E. Rukmangadachari - Engineering Mathematics - Volume - 1-Pearson Education (2009)

Engineering
Mathematics-I
Engineering
Mathematics-I
E. Rukmangadachari
Professor of Mathematics,
Department of Humanities and Sciences,
Malla Reddy Engineering College,
Secunderabad
Chennai • Delhi • Chandigarh

Copyright © 2010 Dorling Kindersley (India) Pvt. Ltd.
This book is sold subject to the condition that it shall not, by way of trade or otherwise, be lent, resold,
hired out, or otherwise circulated without the publisher’s prior written consent in any form of binding
or cover other than that in which it is published and without a similar condition including this condition
being imposed on the subsequent purchaser and without limiting the rights under copyright reserved
above, no part of this publication may be reproduced, stored in or introduced into a retrieval system, or
transmitted in any form or by any means (electronic, mechanical, photocopying, recording or otherwise),
without the prior written permission of both the copyright owner and the above-mentioned publisher
of this book.
ISBN 978-81-317-2941-0
First Impression
Published by Dorling Kindersley (India) Pvt. Ltd., licensees of Pearson Education in South Asia.
Head Office: 7th Floor, Knowledge Boulevard, A-8(A), Sector-62, Noida, India.
Registered Office: 14 Local Shopping Centre, Panchsheel Park, New Delhi 110 017, India.
Laser typeset by MacroTex Solutions, Chennai.
Printed in India at Chennai Micro Prints Pvt. Ltd., Chennai.

To
My Beloved Parents
Enikapati Krishnamachari,
Rajamma, Ademma
About the Author
E. Rukmangadachari is former head of Computer Science and Engineering
as well as Humanities and Sciences at Malla Reddy Engineering College,
Secunderabad. Earlier, he was a reader in Mathematics (PG course)
at Government College, Rajahmundry. He is an M.A. from Osmania
University, Hyderabad, and an M.Phil. and Ph.D. degree holder from Sri
Venkateswara University, Tirupathi.
A recipient of the Andhra Pradesh State Meritorious Teachers’ Award
in 1981, Professor Rukmangadachari has published over 40 research papers
in national and international journals. With a rich repertoire of over 45
years’ experience in teaching mathematics to undergraduate, postgraduate
and engineering students, he is currently the vice-president of the Andhra
Pradesh Society for Mathematical Sciences. An ace planner with fine managerial skills, he was the
organising secretary for the conduct of the 17th Congress of the Andhra Pradesh Society for Mathematical
Sciences, Hyderabad.
vi
Road Map to the Syllabus
JAWAHARLAL NEHRU TECHNOLOGICAL UNIVERSITY
HYDERABAD
UNIT I SEQUENCES–SERIES
Basic Definitions of Sequences and Series – Convergences and Divergence – Ratio Test – Comparison
Test – Integral Test – Cauchy’s Root Test – Raabe’s Test – Absolute and Conditional Convergence
( REFER
CHAPTER 9
UNIT II FUNCTIONS OF SINGLE VARIABLE
Rolle’s Theorem – Lagrange’s Mean Value Theorem – Cauchy’s Mean Value Theorem – Generalized
Mean Value Theorem (all Theorems Without Proof) Functions of Several Variables – Functional
Dependence – Jacobian-Maxima and Minima of Functions of Two Variables With Constraints and
Without Constraints
( REFER
CHAPTERS 3 To 4
UNIT III APPLICATION OF SINGLE VARIABLES
Radius, – Centre and Circle of Curvature – Evolutes and Envelopes Curve Tracing – Cartesian, Polar
and Parametric Curves
( REFER
CHAPTERS 5 To 6
vii
UNIT IV INTEGRATION & ITS APPLICATIONS
Riemann Sums, Integral Representation for Lengths, Areas, Volumes and Surface Areas in Cartesian
and Polar Coordinates – Multiple Integrals – Double and Triple Integrals – Change of Order of
Integration – Change of Variable
( REFER
CHAPTERS 7 To 8
UNIT V DIFFERENTIAL EQUATIONS OF FIRST ORDER AND THEIR APPLICATIONS
Overview of Differential Equations – Exact, Linear and Bernoulli – Applications to Newton’s Law of
Cooling – Law of Natural Growth and Decay – Orthogonal Trajectories and Geometrical Applications
( REFER
CHAPTER 1
UNIT VI HIGHER ORDER LINEAR DIFFERENTIAL EQUATIONS AND THEIR APPLICATIONS
Linear Differential Equations of Second and Higher Order With Constant Coefficients – RHS Term
of the Type f(X) = eax, sin ax, cos ax, and xn, eax V(x), xn V(x), Method of Variation of Parameters
– Applications Bending of Beams – Electrical Circuits, Simple Harmonic Motion
( REFER
CHAPTER 2
UNIT VII LAPLACE TRANSFORM AND ITS APPLICATIONS TO ORDINARY DIFFERENTIAL

EQUATIONS
Laplace Transform of Standard Functions – Inverse Transform – First Shifting Theorem, Transforms
of Derivatives and Integrals – Unit Step Function – Second Shifting Theorem – Dirac’s Delta Function
– Convolution Theorem – Periodic Function – Differentiation and Integration of Transforms –
Application of Laplace Transforms to Ordinary Differential Equations
( REFER
CHAPTER 12
UNIT VIII VECTOR CALCULUS
Vector Calculus: Gradient–Divergence–Curl and Their Related Properties – Potential Function – Laplacian and
Second Order Operators – Line Integral – Work Done – Surface Integrals – Flux of a Vector Valued Function –
Vector Integrals Theorems: Green’s–Stoke’s and Gauss’s Divergence Theorems (Statement & Their Verification)
( REFER
CHAPTERS 10 To 11
Contents
About the Author vi Exercise 2.5 2-21
Road Map to the Syllabus vii Exercise 2.6 2-24
Preface xiii Exercise 2.7 2-26
Exercise 2.8 2-28
1. Ordinary Differential Equations Exercise 2.9 2-30
1.1 Introduction 1-1 Exercise 2.10 2-35
1.2 Differential Equations 1-2 Exercise 2.11 2-37
1.3 First Order and First Degree Exercise 2.12 2-40
Differential Equations 1-4 Exercise 2.13 2-43
Exercise 1.1 1-6 2.3 Application of Linear
Differential Equations—
Exercise 1.2 1-8
Mechanical and Electrical
Exercise 1.3 1-12 Oscillatory Circuits and
Exercise 1.4 1-17 Deflection of Beams 2-44
Exercise 1.5 1-19 Exercise 2.14 2-57
Exercise 1.6 1-23
Exercise 1.7 1-28 3. Rolle’s Theorem and Mean
Exercise 1.8 1-30 Value Theorems
1.4 Applications of Ordinary
Differential Equations 1-30 3.1 Introduction 3-1
Exercise 3.2 3-8
Exercise 1.10 1-36
Exercise 3.3 3-14
2. Linear Differential Equations of 4. Functions of Several Variables

Second and Higher Order
4.1 Introduction 4-1
2.1 Introduction 2-1 4.2 Functions of Several
Exercise 2.1 2-7 Variables 4-1
Exercise 2.2 2-11 4.3 Function of Two Variables 4-1
Exercise 2.3 2-13 4.4 Partial Differential
2.2 General Solution of Linear Coefficient 4-2
Equation f(D)y = Q(x) 2-14 Exercise 4.1 4-4
Exercise 2.4 2-17 4.5 Total Differential Coefficient 4-5
ix
x Contents
4.6 Chain Rules for Partial Exercise 5.5 5-20

Differentiation 4-5
Exercise 4.2 4-7 6. Curve-Tracing
4.7 Homogeneous Functions 4-8
6.1 Introduction 6-1
4.8 Euler’s Theorem on
Homogeneous Functions 4-9 6.2 Curve-Tracing—Cartesian
Equations 6-1
4.9 Extension of Euler’s
Theorem 4-10 6.3 Tracing of Curves—Polar
Coordinates 6-12
Exercise 4.3 4-13
6.4 Tracing of Curves—
4.10 Jacobian 4-13
Parametric Coordinates 6-20
4.11 Standard Jacobians 4-15
Exercise 6-22
Exercise 4.4 4-19
6.5 Asymptotes 6-24
4.12 Maxima and Minima 4-19
Exercise 4.5 4-22
Exercise 4.6 4-26 7. Applications of Integration
4.13 Constrained Maxima 7.1 Introduction 7-1
and Minima 4-27 7.2 Antiderivative 7-1
Exercise 4.7 4-29 7.3 Length of a Curve 7-3
Exercise 7.1 7-9
7.4 Volumes of Solids of
5. Radius of Curvature Revolution 7-10
5.1 Introduction 5-1 Exercise 7.2 7-17
5.2 Curvature of a Circle 5-1 7.5 Surface Area of a Solid
Exercise 5.1 5-3 of Revolution 7-18
5.3 Radius of Curvature— Exercise 7.3 7-23
Cartesian Formula 5-3
5.4 Radius of Curvature—
Parametric Form 5-5 8. Multiple Integrals
5.5 Radius of Curvature— 8.1 Multiple Integrals 8-1
Polar Form 5-7 Exercise 8.1 8-7
5.6 Pedal Equation or p–r 8.2 Change of the Order of
Equation of a Curve 5-9 Integration 8-8
5.7 Radius of Curvature at 8.3 Transformation of
the Origin 5-11 Coordinates 8-11
5.8 Centre of Curvature 5-13 8.4 Double Integral—
Exercise 5.4 5-14 Volume Enclosed by a
5.9 Envelopes 5-15 Cylindrical Surface 8-15
Contents xi
8.5 Multiple Integrals— 9.20 Cauchy’s nth Root Test 9-20

