 a)xy + xy = 0 (9.15)

where a is a constant.
Solution.

84
With a being a constant, we have p(x) = 1 a, so in the series expansion for p, p0 = 1 a.
Moreover, q(x) = x , so q0 = 0. Thus, for the given DE the indicial equation is

r(r 1) + (1 a)r = 0

Rearranging, we see that r(r 1 + 1 a) = r(r a) = 0, and thus the roots of the indicial equation
are r = 0 and r = a.
In the case that r = 0, the Method of Frobenius is providing an analytic solution to (3.15) of the
form
X1
y1 = cn xn
n=0

Dividing both sides of (3.15) by x and substituting this expression for y using the standard series
methods we have done before, it follows that
1
X
y1 = [(n + 1)(n + 1 a)]cn+1 + cn ]xn
n=0

from which we obtain the recurrence relation

1
cn+1 = cn (9.16)
(n + 1)(n + 1 a)

It follows from (3.16) that the closed form expression for cn is

( 1)k
cn+1 = c0 ,n 1
k!(1 a)(2 a)...(n a)
so we find that !
1
X ( 1)k
y1 (x) = c0 1+ xn (9.17)
n=1
k!(1 a)(2 a)...(n a)
which is valid for all x provided that a 6= 1, 2, ..... Note that from this recurrence relation, every cn
is a function of c0 , and thus there cannot be two linearly independent solutions to the Bessel-Cli↵ord
equation that are analytic at 0. Indeed, every solution linearly independent of y1 (x) must be singular
at 0. And while the equation has a singular point at the origin, there is an analytic solution there
for every a except when a is a positive integer.

Exercise Find the second solution to the Bessel-Cli↵ord equation using r = a.

Note: The method by Frobenius shows that a certain class of linear second-order DE’s with a
singular point at the origin can be represented in series form by a slight generalization of a Taylor
series. Thus we can employ Frobenius method to solve other forms of DE’s that have singular point
at the origin.

Example 9.3.2. Use Frobenius method to solve the following di↵erential equation

3xy 00 + y 0 y=0

85
Solution
Since x = 0 is a regular singular point of the di↵erential equation, we try a solution of the form
X1
y= cn xn+r . Now
n=0

1
X 1
X
y0 = (n + r)cn xn+r 1
and y 00 = (n + r)(n + r 1)cn xn+r 2 ,
n=0 n=0

so that
1
X 1
X 1
X
3xy 00 + y 0 y = 3 (n + r)(n + r 1)cn xn+r 1
+ (n + r)cn xn+r 1
cn xn+r
n=0 n=0 n=0
1
X 1
X
= (n + r)(3n + 3r 2)cn xn+r 1
cn xn+r
n=0 n=0
2 3
6 1
X 1
X 7
6 r 1 n 1 7
= x 6r(3r 2)c0 x + (n + r)(3n + 3r 2)cn x cn xn 7
4 5
|n=1 {z } |n=0{z }
k=n-1 k=n
" 1
#
X
= xr r(3r 2)c0 x 1
+ [(k + r + 1)(3k + 3r + 1)ck+1 ck ]xk = 0,
k=0

which implies

r(3r 2)c0 = 0
(k + r + 1)(3k + 3r + 1)ck+1 ck = 0, k = 0, 1, 2, . . .

Since nothing is gained by taking c0 = 0, we must then have

r(3r 2) = 0

and
ck
ck+1 = , k = 0, 1, 2, . . .
(k + r + 1)(3k + 3r + 1)
2
The two values of r satisfying r(3r 2) = 0, r1 = and r2 = 0, when substituted into the recurrence
3
equation, gives two di↵erent recurrence relations,
2 ck
r1 = , ck+1 = , k = 0, 1, 2, . . . (9.18)
3 (3k + 5)(k + 1)

and
ck
r2 = 0, ck+1 = , k = 0, 1, 2, . . . (9.19)
(k + 1)(3k + 1)