Triple Integrals 8-17 Exercise 9.4 9-21
8.6 Applications of 9.21 D’Alembert’s Ratio Test 9-22
Integration— Exercise 9.5 9-25
Quadrature 8-21
9.22 Cauchy—Maclaurin’s
Exercise 8.4 8-24 Integral Test 9-26
Exercise 8.5 8-30 9.23 Cauchy’s Condensation Test 9-27
8.7 Physical Applications 8-31 Exercise 9.6 9-28
Exercise 8.6 8-36
9.24 Raabe’s Test (Higher
Ratio Test) 9-29
9. Sequences and Series 9.25 Logarithmic Test 9-31
9.1 Introduction 9-1 9.26 De Morgan—Bertrand Test 9-33
9.2 Real Sequence 9-1 Exercise 9.7 9-34
9.3 Bounded Sequences 9-2 9.27 Alternating Series 9-35
9.4 Convergent Sequences 9-3 Exercise 9.8 9-37
9.5 Divergent Sequences 9-4
9.6 Monotonic Sequences 9-5
10. Vector Differential Calculus
9.7 Cauchy Sequences 9-6
9.8 Cauchy’s General Principle 10.1 Vector Differential Calculus 10-1
of Convergence (or Cauchy’s Exercise 10.1 10-6
Convergence Criterion 9-7 10.2 Differential Operators 10-6
Exercise 9.1 9-9 10.3 Scalar and Vector
9.9 Infinite Series 9-10 Differential Operators 10-9
9.10 Convergent, Divergent Exercise 10.2 10-13
and Oscillatory Series 9-11 10.4 Identities 10-14
9.11 Geometric Series 9-11 Exercise 10.3 10-19
9.12 Properties of Series 9-13 10.5 Curvilinear Coordinates 10-19
9.13 Zero Test—A Necessary 10.6 Special Orthogonal
Condition for Convergence 9-13 Coordinate Systems 10-21
9.14 Cauchy’s General Exercise 10.4 10-27
Principle of Convergence 9-13
9.15 Tests for Convergence
and Divergence of Series 9-14 11. Vector Integral Calculus
9.16 Series of Positive Terms 9-14 11.1 Vector Integral Calculus 11-1
9.17 Comparison Tests (Type I) 9-15 11.2 Surface Integrals 11-11
9.18 Comparison Tests (Type II) 9-16 Exercise 11.1 11-14
9.19 Limit Form of Comparison 11.3 Volume Integral 11-15
Theorem 9-17 Exercise 11.2 11-16
Exercise 9.2 9-19 11.4 Gauss’s Divergence
Exercise 9.3 9-20 Theorem (G.D.T.) 11-16
xii Contents
Exercise 11.3 11-20 12.8 Periodic Functions 12-29

11.5 Stokes’ Theorem 11-27 12.9 Inverse Transforms—By
Exercise 11.5 11-30 the Method of Partial
Fractions 12-32
Exercise 12.7 12-39
12. Laplace Transforms 12.10 Solution of Linear
12.1 Introduction 12-1 Differential Equation by
12.2 Laplace Transformation 12-1 the Method of Laplace
Transforms 12-40
Exercise 12.1 12-5
12.3 Existence of Laplace Exercise 12.8 12-44
Transform 12-9 12.11 Heaviside’s Partial
Exercise 12.2 12-15 Function Expansion
Theorem 12-45
Exercise 12.3 12-17
Exercise 12.9 12-46
Exercise 12.4 12-18
12.4 Heaviside Function or
Unit-step Function 12-19 Question Bank A-1
12.5 Unit Impulse Function or Multiple Choice Questions A-1
Dirac Delta Function 12-22 Fill in the Blanks A-16
12.6 Convolution Functions 12-23 Match the Following A-31
12.7 Solution of Initial True or False Statements A-38
Value Problems (IVPs)
Using Convolution Question Papers A-41
Theorem 12-26 Bibliography B-1
Exercise 12.5 12-28 Index I-1
Preface
I am pleased to present this edition of Engineering Mathematics-I to the first year B.Tech. students
of Jawaharlal Nehru Technical University (JNTU). The first edition of the book has received good
comments from students and teachers alike, and this has encouraged me to bring out this new edition
strictly in accordance with the revised syllabus (2009–2010) for the first year B.Tech. students of
JNTU.
As in the first edition, the topics have been dealt with in a coherent manner, supported by illustrations
for better comprehension. The chapter on Laplace Transforms has been revised, with more worked-out
examples. Each chapter is replete with examples and exercises, along with solutions and hints, wherever
necessary. The examples in the chapter on Applications of Integration have been augmented with more
problems that give a deeper insight into the topic.
This edition has a question bank, with over 280 Multiple Choice Questions (MCQs) added to the
215 MCQs of the first edition. The question bank also includes Fill in the Blanks, Match the Following
and True or False statements, thus providing the student with an abundant repository of exam-specific
problems. Further, a collection of fully solved question papers has been included at the end of the book
for the student’s benefit.
Suggestions for the improvement of the book are welcome and will be gratefully acknowledged.
Acknowledgements
I express my deep sense of gratitude to Sri Ch. Malla Reddy, Chairman, and Sri Ch. Mahender Reddy,
Secretary, Malla Reddy Group of Institutions (MRGI), whose patronage has given me the opportunity
to write this book.
I am thankful to Prof. R. Madan Mohan, Director (Academics); Col. G. Ram Reddy, Director,
(Administration), MRGI; and Dr R. K. Murthy, Principal, Malla Reddy Engineering College,
Secunderabad, for their kindness, guidance, and encouragement.
E. Rukmangadachari
xiii
Ordinary Differential Equations
1
1.1 Introduction dy a
2. y = x + dy (1,2)
The general laws of science, engineering, medicine, dx dx
social sciences, population dynamics and the like
where the rate of change of quantities is involved d 2y
3. + m2 y = 0 (2,1)
are usually modelled as differential equations. dx2
Formation of differential equations, solution and 2
interpretation of the results are of practical interest, d 2y dy
4. y 2 − = y2 log y (2,1)
especially for engineers and physicists, as they deal dx dx
with many engineering and physical problems. So, 2 3/2
we study in this chapter, methods of solution of dy d 2y
ordinary differential equations of the first order and 5. 1+ =a (2,2)
dx dx2
first degree, their applications to Newton’s law of
cooling the law of natural growth and decay, and 2
orthogonal trajectories. dy dy
6. y = x + a 1 + (1,2)
Further, methods of solution of linear differential dx dx
equations of second and higher order with constant
coefficients are also considered in this chapter. dx dy
7. y + x (a − t) + xy = 0 (1,1)
dt dt
1.1.1 Differential Equation
An equation expressing a relation between functions, d 2x
8. − k 2 y = a cos pt
their derivatives and the variables is called a dt 2
d 2y
differential equation. + k 2 x = a sin pt (2,1)
Differential equations are classified into dt 2
(1) ordinary and (2) partial differential equations.
Partial differential equation
Ordinary differential equation A differential equation containing partial derivatives
of a function (or functions) of more than one variable,
A differential equation containing derivatives of a
with each derivative referring to one of the variables
function (or functions) of a single variable is called
is called a partial differential equation.
an ordinary differential equation.
Example 1.2
Example 1.1
(order, degree)
(order, degree)
d 2y ∂2 z ∂2 z
1. =0 (2, 1) 1. + 2 =0 (2,1)
dx2 ∂x 2 ∂y
1-2 Engineering Mathematics I
∂2 z Example 1.3
2. =0 (2,1)
∂x∂y 1. The following differential equations are linear:
2 2 d 2y dy
∂z ∂z (i) a0 (x) + a1 (x) + a2 (x) y = b (x)
3. + =0 (1,2) dx 2 dx
∂x ∂y d 2y
(ii) + m2 y = 0
∂z ∂z dx2
4. y2 + xy = nxz (1,1)
∂x ∂y 2. The following differential equations are non-
linear:
∂2 u ∂2 u 2
5. x + y = nu (2,1) dy dy
∂x2 ∂y2 (i) = x2 y (ii) = tan y
dx dx
∂z ∂z
6. x +y =z (1,1) 1.2 Differential Equations
∂x ∂x
Differential equations arise in many engineering and
physical problems. The approach of an engineer to
With each differential equation, we associate a pair of
the study of differential equations has to be practical,
positive integers (n, m) called the order and degree,
and so, it consists of
respectively, of the differential equation.
(1) formation of a differential equation from the
physical conditions, called modelling;
Order of a differential equation (2) solution of a differential equation under the
The order of a differential equation is the order of the initial/boundary conditions; and
highest derivative appearing in it. (3) the physical interpretation of the results.
The general form of an nth order ordinary
differential equation in variables x and y is
F(x, y, y , y , · · · , y(n) = 0) 1.2.1 Formation of a Differential
Equation
d ny
where y(n) = . An ordinary differential equation is obtained
dxn when we eliminate arbitrary constants (also called
parameters) from a given relation involving the
Degree of a differential equation variables; the order of the equation being equal to
the number of the constants to be eliminated.
The degree of a differential equation is the degree A partial differential equation is obtained when
or power of the highest derivative, when the we eliminate arbitrary constants or functions from a
equation is freed from radicals and fractions in given relation involving the variables/functions.
respect
of the derivatives, i.e., when the expression We consider here only ordinary differential
F x, y, y , y , · · · , y(n) is written as a rational equations and hence by a differential equation we
integral algebraic expression in y(n) . mean an ordinary differential equation.
If F cannot be expressed in this manner then the Differential equations occur in physical, bio-
degree of the differential equation is not defined. logical, chemical, engineering and geometrical
The order and degree of the differential equations problems, and also in problems relating to eco-
given above are indicated against each of them. nomics, ecology, population studies, medicine, etc.
We can also classify differential equations as linear
and non-linear. A differential equation is called linear
Let φ x, y, c1 , c2 , · · · , cn = 0 (1.1)
if the sum of the powers of the function and its
derivatives in each term is at most equal to unity, be a relation involving variables x and y and n
and otherwise, non-linear. arbitrary constants c1 , c2 , · · · , cn .
Ordinary Differential Equations 1-3
If we differentiate Eq.(1.1) n times with respect to Example 1.6

‘x’ and eliminate the n arbitrary constants between Obtain the differential equation by eliminating λ
the (n + 1) relations, we get an ordinary differential from
equation of the form x2 y2
+ =1 (1)
F x, y, y , y , · · · , y(n) = 0 (1.2) a2 + λ b2 + λ
Solution Differentiating Eq. (1) with respect to
dy d 2y
where y = , y = 2 ,

and ‘x’ once
dx dx
n
d y 2x 2yy
y(n) = n . + 2 =0
dx a2 +λ b +λ
x yy x + yy
Example 1.4 ⇒ 2 = 2 = 2 (2)
Eliminate a and b from y = ax2 + bx and form the a +λ − b +λ a − b2

differential equation. Sum of numerators
=
Sum of denominators
Solution
1 x + yy
y = ax2 + bx (1) ⇒ 2 = 2
a +λ x a − b2
Differentiating twice with respect to ‘x’ we have 1 x + yy
and 2 = (3)
b +λ −yy a2 − b2
y = 2ax + b (2)

y = 2a (3) Substituting in Eq. (1)
1
(3) ⇒ a = y x x + yy y x + yy
2 1= 2 +
(2) ⇒ b = y − 2ax = y − xy a − b2 −y a2 − b2

1 x + yy xy − y
(1) ⇒ y = y · x2 + y − xy x =
2 y a2 − b2
1
= xy − x2 y or x + yy xy − y = a2 − b2 y
2
2
x y − 2xy + 2y = 0. which is the required differential equation.
Example 1.5
Example 1.7
Form the differential equation by eliminating the Determine the differential equation by eliminating
constants A and B from the constants A and B from
y = Ae2x + Be−2x (1)
y = A sin x + B cos 2x (1)
Solution Differentiating Eq. (1) twice with
respect to ‘x’ Solution Differentiating Eq. (1) twice with
respect to ‘x’
y = 2Ae2x − 2Be−2x (2)
y = A cos x − 2B sin 2x (2)
y = 4Ae2x + 4Be−2x (3)
y = −A sin x − 4B cos 2x (3)
From Eqs. (1) and (3) we obtain
From Eqs. (1) and (3), we obtain
y − 4y = 0
y + 4y = 3A sin x (4)
which is the required differential equation. y + y = −3B cos 2x (5)
Substituting the values of A and B in Eq. (2) In the context of differential equations, solution
and ‘integral’ have the same meaning; and the general
y + 4y y +4 solution is sometimes called the primitive.
y = cos x + 2 sin 2x
3 sin x 3 cos 2x
⇒ 3y = y (cot x + 2 tan 2x) Example 1.8
y = Ae3x + Be−3x , where A and B are arbitrary
+y (4 cot x + 2 tan 2x) (6)
constants, is the general solution while
which is the required differential equation.
y = −2e3x and y = 10e−3x
are particular solutions of the second order and first

1.2.2 Solution of a Differential
degree differential equation
Equation
A relation φ(x, y) = 0 defining a function y = f (x) in d 2y
some interval f , f , · · · , f (n) − 9y = 0
I , which

has derivatives
(n)
dx2
such that F x, f , f , f , · · · , f = 0, i.e., satisfying
the differential equation, is called a solution of the It has no singular solution.
differential equation:
Example 1.9
y = (x + a)2 is the general solution;
F x, y, y , y , · · · , f (n) = 0.
y = x2 is a particular solution; and
General (or complete) solution y = 0 is a singular solution of the first order and
second degree differential equation
A relation φ x, y, c1, c2, · · · , cn = 0 containing n
2
independent arbitrary constants ci which is a solution dy
of the differential equation − 4y = 0.
dx

F x, y, y , y , · · · , y(n) = 0
Note that the singular solution y = 0 cannot be
obtained from the general solution for any value of
is called the general (complete) solution of the
the constant ‘a’.
differential equation.
Formation of a differential equation is straight-
forward while the solution of a differential equation
Particular solution is not. So, we have to classify them and find methods
Any solution obtained from the general solution of of solution.
a differential equation, by giving particular values
to the arbitrary constants in it, is called a particular 1.3 First Order and First Degree
solution of the differential equation. Differential Equations
The simplest type of the ordinary differential
Singular solution equations is that of the first order and first degree
and is of the form
A solution φ(x, y) = 0 of the differential equation dy
F x, y, =0 (1.3)
F x, y, y , y , · · · , y(n) = 0 dx
dy
which is neither a general solution nor a particular Solving for we can write Eq. (1.3) as
solution of it, is called a singular solution of the dx
differential equation. Only some equations have dy
singular solutions. = f (x, y) (1.4)
dx
and if we write Example 1.11

dy x(2 log x + 1)
M x, y Solve = .
f x, y = − dx sin y + y cos y
N x, y
Solution Separating the variables, we have
then Eq. (1.4) can be put in the form
x(2 log x + 1)dx − (sin y + y cos y)dy = 0
Mdx + Ndy = 0 (1.5)
Integrating, we get x2 log x = y sin y + c
where M and N are functions of x and y. In order to
study the methods of solution, we classify the first Example 1.12

order and first degree equations as follows: dy x2 − 1 (y + 2)
Solve = 2 .
(1) Variables separable dx y − 1 (x + 2)
(2) Reducible to variables separable
Solution Separating the variables,
(3) Homogeneous equation
(4) Non-homogeneous but reducible x2 − 1 y2 − 1
dx = dy
to homogeneous type x+2 y+2
2
(5) Exact equation x − 22 + 3 y 2 − 22 + 3
⇒ dx = dy
(6) Inexact equation but rendered x+2 y+2

exact using integrating factors 3 3
⇒ x−2+ dx = y − 2 + dy
(7) First order linear equation. x+2 y+2
(8) Reducible to linear equation
Integrating, we get
(i) Bernoulli’s equation
x2
(ii) Others − 2x + 3 log |x + 2|
Now we take up the above types of equations one by 2
one and discuss the methods of solution. y2
= − 2y + 3 log |y + 2| + c1
2
1.3.1 Variables Separable Equation |x + 2|
⇒ x2 − y2 − 4 x − y + 6 log = 2c1
If the first order and first degree differential equa- |y + 2|
tion is of the form f1 (x)g2 (y)dx + f2 (x)g1 (y)dy = 0 x+2
⇒ x − y x + y − 4 + 6 log =c
then it can be written as M (x)dx + N (y)dy = 0 y+2
where M (x) = (f1 /f2 )(x), is a pure function of x;
where c = 2c1 .
and N (y) = (g1 /g2 )(y), a pure function of y, so that
the equation can be readily integrated and the solu- Example 1.13
tions obtained. dy
Solve = (x + y + 2)2 .
Example 1.10 dx
Solve e−4y log x dx + x cos 3y dy = 0. Solution Put x + y + 2 = u
Solution If we multiply the equation by dy du
1 4y 1 ⇒ = − 1 = u2
e it becomes log xdx + e4y cos ydy = 0 dx dx
x x du
Integrating, we get or = dx
u +1
2
2 2
log x + e4y 4 cos 3y + 3 sin 3y = c Integrating, we get
5
1
where c is an arbitrary constant. x+c = tan−1 x + y + 2 .
2
EXERCISE 1.1 (b) ex tan y dx + (1 − ex ) sec2 y dy = 0.

Ans: tan y = c (1 − ex )
1. Form the differential equation in each of the
following cases by eliminating the parameters (c) x 1 + y2 dx + y 1 − x2 dy = 0.
mentioned against each. Ans: 1 + y2 − 1 − x2 = c
(a) y = ax + bx2 (a, b). (d) 1 − x2 sin−1 x dy + y dx = 0.

Ans: 2y + x2 y = 2xy
Ans: y sin−1 x = c
(b) x = A cos(pt + B) (A, B).
Ans: x + p2 x = 0 ( denotes dy 1 − y2
differentiation with respect to ‘t’) (e) + = 0.
dx 1 − x2
a Ans: sin−1 x + sin−1 y = c
(c) y = mx + (m).
m
Ans: y = xy + ya
1 + y2 dx + y 1 + x2 dy = 0.
(f) x √
(d) y = ax2 + bx + c (a, b, c). Ans: 1 + x2 + 1 + y2 = c
Ans: y = 0 2
(g) ydx − xdy + 3x2 y2 ex dx = 0.
(e) (x − h) + (y − k) = r
2 2 2
(h, k). 3
Ans: (x/y) + ex = c
3
Ans: 1 + y12 = r 2 y22 ,

dy d 2y (h) y2 − x2 y + 2xy = 0
y1 = , y2 = 2
dx dx Ans: x4 + 2e2x = c + 4ey
(f) y = ex (A cos x + B sin x) (A, B). √
(i) 1 + x2 1 + y2 dx + dy = 0.
Ans: y2 − 2y1 + 2y = 0 x 1
Ans: 1 + x2 + sinh−1 x + sinh−1 y = c
(g) Find the differential equation for the family 2 2
of circles with their centres on the x-axis.
dy
(Hint: x2 + y2 + 2gx + c = 0 g, c (j) 2xex
2 +y2
= 1 + 2y e−y .
parameters) dx
2 2
Ans: 1 + y12 + yy2 = 0 Ans: ex + ey+y = c
2 dy
(h) Form the differential equation for the family (k) x − y = a2 .
of circles, touching the x-axis at (0, 0). dx
a x−y−a
(Hint: x2 + y2 − 2fy = 0, f parameter) Ans: y = log +c
Ans: (y2 − x2 )y + 2xy = 0 2 x−y+a
(i) Form the differential equation of all parabo- dy

(l) = cos x + y .
las each having its latus-rectum = 4a and its dx
axis parallel to the x-axis. Ans: tan[ x + y /2] = x + c
(Hint: (y−k)2 = 4a(x−h); h, k parameters)
Ans: 2ay2 + y13 = 0 dy x 2 log x + 1
(m) = .
dx sin y + y cos y
2. Solve Ans: y sin y = x2 log x + c
dy y+b
(a) = . dy
dx x+a (n) x (x − 1) + y y − 1 = 0.
Ans: x + a = c y + b
dx
Ans: (x − 1) y − 1 = cxy
dy where v = y/x (or v = x/y) and c is an arbitrary

x2 + 1 + y2 + 1 = 0
(o) dx constant.
y (0) = 1. [JNTU 2001] We extend the meaning of ‘homogeneity’ to
Ans: x + y + xy = 1
include differential equations with functions of the
dy following type since they are solvable by the above
(p) x + y + 4 = 0. [JNTU 2001] method.
dx
Ans: x y + 4 = c
Example 1.16
x2 + y2
f x, y = ex/y + sin(y/x)
xy
1.3.2 Homogeneous Equations
A polynomial f (x, y) is called a homogeneous Example 1.17
function of degree n if f (tx, ty) = t n f (x, y) for some x2 + y2
f x, y = sin 2 + tan−1 (y/x)
t ∈ R or equivalently, f (x, y) = xn f (1, y/x). x − y2
Example 1.14 Example 1.18

f (x, y) = 2x + 3y is a homogeneous function of f x, y = log x2 − 2xy + y2 x2 + y2
degree 1 since
y y
f (x, y) = x 2 + 3 = xf (1, ). Example 1.19
x x dy y2
Solve = .
Example 1.15 dx xy + x2
g(x, y) = 3x2 −4xy +7y2 is a homogeneous function
of degree 2 since Solution Put y = vx

y y 2 dy dv
g(x, y) = x 3 − 4
2
+7 = x2 f (1, y/x) ⇒ =v + x
x x dx dx
A first order and first degree differential equation The equation transforms to
dy dv v2 dv v2 − v2 − v
= f (x, y) (1) v+x = 2 ⇒ x =
dx dx v +1 dx v+1
is called a homogeneous type if f (x, y) is a rational dx v + 1
function which is homogeneous and of degree zero. Separating the variables + dv = 0
x v
That is, f (x, y) is the quotient of two polynomials
each of the same degree n. If we put y = vx (or Integrating, log x + log v + v = constant
x = vy) Eq. (1) transforms to Dropping logarithms we may put the solution as
y
dv dv 1 ye x = c
v+x = f (1, v) or v + y =
dx dy f (v, 1)
where c is an arbitrary constant.
where the variables are separable. Separating the
variables and integrating, we get the general solution Example 1.20
as Solve xdy − ydx = x2 + y2 dx
⎛ ⎞
dv
x = ce Solution Put y = vx ⇒ dy = vdx + xdv
⎜ f (1, v) − v ⎟
⎝ f (v, 1) dv ⎠ (2) The equation becomes
or y = ce
1 − vf (v, 1) x(vdx + xdv) − vxdx = x 1 + v2 dx
Separating the variables 2. xdy − ydx = x2 + y2 dx.