86
Iteration of (9.18), gives
c0
c1 =
5·1
c1 c0
c2 = =
8·2 2!5 · 8
c2 c0
c3 = =
11 · 3 3!5 · 8 · 11
c3 c0
c4 = =
14 · 4 4!5 · 8 · 11 · 14
..
.
c0
cn = , n = 1, 2, 3, . . .
n!5 · 8 · 11 · · · (3n + 2)

whereas iteration of (9.19) yields

c0
c1 =
1·1
c1 c0
c2 = =
2·4 2!1 · 4
c2 c0
c3 = =
3·7 3!1 · 4 · 7
c3 c0
c4 = =
4 · 10 4!1 · 4 · 7 · 10
..
.
c0
cn = , n = 1, 2, 3, . . .
n!1 · 4 · 7 · · · (3n 2)

Thus we obtain two series solutions

" 1
#
2
X 1
y1 = c0 x 3 1+ xn
n=1
n!5 · 8 · 11 · · · (3n + 2)

and " #
1
X 1
y2 = c0 x0 1 + xn .
n=1
n!1 · 4 · 7 · · · (3n 2)
Hence by the superposition principle
" 1
# " 1
#
2
X 1 2
X 1
n+ 3
y = C1 y1 (x)+C2 y2 (x) = C1 x +
3 x +C2 1 + xn .
n=1
n!5 · 8 · 11 · · · (3n + 2) n=1
n!1 · 4 · 7 · · · (3n 2)

87
Tutorial 9.3.1. Attempt all questions.

1. Use Euler method and Frobenius method to solve the following di↵erential equation and com-
pare your solutions.
00 0
x2 y 4xy + 6y = 0.

2. Determine which points are ordinary points and which points are singular points for the given
di↵erential equation.
x+1 cos x
(i) y 00 + ex y 0 + (2x2 1)y = 0 (ii) y 00 + y0 + 2 y = 0
(x + 2)(x 3) x
2 00 0 2
(iii) (x + 1) (2x 1)y 2xy + (x 1)y = 0.

3. For each di↵erential equation find two linearly independent power series solutions about the
ordinary point x = 0.
(i) y 00 xy = 0 (ii) y 00 2xy 0 + y = 0 (iii) (x2 1)y 00 + 4xy 0 + 2y = 0
(iv) y 00 (x + 1)y 0 y = 0 (v) y 00 xy 0 (x + 2)y = 0 (vi) (x2 + 1)y 00 6y = 0.

4. Show that the indicial roots do not di↵er by an integer. Use the method of Frobenius to obtain
two linearly independent series solutions about the regular singular point x0 = 0.
(i) 2xy 00 y 0 + 2y = 0 (ii) 3xy 00 + (2 x)y 0 y = 0 (iii) 9x2 y 00 + 9x2 y 0 + 2y = 0
(iv) x(x 2)y 00 + y 0 2y = 0 (v) x2 y 00 (x 29 )y = 0 (vi) 2xy 00 + 5y 0 + xy = 0.

5. The regular singular points of the hyper geometric equation

x(1 x)y 00 + [ (↵ + + 1)x]y 0 ↵ y = 0,

↵, , real constants are 0 and 1.

(a) Show that the indicial roots are 0 and 1 .

(b) Show that a solution corresponding to the indicial root 0 is given by

↵ ↵(↵ + 1) ( + 1) 2 ↵(↵ + 1)(↵ + 2) ( + 1)( + 2) 3

y1 (x) = 1 + x+ x + x + ···
1! 2! ( + 1) 3! ( + 1)( + 2)

provided is not zero or a negative integer. This solution is known as the hypergeo-
metric series.

6. Verify that the given equation has a regular singular point at x = 0, and find all solutions of
the Frobenius type.
(i) x2 y 00 + x(1 + x)y 0 y = 0 (ii) xy 00 + (3 + x2 )y 0 + 2xy = 0 (iii) (x + x2 )y 00 2y 0 2y = 0.

88