Ans: cx2 = x2 + y2 + y
dx dv
=√
x 1 + v2 dy y (x+y)
3. = +e x .
Integrating dx x
(x+y)
Ans: e x log x = c
log x − log(v + 1 + v2 ) = Constant
4. 2xydy = (x2 + y2 ). [JNTU 1999]
y
Dropping logarithms, replacing v by and sim- Ans: cx = (x2 − y2 )
x
plifying we get
5. x2 ydx = (x3 + y3 )dy.
x
y+ x2 + y2 = cx 2 Ans: y = ce y
where c is an arbitrary constant. 6. (x2 − y2 )dx = xydy.

Ans: x2 (x2 − 2y2 ) = c
Example 1.21
x x x
Solve 1 + e y dx + e y 1− dy = 0 7. x tan
y y dy
+ x sec2 ·
y
= y sec2 .
y x x dx x
y
Solution Put x = vy Ans: x tan = c
x
dx dv
⇒ =v+y x x dy
dy dy 8. ye y = xe y + y .
x
dx
The equation transforms to Ans: y = ce y
dv ev (1 − v)
v+y =− y dy y
dy 1 + ev 9. x sin = y sin + x.
x dx x
dv e (1 − v)
v y
Ans: x2 sin + tan
y
=c
⇒ y =− −v
dy 1 + ev x x
−ev (1 − v) − v(1 + ev )
= dy y3 + 3x2 y
1 + ev 10. = 3 .
dx x + 3xy

2

Separating the variables we have Ans: x10 = cy6 x2 − y2
dy 1 + ev
+ dv = 0 x x x
y v + ev 11. 1 + 2e y dx + 2e y 1− dy = 0.
y
⇒ y (v + ev ) = c, on integration Ans: xecos(y/x) = c
x
⇒ x + ye = c y
dy y2
where c is an arbitrary constant. 12. x + = y.
dx x
x y
EXERCISE 1.2 Ans: sin−1 = log
y c
Solve:
x x
dy 13. xy log dx + y − x log
2 2
dy = 0.
1. y2 = (xy − x2 ) . y y
dx x2 y
Ans: y = ce x
y Ans: 2
2 log + 1 = log cy
4y x
dy y Integrating, we obtain the general solution as

14. = √ . [JNTU 1995]
dx x + xy
√x
2 ax2 + 2bxy − my2 + 2cx − 2ny + C = 0
Ans: y = ce y
(This is an exact equation. We will study exact equa-
15. y2 − 2xy dx = x2 − 2xy dy. [JNTU 1995] tions in detail in Section 1.3.4)
Ans: xy(x − y) = c
Case (5) b + l = 0
a b c
1.3.3 Non-Homogeneous Equations (i) = = λ =
l m n
The first order and first degree differential equation Put ax + by = u. The variables are then
of the type separable and the equation becomes integrable.
a b
dy ax + by + c (ii) = . Shift the origin to the point (h, k)
= (1.6) l m
dx lx + my + n by the substitution
x = X + h, y = Y + k
where a, b, c, l, m, n are constants, is called a non- Now, choose h, k such that
homogeneous differential equation.
If c = n = 0, then Eq. (1.6) reduces to a ah + bk + c = 0
homogeneous equation whose solution we have lh + mk + n = 0
discussed earlier. We assume that at least one of c
and n is not zero. We consider below the various a b
which is possible since = 0 and the
cases and the method of solution in each case. In l m
this section, C denotes an arbitrary constant instead differential equation now reduces to the homoge-
of c. neous type
dY aX + bY
=
a b c dX lX + mY
Case (1) = = = λ (say):
l m n which can be solved by putting Y = vX or X = vY .
dy
Eq. (1.6) reduces to = λ whose general solution
dx Example 1.22
is y = λx + C
dy 3x − y + 2
Case (2) (i) b = l = 0 or Solve = .
dx 6x − 2y + 4
(ii) a = m = 0
In either case, the variables are separable and hence Solution Here,
the equation is integrable.
a = 3, b = −1, c = 2
Case (3) c = 0 or n = 0 or both c = 0, and n = 0
(at least one of c and n is non-zero) l = 6, m = −2, n = 4
(i) a = b = 0 (c = 0) or a b c 1
(ii) l = m = 0 (n = 0) Since = = = , the differential equation
l m n 2
Eq. (1.6) can be solved by putting u = lx + my or dy 1
u = ax + by as the case may be. reduces to = whose general solution is
dx 2
1
Case (4) b + l = 0 y = x + C, where C is an arbitrary constant.
2
Separating the variables and grouping the terms,
we can write Eq. (1.6) as Example 1.23
(ax + c)dx + bd(xy) − (my + n)dy = 0 dy 2x + 3
∵ l = −b; ydx + xdy = d(xy) Solve = .
dx 4y − 5
Solution Here Separating the variables,

u−2
a = 2, b = 0, c = 3 dx = du
3u − 1
l = 0, m = 4, n = −5 3(u − 2) du
⇒ 3dx = du = du − 5
Since b = l = 0, the variables are separable. 3u − 1 3u − 1
Separating the variables, we have Integrating, we get
(2x + 3)dx − (4y − 5)dy = 0 5
3x + constant = u − log(3u − 1)
3
Integrating, x2 − 2y2 + 3x + 5y + C = 0 5
3x + constant = 3x + y − log(9x + 3y − 1)
where C is an arbitrary constant. 3
Dropping logs, the general solution is
Example 1.24
(9x + 3y − 1)5 = Ce3y .
dy 4y − 3
Solve = .
dx 2x + 5 Example 1.26
Solution Here, dy x − 2y + 4
Solve = .
dx 3
a = 0, b = 4, c = −3
l = 2, m = 0, n = 5 Solution Here,
Since a = m = 0, the variables are separable. a = 1, b = −2, c = 4

Separating the variables: l = 0, m = 0, n=3
dy dx Put u = x − 2y
=
4y − 3 2x + 5
du dy u+4 −2u − 5
=1−2 =1−2 =
Integrating, we get dx dx 3 3
1 1 1 Separating the variables,
log(4y − 3) − log(2x + 5) = log C 3
4 2 4 dx + du = 0
2u + 5
Dropping logs and removing fractions
Integrating, we get
(4y − 3) = k(2x + 5)2 .
3
x + constant = log(2u + 5)
Example 1.25 2
Dropping logs, the general solution is
dy 5
Solve = .
dx 3x + y − 20 (2x − 4y + 5)3 = Ce2x .
Solution Here, Example 1.27
a = 0, b = 0, c = 5 dy y+x−2
Solve = . [JNTU 1997, 1999]
l = 3, m = 1, n = −2 dx y−x−4
Put u = 3x + y Solution Here,

a = 1, b = 1, c = −2
du dy 5 3u − 1
=3+ =3+ = l = −1, m = 1, n = −4 b+l =1−1=0
dx dx u−2 u−2
Separating the variables and grouping the terms du

= du + 5
we can write the differential equation as 4u + 1
(x − 2)dx + ydx + xdy − (y − 4)dy = 0 Integrating, we get

⇒ (2x − 4)dx + 2d(xy) − (2y − 8)dy = 0, 5
2x + constant = u + log(4u + 1)
4
on multiplying by 2 and noting that 4(x − 2y) + constant = 5 log(4x + 8y + 1)
ydx + xdy = d(xy).
Integrating, we have the general solution as Dropping logs, the general solution is obtained as
x2 + 2xy − y2 − 4x + 8y + C = 0. (4x + 8y + 1)5 = Ce(4x−8y) .

dy 2x − 4y + 3 dy 2x + 5y + 1
Solve = . Solve = .
dx 4x − 8y + 1 dx 5x + 2y − 1
Solution Here, Solution Here,
a = 2, b = −4, c = 3 a = 2, b = 5, c = 1
l = 4, m = −8, n = 1 (b + l = 0) l = 5, m = 2, n = −1
Separating the variables and grouping the terms a b 2 5
= = −21 = 0
we can write the differential equation as l m 5 2
(2x + 3)dx − 4(ydx + xdy) + (8y − 1)dy = 0 Put x = X + h, y = Y + k

The differential equation reduces to
Integrating, we get
dY 2X + 5Y
x2 − 4xy + 4y2 + 3x − y + C = 0. =
dX 5X + 2Y
Example 1.29 Provided (h, k) are chosen such that
dy x + 2y − 1 2h + 5k + 1 = 0
Solve = .
dx 2x + 4y + 3 5h + 2k − 1 = 0
Solution Here, By rule of cross-multiplication
a = 1, b = 2, c = −1 h k
=
a b 1 c 1 5(−1) − 2 · 1 1 · 5 − 2(−1)
= = = = −
l m 2 n 3 1
l = 2, m = 4, n = 3 =
2·2−5·5
du dy
Put u = x + 2y ⇒ =1+2 1 1
= (h, k) = ,−
dx dx 3 3
u−1 4u + 1
= 1+2 = Put Y = vX
2u + 3 2u + 3
dY dv
Separating the variables and multiplying by 2 ⇒ = v+X
dX dX
2u + 3 (4u + 1) + 5 2 + 5v
2 · dx = 2 · du = du =
4u + 1 4u + 1 5 + 2v
dv 2 + 5v This method applies if l = b and m = a.

⇒X = −v
dX 5 + 2v
2 + 5v − 5v − 2v 2 EXERCISE 1.3
=
5 + 2v Solve the following
Case
Separating the variables
dy 3x − 6y + 3
dX 2v + 5 1. = 1
2· + 2 dv = 0 dx x − 2y + 1
X v −1 Ans: y = 3x + C
Integrating, we get
dy 3x + 2
5 v−1 2. = 2(i)
2 log X + log(v2 − 1) + log = Constant dx 2y − 3
2 v+1 Ans: 3x2 − 2y2 + 4x + 6y + C = 0
5 v−1
log x2 (v2 − 1) + log = Constant
2 v+1 dy y+3
3. = 2(ii)
dx 2x − 7
Dropping logs after multiplying by 2
Ans: (y + 3)2 = C(2x − 7)

Y −X 5 dy 9
(Y + X ) (Y − X )
2 2
= c1 4. = 3(i)
Y +X dx 5x − 7y + 13
⇒ (Y − X )7 = c1 (Y + X )3 Ans: (25x − 35y + 2)63 = Ce5y

1 1
But (X , Y ) = (x − h, y − k) = x − , y + dy 2x − 5y − 6
3 3 5. = 3(ii)
2 dx 11
Y −X =y−x+ ; Y +X =y+x Ans: (10x − 25y − 52)11 = Ce−5x
3
dy x−y+2
The general solution is 6. = 4
dx x+y−1
(3y − 3x + 2)7 = C(y + x)3 . Ans: y + 2xy − x − 2y − 4x + C = 0
2 2
Alternative method (Method of proportions) dy 2x − y + 1

7. = 4
dy dx dx x + 2y − 3
= Ans: y2 + xy − x2 − 3y − x + C = 0
2x + 5y + 1 5x + 2y − 1
d(y + x) d(y − x) dy 2x + y + 6
= = 8. = 4
7(x + y) 3(y − x) + 2 dx −x + y − 3
Sum of numerators Ans: 2x2 + 2xy − y2 + 12x + 6y + C = 0
=
Sum of denominators
dy 2x + 3y + 1
Difference of numerators 9. = 4
= dx −3x + 4y − 1
Difference of denominators
Ans: x2 + 3xy + 2y2 + x − y + C = 0
Integrating, we get
dy y−x
1 1 10. = 4 or 5(i)
log(x+y) = log(3y−3x+2)+constant dx y−x+2
7 3 Ans: x2 + 2xy − y2 − 4y + C = 0
Multiplying by 21 and dropping logs, we have the
dy x − 3y + 4
general solution 11. = 4 or 5(i)
dx 3x − 9y − 7
(x + y)3 = C(3y − 3x + 2)7 Ans: x − 6xy + 9y + 8x + 14y + C = 0
2 2
dy x + 2y − 3 1 y
12. = 5(ii) or as d(xy) = 0, d(x2 + y2 ) = 0 and d = 0,
dx 2x + y − 3 2 x
method of proportions respectively.
Ans: (y − x)3 = C(y + x − 2) The following theorem gives a criterion for exact
equations.
dy ax + by − a Theorem 1.1 If M (x, y) and N (x, y) are real-
13. = q 5(ii) or
dx bx + ay − b valued functions having continuous first partial
method of proportions derivatives on some rectangle R :| x − x0 |≤ a, |
Ans: (y − x + 1)a+b (y + x − 1)a−b = C y − y0 |≤ b, then a necessary and sufficient condition
for the equation
dy 2x + y − 2
14. = 5(ii)
dx √ + y −3
3x
Mdx + Ndy = 0 (1.7)
2+ 3 y √
Ans: √ log +1− 3
2 3 √ x −1
to be exact in R is that
2− 3 y √
− √ log +1+ 3 ∂M ∂N
2 3 x−1 = in R. (1.8)
+ log(x + 1) = C ∂y ∂x
dy x+y+1 Proof: (i) The condition is necessary: Suppose

15. = 5(i) or
dx x+y−1 Eq. (1.7) is exact. Then there exists a function u(x, y)
method of proportions such that
Ans: y = x + log(x + y) + C
∂u ∂u
du =dx + dy
16. (2x + 3y − 7)xdx + (3x + 2y − 8)ydy = 0
2 2 2 2
∂x ∂y
4 = Mdx + Ndy (1.9)
[Hint: Put x2 = X , y2 = Y and group the terms ∂u ∂u
(2X − 7)dX + 3 (YdX + XdY ) ⇒M = , N = (1.10)
∂x ∂y
+ (2Y − 8)dY = 0]
Ans: X 2 + 3XY + Y 2 − 7X − 8Y + C = 0; Differentiating M and N partially with respect to
x4 + 3x2 y2 + y4 − 7x2 − 8y2 + C = 0 y and x, respectively
dy (2ex − ey + 1)ex
17. = 4 ∂M ∂ ∂u ∂2 u
dx (ex + 2ey )ey = = (1.11)
Ans: e2x + ex = C + ex+y + e2y ;
∂y ∂y ∂x ∂y∂x

∂N ∂ ∂u ∂2 u
dy y−x+2 = = (1.12)
18. = ∂x ∂x ∂y ∂x∂y
dx 3y − 3x + 6
Ans: 3y = x + C From Eqs. (1.11) and (1.12) it follows that

1.3.4 Exact Equations ∂M ∂N ∂2 u ∂2 u
= ∵ = (1.13)
Definition A differential equation which is ∂y ∂x ∂y∂x ∂x∂y
obtained from its primitive by mere differentiation
without any further operation is called an exact (ii) The condition is sufficient: Assume that Eq.
equation. (1.8) holds.

The equations xdy + ydx = 0, xdx + ydy = 0 and
xdy − ydx Let v(x, y) = Mdx
= 0 are exact since these can be written
x2 y constant
where the integral is evaluated partially with respect are exact differentials will not only prove the
∂v exactness of the equation but also yield the general
to x treating y as constant so that = M.
∂x solution, on integration, as the illustrative examples
will show.
∂N ∂M ∂ ∂v
Now, = = (1.14)
∂x ∂y ∂y ∂x Note 2

∂2 v ∂ ∂v The following formula
= =
∂y∂x ∂x ∂y
Mdx + (terms of N
∂ ∂v
⇒ N− =0 y constant
∂x ∂y
not containing x)dy = c (1.16)
∂v
⇒N− = φ(y), a function of y alone
∂y though in many cases, gives the correct solution, fails
∂v in some cases and is therefore not advisable.
⇒N = + φ(y)
∂y (Example 1.31 shows the failure of this method to
give the correct solution.)
Hence the equation becomes
Example 1.31
∂v ∂v
0 = Mdx + Ndy = dx + + φ(y) dy Solve y sin 2x dx = (y2 + cos2 x)dy.
∂x ∂y

Solution Here
= d v + φ(y)dy (1.15)
M = y sin 2x; N = −y2 − cos2 x
∂M ∂N
which proves that the equation is exact. = sin 2x; = −2 cos x(− sin x)
∂y ∂x
Procedure for solving an exact equation = sin 2x
∂M ∂N
∂M ∂N The equation is exact, since =
(1) Test if = . ∂y ∂x
∂y ∂x

(2) Compute v = Mdx Now v = Mdx
y constant
y constant
(or) u= Ndy.
x constant
= y sin 2xdx
y constant
∂v y
(3) Compute N− dy = − cos 2x
∂y
2
∂u
(or) M− dx. ∂v 1
∂x N− = −y2 − cos2 x + cos 2x
∂y 2
(4) The 1
general solution
is = −y2 − cos2 x + 2 cos2 x − 1
∂v 2
Mdx + N− dy = c
∂y 1
= −y −
2
y constant
2
∂u
(or) Ndy + M− dx = c.
∂x
∂v 1 1
x constant
N− dy = − y3 − y
∂y 3 2
Note 1
In most cases, collecting and grouping terms which Hence the general solution is
y 1 1 Alternative method: Grouping the terms

− cos 2x − y3 − y = constant
2 3 2
1 3 (3x + 5)dx − 2(ydx + xdy) + (4y + 1)dy = 0
⇒ y cos x + y = c
2
3 3 2
or d x + 5x − 2d(xy) + d(2y2 + y) = 0
y 1 3 2
⇒ (cos 2x + 1) + y = c.
2 3
This shows that the equation is exact. Integrating
Note it, we get
If we compute the solution using Eq. (1.16) we get
3 2
y x + 5x − 2xy + 2y2 + y = Constant
− cos 2x − y2 dy = constant 2
2 or 3x2 − 4xy + 4y2 + 10x + 2y + c = 0.
y y3
⇒ cos 2x + =c Example 1.33
2 3
dy 3x2 − 2xy − 5
Solve = 2 .
1 dx x + y2 − 2y
in which the term y is missing.
2 Solution Writing the equation in the standard
Example 1.32 form
dy 3x − 2y + 5
Solve + = 0. l(3x2 − 2xy − 5)dx − (x2 + y2 − 2y)dy = 0
dx −2x + 4y + 1
M = 3x2 − 2xy − 5; N = −x2 − y2 + 2y
Solution In the standard form, the equation is
My = −2x = Nx = −2x
(3x − 2y + 5)dx + (−2x + 4y + 1)dy = 0
The equationis exact.
M = 3x − 2y + 5; N = −2x + 4y + 1
v= Mdx = x3 − x2 y − 5x;
My = −2 = Nx = −2
y constant
The equation is exact since ∂M /∂y = ∂N /∂x ∂v

= −x2
∂y

v= Mdx = 3x − 2y + 5 dx

y constant ∂v 2 1
N− dy = −y + 2y dy = y3 + y2
3 2 ∂y 3
= x − 2xy + 5x
2
∂v The general solution is,
= −2x
∂x 1
∂v x3 − x2 y − 5x − y3 + y2 + constant = 0
N− = −2x + 4y + 1 − (−2x) 3
∂y ⇒ 3x3 − y3 − 3x2 y + 3y2 − 15x + c = 0.
= 4y + 1
Alternative method: Grouping the terms
∂v
N− dy = 2y2 + y
∂y (3x2 − 5)dx + (2y − y2 )dy − (2xydx + x2 dy) = 0

The general solution is 1 3
⇒ d(x − 5x) + d y − y − d(x2 y) = 0
3 2
3
∂v 3
v+ N− dy = x2 − 2xy + 5x + 2y2 + y The equation is exact. Integrating and multiplying
∂y 2
by 3, we get the general solution as
= Constant
⇒ 3x2 − 4xy + 4y2 + 10x + 2y + c = 0. 3x3 − 3x2 y − y3 + 3y2 − 15x + c = 0.
Example 1.34 The equation is exact.

x x x
Solve 1 + e y dx + 1 − e y dy = 0.
y v= Mdx = xy2 − x2 y;
[JNTU 2000]
y constant
Solution Here ∂v
= 2xy − x2
∂y
x x x
M = 1 + ey ; N = 1 − ey ∂v
y N− = −x2 + 2xy − 2xy + x2 = 0
∂y
x x
My = e y − The general solution is xy2 − x2 y = c.
y

x 1 1 yx Alternative method: Grouping the terms
= Nx = 1 − − e
y y y
x x y2 dx + 2xydy = x2 dy + 2xydx
= − 2ey
y ⇒ d xy2 − d x2 y = 0
(The equation is exact.)
The equation is exact.
The general solution is xy2 − x2 y = c.
x
v= Mdx = x + ye y ;
Example 1.36
y constant xdy − ydx
Solve xdx + ydy = .
∂v x x x x x x2 + y2
= 1 · ey + y − 2 ey = 1− ey
∂y y y
Solution
∂v x x x x
1
N− = 1− ey − 1 − ey = 0 LHS = d(x2 + y2 )
∂y y y 2
x xdy−ydx
The general solution is x + ye y = c. x2
RHS =
x2 +y2
Alternative method: Grouping the terms x

x x x (dividing the numerator and denominator by x2 )
dx + e dy + dx − dy e y = 0
y
y
d yx −1 y
x x x = y 2 = d tan .
⇒ dx + e dy + y · e d
y y =0 1+ x
y x
x
⇒ dx + d ye y =0 From the above results, we observe that the equa-
tion is exact. Its general solution is
The equation is exact. y
x x2 + y2 = 2 tan−1 + c.
The general solution is x + ye y = c. x

Solve (y − 2xy)dx = (x − 2xy)dy.
2 2 Solve (ey + 1) cos x dx + ey sin xdy = 0.
[JNTU 1995]
Solution

Solution ey cos xdx + ey sin xdy + cos xdx = 0
M = y2 − 2xy; N = 2xy − x2 ⇒ d (ey sin x) + d(sin x) = 0
My = 2y − 2x = Nx = 2y − 2x (The equation is exact.)
Integrating, we get the general solution as y(xy + ex )dx − ex dy

11. = 0.
y2
(ey + 1) sin x = c. ye dx − ex dy
x
Hint: xdx + = 0.
EXERCISE 1.4 y2
x2 ex
Ans: 2 + y = c
Solve:
1. (2x + 3y + 4)dx + (3x − 6y + 2)dy = 0. ydx − xdy
12. + xex dx = 0.
Ans: x2 + 3xy − 3y2 + 4x + 2y + c = 0 y2
Ans: yx + (x − 1)ex = c
2. (3x2 − 9x2 y2 + 2xy)dx

+ (6y2 − 6x3 y2 + x2 )dy = 0. 1.3.5 Inexact Equation—Reducible
Ans: x + 2y − 3x3 y2 + x2 y = c
3 3 to Exact Equation by
Integrating Factors
Integrating factor (I.F.)
3. ex (sin x + cos x) sec y dx + ex sin x sec y
tan y dy = 0. If the differential equation Mdx + Ndy = 0 becomes
Ans: ex sin x sec y = c exact when we multiply it by a function μ(x, y) then
μ(x, y) is called an integrating factor of the equation.
Consider the equation ydx − xdy = 0
4. (cos x cos y − cot x) dx = sin x sin y dy.
Here
Ans: sin x cos y = log | c sin x |
∂M ∂N
M = y, N = −x ⇒ = 1 = = −1
∂y ∂x
5. y 1 + 1x + sin y dx
+ {x + log x + x cos y} dy = 0. The equation is not exact. If we multiply it by 1
y2
Ans: xy + y log x + x sin y = c we get
1 x
dx − 2 dy = 0.
y y
6. 4y sin 2xdx + (2y + 3 − 4 cos2 x)dy = 0.
Ans: y2 + y − 2y cos 2x = c Now M = 1y , N = − yx2 so that
∂M 1 ∂N
xdy − ydx =− 2 = .
7. xdx + ydy + = 0. ∂y y ∂x
x2 − y2
x+y
Ans: (x2 + y2 )e x−y = c So, the equation becomes exact. y12 is an integrating
factor. We can easily check that x12 , xy1 , x2 +y
1
2 are also
8. (x2 − ay)dx = (ax − y2 )dy. integrating factors for the equation.
Ans: x3 + y3 = 3axy + c Integrating factor can be found by inspection,
after grouping of terms. Table 1.1 gives the list of
xdx + ydy xdy − ydx integrating factors.
9. + 2 = 0.
x2 + y2 x − y2
Example 1.38
Ans: (x2 + y2 ) x−y = c
x+y
Solve the following differential equations after
finding the integrating factor in each case:
dy y cos x + sin y + y
10. + = 0. Differential Equation Integrating Factor
dx sin x + x cos y + x 1
Ans: xy + y sin x + x sin y = c (i) ydx−xdy+y xe dx = 0
2 x
y2
Table 1.1
S. No. Group of terms Integrating factor Exact differential
1. xdy + ydx 1 d(xy)
1 xdy + ydx
2. xdy + ydx = d(log xy)
xy xy
1−a
1 xdy + ydx xy
3. xdy + ydx a = d (a = 1)
(xy)a xy 1−a
2 2xdx + 2ydy
4. xdx + ydy = d[log(x2 + y2 )]
x2 + y2 x2 + y2
2
2 2xdx + 2ydy (x + y2 )1−a
5. xdx + ydy = d (a = 1)
(x2 + y2 )a (x2 + y2 )a 1−a
6. xdx + ydy 2 d(x2 + y2 )
1 xdy − ydx y
7. xdy − ydx =d
x2 x2 x

1 ydx − xdy x
8. xdy − ydx − =d
y2 y2 y
1 dy dx y
9. xdy − ydx − = d log
xy y x x
xdy−ydx
1 xdy − ydx x2 −1 y
10. xdy − ydx = 2 = d(tan )
x2 + y2 x2 + y2 1 + yx x
ydx−xdy
1 y2 −1 x
11. xdy − ydx − = d tan
x2 + y2 2 y
1 + yx

y 2xydy − y2 dx y2
12. 2xdy − ydx = d
x2 x2 x
2
x 2xydx − x dy 2
x
13. xdy − 2ydx − =d
y2 y2 y
2 xdy−ydx
2 2xdy − 2ydx x2 x+y
14. xdy − ydx = = d log
x2 − y2 x2 − y2 1 − yx
2 x−y
1 derive the general solution as follows:

(ii) ydx − xdy + 2x2 y sin x2 dx = 0
xy
2 y2 1 ydx − xdy
(iii) ydx − xdy + 2x ye dy = 0 (i) + xex dx = 0
x2 y2
2 dx 1
(iv) ydx − xdy + x + y2 √ =0 2 x
1−x 2 x + y2 ⇒ d + xex dx = c
2 1 y
(v) xdx + ydy + x + y2 tan xdx = 0 x
x2 + y2 ⇒ + (x − 1) ex = c.
1 y
(vi) y(1 + xy)dx + x(1 − xy)dy = 0
(xy)2
Solution The suitable integrating factor in each
case is shown against each differential equation. ydx − xdy
After multiplying by the integrating factor, we can (ii) + 2x sin x2 dx = 0
xy
dx dy 3. xdy + 2ydx = 2y2 xdy

⇒ − + sin x2 d x2 = 0 2
x y Ans: xy2 = cey
x
⇒ log = cos x2 + c. 4. (x2 +y2 )(xdy − ydx) = (x2 − y2 )(xdx+ydy)
y
Ans: (x + y) = c(x − y)(x2 + y2 )
ydx − xdy 2 5. 2xydy − y2 dx + x3 ex dx = 0

(iii) + ey dy2 = 0 Ans: y2 + x (x − 1) ex + cx = 0
x2
y 2
Given below are the rules for finding integrating
⇒ −d + d ey = 0
x factors of Mdx + Ndy = 0 depending on the nature
y
⇒ − + ey = c of functions M and N .
x
y Rule 1
⇒ e =c+ .
y
x If Mdx + Ndy = 0 is a homogeneous differential
ydx − xdy dx 1
(iv) +√ =0 equation and Mx + Ny = 0, then is an
x +y
2 2
1 − x2 Mx + Ny
ydx−xdy
integrating factor of the equation.
x2 dx
⇒ 2 +√ =0 Note 1
1 + yx 1 − x2
Recall that we have already solved homogeneous
−d yx dx differential equations reducing them to variables sep-
⇒ 2 +√ =0
1 + yx 1 − x2 arable form by putting y = vx or x = vy. Rule 1
y provides an alternative method for its solution.
⇒ sin−1 x = tan−1 + c.
x Note 2
M N
If Mx + Ny = 0, then = and the equation
2xdx + 2ydy y −x
(v) + tan xdx = 0 becomes ydx − xdy = 0 whose general solution is
x2 + y2 x
= c.
⇒ d log x2 + y2 + d log (sec x) = 0 y

⇒ x2 + y2 (sec x) = c. Rule 2
If Mdx + Ndy = 0 is of the form

(vi) ydx + xdy + xy ydx − xdy = 0 f1 (xy)ydx + f2 (xy)xdy = 0

d xy dx dy and Mx − Ny = 0, then 1
is an integrating factor
⇒ 2 + − =0 Mx−Ny
xy x y of the equation.
1 x Rule 3
⇒ − + log = c. ∂M ∂N
xy y If − N = p(x), purely a function of x
∂y ∂x
EXERCISE 1.5 alone, then e p(x)dx is an integrating factor of the
equation Mdx + Ndy = 0.
Find the integrating factors by inspection and solve
Note that the linear equation dy
dx
+ Py = Q where
the following.
P, Q are functions of x alone may be written in the

1. xdy − y − 3x2 dx = 0 form Mdx +Ndy = 0 where M = Py −Q and N = 1
y so that
Ans: + 3x = c
x ∂M ∂N
∂y
− ∂x P−0
2. ye −x −x
+ e dy = 2xy dx 2 =
−x
N 1
Ans: e y = x2 + c = P,
y2
purely a function
of x and the integrating factor in ∴ The general solution is log x = 2x2
+ c.
this case is e p(x)dx , as we will see in linear equations
Example 1.40
discussed below.
Rule 4
Find an integrating factor and solve
∂N ∂M dy y x2 + y2
If − M = q(y), purely a function of y = + .
∂x ∂y dx x x2
alone then e q(y)dy is an integrating factor of the Putting
Solution
the equation in the standard
equation Mdx + Ndy = 0. form xy + x2 + y2 dx − x2 dy = 0
Rule 5 M = xy + x2 + y2 , N = −x2
If Mdx + Ndy = 0 is expressible in the form ∂M ∂N
= x + 2y, = −2x
xa yb (mydx + nxdy) + xc yd (pydx + qxdy) = 0 ∂y ∂x
∂M ∂N
where a, b, c, d, m, n, p, q are all constants such that Since = , the equation is not exact
mq − np = 0, then xh yk is an integrating factor of ∂y ∂x
the equation for some suitable constants h, k to be Mx + Ny = x xy + x2 + y2 − x2 y

determined from the two equations a+h+1 m
= b+k+1
n = x x2 + y2 = 0
and c+h+1
p
= d+k+1
q
.
The given equation is a homogeneous differential
Example 1.39 equation.
1 1
Solve (x2 +y2 )dx−xydy = 0 by finding an integrating ∴ Integrating factor = = 2
factor. Mx + Ny x x + y2
Solution This is a homogeneous differential Multiplying the equation by the integrating factor
equation. dx xdy − ydx
= 2
Here, x x + y2

M = x2 + y2 , N = −xy; dx
xdy−ydx
d yx
x2
∂M ∂N ⇒ = 2 = 2
= 2y, = −y x 1 + yx 1 + yx
∂y ∂x y
∂M ∂N ⇒ log x = tan−1 + c.
Since = , the equation is not exact x
∂y ∂x
Example
1.41

Mx + Ny = (x + y )x − (xy)y = x = 0
2 2 3 Solve xy + 1 ydx + −xy + 1 xdy = 0.
Multiplying by the integrating factor, Solution This equation is of the form

1 1 f1 xy ydx + f2 xy xdy = 0
= 3
Mx + Ny x 1
By Rule 2, the integrating factor is if
1 y2 y Mx − Ny
+ 3 dx − 2 dy = 0
x x x Mx − Ny = 0.

1 y2 Mx−Ny = xy + 1 yx− −xy + 1 yx = 2x2 y2 = 0
v= Mdx = + 3 dx
x x
y constant 1
Multiplying by the integrating factor = , the
y2 x2 y2
= log x − 2 equation can be written as
2x
∂v y ∂v dx dy d(xy) 1
= − 2; N − =0 − + 2 = 0 ⇒ log x/y = + c.
∂y x ∂y x y xy xy
Example
1.42 Solution
Solve x2 y2 +xy+1 ydx+ x2 y2 −xy + 1 xdy = 0. M = 3xy − 2y2 ; N = x2 − 2xy
∂M ∂N
Solution
The
equation
is of the form = 3x − 4y = 2x − 2y
∂y ∂x
f1 xy ydx + f2 xy xdy = 0
∂M ∂N
Mx−Ny = x3 y3 +x2 y2 +xy− x3 y3 −x2 y2 +xy =
∂y ∂x
= 2x2 y2 = 0 ∂M ∂N
∂y
− ∂x 3x − 4y − 2x − 2y
=
1 1 N x x − 2y
Integrating factor = = 2 2
Mx − Ny 2x y x − 2y 1
1 = = = p (x) , a function of x
Multiplying the equation by 2 2 and omitting the x x − 2y x
xy
constant, By Rule 3
1
1 1 1 1 Integrating factor = e p(x)dx = e x dx = elog x = x
1+ + 2 2 ydx + 1 − + 2 2 xdy = 0
xy x y xy x y Multiplying by integrating factor = x, the equation

1 1 becomes
1 + 2 2 (ydx + xdy) + (ydx − xdy) = 0 2
xy xy 3x y − 2xy2 dx + x3 − 2x2 y dy = 0

1 1 1 d x3 y = d x2 y2 ⇒ x3 y = x2 y2 + c
1 + 2 2 d(xy) + dx − dy = 0
xy x y
is the general solution of the differential equation.
1 x
1 + 2 d xy + d log =0
xy y Example 1.45

Solve xy3 + y dx + 2 x2 y2 + x + y dy = 0.
1 x
The general solution is xy − + log = c.
xy y Solution
Example
1.43
M = xy3 + y N = 2x2 y2 + 2x + 2y
Solve x − y dx − dy = 0. ∂M ∂N
= 3xy2 + 1 = 4xy2 + 2
∂y ∂x
Solution
1 ∂N ∂M xy2 + 1 1
M = x − y, N = −1, − = 2 = =q y ,
M ∂x ∂y xy + 1 y y
∂M ∂N
= −1, = 0;
∂y ∂x a function of y alone.
∂M
∂y
− ∂N
∂x −1 − 0 By Rule 4
= = 1 = p (x)
N −1 1
Integrating factor = e q(y)dy = e y dy
=y
(Constant can be considered as a function of x.)
By Rule 3
Multiplying by integrating factor, the equation can
Integrating factor = e p(x)dx = e 1dx = ex be written as
Multiplying by the integrating factor = ex , the 4
equation can be written as xy dx + 2x2 y3 dy + y2 dx + 2xydy + 2ydy = 0
1 2 4
xex dx = yex dx + ex dy ⇒ (x − 1) ex = yex + c d x y + d xy2 + d y2 = 0
2
or x = y + 1 + ce−x .
General solution is
Example
1.44 1 2 4
Solve 3xy − 2y dx + x2 − 2xy dy = 0.
2 x y + xy2 + y2 = c.
2
Example
1.46
The constants h, k are determined from
Solve y + xy2 dx − xdy = 0.
a+h+1 b+k +1
= and
Solution m n
c+h+1 d +k +1
M = y + xy2 ; N = −x =
p q
∂M ∂N
= 1 + 2xy; = −1 1+h+1 3+k +1
∂y ∂x ⇒ = and
1 2

1 ∂N ∂M 1 0+h+1 0+k +1
− = −1 − 1 − 2xy =
M ∂x ∂y y 1 + xy 3 5
⇒ 2h + 4 = k + 4 and 5h + 5 = 3k + 3
−2 1 + xy 2
= =− =q y ⇒ 2h = k and 5h + 2 = 3k (5)
y 1 + xy y
⇒ (h, k) = (2, 4) (6)
(xy = −1)
An integrating factor is x2 y4 . Multiplying Eq. (1)
By Rule 4, integrating factor is by this integrating factor, we have
2 1
e q(y)dy = e− y dy = e−2 log y = , x3 − y8 dx + 2x4 y7 dy
y2
a pure function of y. +3x2 y5 dx + 5x3 y4 dy = 0 (7)
1
Multiplying the equation by 2 we can write the which, on regrouping and expressing as exact differ-
y
2 ydx − xdy entials, yields
equation as d x + 2 =0
y2 1
whose general solution is d(x4 y8 ) + d(x3 y5 ) = 0
2x 4
x2 + = c. ⇒ x4 y8 + 4x3 y5 = c
y
⇒ x3 y5 (xy3 + 4) = c (8)
Example 1.47
Solve xy3 (ydx + 2xdy) + (3ydx + 5xdy) = 0 which is the required solution.
Solution The given differential equation is Example 1.48

Solve x(3ydx + 2xdy) + 8y4 (ydx + xdy) = 0.
xy (ydx + 2xdy) + (3ydx + 5xdy) = 0
3
(1)
Here M = xy4 + 3y N = 2x2 y3 + 5x Solution The given differential equation is
My = 4xy3 + 3 Nx = 4xy3 + 5 x(3ydx + 2xdy) + 8y4 (ydx + xdy) = 0 (1)
Since My = Nx , the equation is not exact. We can We observe that an integrating factor for this equa-
easily verify that Rules 1−4 are not applicable here. tion is of the form xh yk for some h, k. Multiplying
So, we try to find an I.F. of the form xh yk , by applying Eq. (1) by xh yk , it can be written in the form
Rule 5.
Mdx + Ndy = 0 (2)
Comparing Eq. (1) with the standard form
xa yb (mydx + nxdy) + xc yd (pydx + qydy) = 0 (2) where M = 3xh+1 yk+1 + 8xh yk+5 ,
We have N = 2xh+2 yk + 8xh+1 yk+4
a = 1, b = 3, m = 1, n = 2 ∴ My = 3(k + 1)xh+1 yk
c = 0, d = 0, p = 3, q = 5 (3) +8(k + 5)xh yk+4 (3)
Also, mp − nq = 1.3 − 2.5 = −7 = 0 (4) Nx = 2(h + 2)x y + 8(h + 1)xh yk+4
h+1 k
Exactness condition is which is the required solution.
My = Nx ⇒ 3k = 2h + 1, k + 4 = h Example 1.50
⇒ (h, k) = (13, 9) (4) Solve xy(ydx + xdy) + x2 y2 (2ydx − xdy) = 0.
Now the left-hand side expression of Eq. (1) is Solution Multiplying the equation by an inte-
exact grating factor xh yk , we obtain

3 8 Mdx + Ndy = xh+1 yk+2 + 2xh+2 yk+3 dx
⇒ Mdx + Ndy = d(x15 y10 ) + d(x14 y14 ) = 0
15 14 + xh+2 yk+1 − xh+3 yk+2 dy
⇒ x14 y10 (7x + 20y4 ) = c =0 (1)
which is the required solution. My = Nx
Example 1.49 ⇒ (k + 2)xh+1 yk+1 + 2(k + 3)xh+2 y(k+2)

Solve 2x2 (ydx + xdy) + y(ydx − xdy) = 0.
= (h + 2)x(h+1) y(k+1) − (h + 3)xh+2 yk+2
Solution The given differential equation is
Equating the coefficients of like powers
2x (ydx + xdy) + y(ydx − xdy) = 0
2
(1) k + 2 = h + 2, 2k + 6 = −h − 3
Here M = 2x y + y
2 2
N = 2x − xy
3
⇒ (h, k) = (−3, −3) (2)
∴ My = 2x + 2y Nx = 6x2 − y
2
Now Eq.(1) becomes

Since My = Nx , the equation is not exact.
To find an integrating factor of the form xh yk , we (x−2 y−1 + 2x−1 )dx + (x−1 y−2 − y−1 )dy = 0
multiply Eq. (1) by xh yk so that new values of −d(x−1 y−1 ) + d(2 log x) − log y = 0
k+2
M = 2xh+2 yk+1 + xhy and Integrating, we get
N = 2xh+3 yk − xh+1 yk+1 1 log x2
− + =c
My = 2(k + 1)xh+2 yk + (k + 2)xh yk+1 xy y
Nx = 2(h + 3)xh+2 yk − (h + 1)xh yk+1 the required general solution.
Equating My = Nx , we have EXERCISE 1.6
k = h + 2 k + 3 = −h Solve the following equations by finding integrating

(h, k) = (−5/2, −1/2) factors:

∴ Mdx + Ndy = 2x−1/2 y1/2 + x−5/2 y3/2 dx 1. (x2 y − 2xy2 )dx − (x3 − 3x2 y)dy = 0

+ 2x1/2 y−1/2 − x−3/2 y1/2 dy (Madras 1975, Karnataka 1971, Calcutta Hon.
=0 1975)
2 Ans: (x/y) + log y3 /x2 = c
⇒ 4d(x1/2 y1/2 ) − d(x−3/2 y3/2 ) = 0
3
2. (x2 + y2 )dx − 2xydy = 0
Integrating, we get Ans: x2 − y2 = cx
2
4x1/2 y1/2 − x−3/2 y3/2 = Constant 3. x2 ydx − (x3 + y3 )dy = 0
3
x3 / 3y3
⇒ 6x1/2 y1/2 − x−3/2 y3/2 = c Ans: y = ce
4. y(xy + 2x2 y2 )dx + x(xy − x2 y2 )dy = 0 which is taken as the standard form of the first order
(Rajasthan 1969, Kanpur 1974, Karnataka linear equation. Here the dependent variable y and its
1971, Punjab 1971) dy
1 derivative occur separately and to a first degree.
Ans: log(x2 /y) = +c dx
xy For example,
5. (x3 y3 + x2 y2 + xy + 1)ydx dy
(i) (1 + x) + 2xy = 3x2
+ (x3 y3 − x2 y2 − xy + 1)xdy = 0 dx
(Kanpur 1980, Gorakhpur 1972) dy
1 (ii) cos x + (sin x)y = tan x.
Ans: xy = c + + log y2 dx
xy
If Q(x) ≡ 0, then Eq. (1.17) is called a non-
6. (xy sin xy + cos xy)ydx homogeneous linear equation. If Q(x) ≡ 0, then it is
+ (xy sin xy − cos xy)xdy = 0 called a homogeneous or reduced linear equation.
(Gorakhpur 1974, Kanpur 1977, Lucknow Writing Eq. (1.17) in the form
1975, Rajasthan 1976, Marathawada 1974)
Ans: (x/y) sec xy = c Py − Q dx + dy = 0 (1.19)
7. (x2 + y2 + 2x)dx + 2ydy = 0 and comparing it with

(Srivenkateswara
1984, Calicut 1983)
Ans: ex x3 + y3 = c
Mdx + Ndy = 0 (1.20)
8. (y + log x)dx − xdy = 0 we have M = Py − Q, N = 1

(Marathwada 1994)
Ans: ex + y + log x + 1 = 0 Since
∂M ∂N
9. (3x2 y4 + 2xy)dx + (2x3 y3 − x2 )dy = 0 ∂y
− ∂x
(Calcutta Hon. 1952, 1954; Utkal 1980) = P (x) (1.21)
N
1
Ans: x3 y3 + 2
x a function of x alone e pdx is an integrating factor of
10. (xy3 + y)dx + 2(x2 y2 + x + y4 )dy = 0 Eq. (1.18).
Multiplying Eq. (1.18) by the integrating
Ans: 3x2 y4 + 6xy2 + 2y6 = c
factor e pdx , we have
dy pdx d
11. x(4ydx + 2xdy) + y3 (3ydx + 5xdy) = 0 e + ye pdx = ye pdx = Qe pdx
Ans: x4 y2 + x3 y3 = c dx dx
Separating the variables and integrating, we get
12. xy3 (ydx + 2xdy) + (3ydx + 5xdy) = 0
the general solution of Eq. (1.18) as
Ans: 14 x4 y8 + x3 y5 = c

1.3.6 Linear Equations ye pdx = c + Qe pdx dx (1.22)

A differential equation of the form where c is an arbitrary constant.
dy
p0 (x) + p1 (x) y = p2 (x) (p0 (x) = 0) (1.17)
dx Method of solving a linear equation
where p0 , p1 and p2 are continuous functions of x in
(1) Write the equation in the standard form
some interval I , is called a first order linear differen- dy
tial equation in y. Dividing Eq. (1.17) by p0 , we can + Py = Q.
dx
write it as
dy (2) Find the integrating factor e pdx and multiply
+ P (x) y = Q (x) (1.18) the equation by the integrating factor.
dx
(3) Write the general solution of the differen- which is linear in x.

tial
equation
as y integrating
factor = c +
1
Q integrating factor dx, after evaluating the Integrating factor = e y dy = elog y = y
integral.
1
x integrating factor = c +
y2
Note 1
Integrating factor dy
The words homogeneous and non-homogeneous
1
used here are not to be confused with similar words ⇒ xy = c + dy = c + log y.
used earlier. y
Note 2
Sometimes a linear equation may be written in the Example 1.53
form dy 1
Solve + y = ex + sin x.
dx x
dy
Po (x) + P (x) y = Q (x) (1.23)
dx Solution The equation is linear in y
where Po = P (1.24) 1
Here P (x) = , Q (x) = ex + sin x
x
so that Eq. (1.23) can be written as d(Po y) = Qdx
which, on integration, yields the general solution Integrating factor = e pdx
1
=e x dx
Po y = c + Qdx. (1.25) = elog x = x
Example 1.51 Multiplying the equation by x, we have

dy
Solve 1 + x2 + 2xy = cot x.
dx
dy
Solution This can be written as x + y = x (ex + sin x)
dx
d d
1 + x2 y = cot x ⇒ xy = x (ex + sin x)
dx dx
Integrating xy = c + x (ex − cos x)
Separating the variables and integrating = c + x (ex − cos x) − ex + cos x
xy = c + (x − 1) (ex − cos x) .
1 + x2 y = log sin x + c
where c is an arbitrary constant. Example 1.54

dy
Note 3 Solve x log x + y = 2 log x.
An equation which is not linear in y may be linear in dx
the variable x.
Solution Writing the equation in the standard
Example 1.52
form

Solve y2 dx + xy − 1 dy = 0. dy 1 2
+ y=
Solution Rewriting this equation as dx x log x x
dx 1 1 1 2
+ x = 2, Here P (x) = , Q (x) =
dy y y x log x x
This is a linear equation in y Example 1.56

dx −1
1 Solve 1 + y2 + x = etan y [JNTU 2001S].
Pdx = dx dy
x log x
Solution Writing the equation as
1
= d log x
log x
dx 1 1 −1
= log log x + x= etan y

dy 1 + y 2 1+y 2
Integrating factor = e Pdx = elog(log x) 1 1 −1

Here, P y = , Q y = etan y
= log x 1 + y2 1 + y2
The general solution is
The equation is linear in x
2
y log x = c + log xdx
x
1
P y dy = dy = tan−1 y
= c + 2 log xd log x 1 + y2

Integrating factor = e P(y)dy = etan y
−1
2
= c + log x

Hint: put log x = t,

2 1 1
log xdx dx = dt = 2tdt = t 2 xetan
−1 y
= c+
−1
e2 tan dy
x x 1 + y2
Example 1.55 1
dy Put tan−1 y = t, dy = dt
Solve x + 2y3 = y. 1 + y2
dx
= c + e2t dt
Solution This equation can be written as
1
dx dx 1 = c + e2t
y − x = 2y3 or − · x = 2y2 2
dy dy y −1 y 1 2 tan−1 y
xetan = c+ e .
1 2
Here P y = − , Q y = 2y2
y
The equation is linear in x Example 1.57

1 dy
P y dy = − dy Solve cos2 x + y = tan x. [JNTU 1999S]
y dx
1
= − log y = log Solution Writing the equation as
y
1
Integrating factor = e P(y)dy = elog y = dy
1
y + sec2 x y = tan x sec2 x

dx
The general solution is Here P (x) = sec2 x, Q (x) = tan x sec2 x

1 1 This is linear in y. Also,
x· = c + 2y2 · dy
y y
= c+y 2
P (x) dx = sec2 x dx = tan x
x
or = c + y2 Integrating factor = etan x
y
The general solution is Solution This is a linear equation in y with

1 log x
yetan x = c + tan x sec2 xetan x dx P (x) = , Q (x) =
x x
Put tan x = t, sec2 xdx = dt But, as it is, it can be written as

= c+ t
t e dt
d xy = log x dx

= c + (t − 1) et
log x dx = x log x − x
Integration by parts
y = ce− tan x + (tan x − 1) . Integrating, the general solution is
Example 1.58 xy = c + (x − 1) log x.

Solve xy + y + 4 = 0. [JNTU 2001]
Example 1.61
Solution Writing the equation in the form dy
Solve x (x − 1) − y = x2 (x − 1)3 .
dy 1 4 dx
+ y=− ,
dx x x Solution The equation can be put in the standard
we note that this is a linear equation in y with form
dy 1
1 4 − · y = x (x − 1)2
P(x) = , Q(x) = − dx x (x − 1)
x x This is linear in y with
But, as it is, the equation can be written as
1
d(xy) + 4dx = 0 P (x) = − , Q (x) = x (x − 1)2
x (x − 1)

1 1
which, on integration, yields the general solution Now, P (x) dx = − dx
x x−1
xy + 4x + c = 0.
x
= log
Example 1.59 x−1

dy
Solve x2 − 1 + 2xy = 1. [JNTU 1999] Integrating factor = e P(x)dx
dx
= elog x/(x−1)
x
Solution This is a linear equation in y with =
x−1
2x 1
P(x) = , Q(x) = The general solution is
x−1 x2 − 1

x x
But, as it is, the equation can be written as y = c + (x − 1)2 x dx
x−1 x−1
d 2
xy 3
(x − 1)y = 1 = c+ x − x2 dx
dx x−1
= c + x4 − x3 .
2 4 3
x − 1 y = x + c.
Example 1.62
dy 2 5x2
Example 1.60 Solve − y= .
dy dx x (2 + x) (3 − 2x)
Solve x + y = log x. [JNTU 1996S] [JNTU 2004 (Set 3)]
dx
Solution Divide by x2 dy
7. x + 2y = x2 log x.
dx
1 dy 2 5 Ans: 16x2 y + x4 1 − 4 log x = c
− 3y =
2
x dx x (2 + x) (3 − 2x)
dy
1 1 2 8. + y tan x = cos3 x.
⇒ d y =5 + dx dx
x2 2 + x 3 − 2x Ans: 4y sec x = 2x + sin 2x + c
y
⇒ 2 = 5 [log (2 + x) − log (3 − 2x)] + c.
x dy
9. 1 − x2 + 2xy = x 1 − x2 .
Example 1.63 dx
dy Ans: y = 1 − x2 + (1 − x2 )
Solve (x + 1) − ny = ex (x + 1)n+1 .
dx dy 2
[JNTU 2004 (Set 2)] 10. x 1 − x2 + 2x − 1 y = x3 .
dx
Solution Divide by (x + 1)n+1 Ans: y = x + cx 1 − x2
dy dx
(x + 1)−n − n (x + 1)−(n+1) y = ex 11. = sec y − x tan y.
dx dy
Ans: x sec y = tan y + c
⇒ d (x + 1)−n y = ex dx + c
12. yey dx = y3 + 2xey dy .
⇒ (x + 1)−n y = c + ex . x
Ans: 2 + e−y = c
y
EXERCISE 1.7
Solve the following: dy
13. + y cot x = 4x cosec x y (π/2) = 0.
dx
dy π2
1. x log x + y = x sin 2x. Ans: y sin x = 2x2 −
dx 2
[JNTU 2003, 2004]
1 dy
Ans: y log x = C − cos 2x 14. = y/ 2y log y + y − x .
2 dx c
Ans: x = + y log y
dy y
2. x cos x + y (x sin x + cos x) = 1.
dx
Ans: xy sec x = tan x + C 15. 1 − y2 dx = sin−1 y − x dy.
−1
Ans: x = sin−1 y − 1 + ce− sin y
dy
3. + 2y = ex (3 sin 2x + 2 cos 2x).
dx dy
Ans: y = Ce−2x + ex sin 2x 16. sin x cos x + y = xex cos2 x
dx
dy
4. y + y tan x = sin 2x, y (0) = 1. Hint: tan x + y sec2 x = xex ,
dx
[Hint :y sec x = sin2 x + c on multiplication by sec2 x

or y = cos x − cos3 x + c cos x
y (0) = 1 ⇒ 1 = c ⇒ d y tan x = xex dx
Ans: y = 2 cos x − cos3 x Ans: y tan x = c + (x − 1)ex

5. dx = x + y + 1 dy. 1.3.7 Bernoulli’s Equation
Ans: x + y + 2 = cey
An equation of the form
6. y + y cot
x = 2x cosec x. dy
Ans: y = x2 + c cosec x + Py = Qya (a ∈ R, a = 1) (1.26)
dx
is called a Bernoulli’s equation. Example 1.65

A Bernoulli’s equation is reducible to the linear Solve y(2xy + ex )dx = ex dy.
form by the substitution z = y1−a .
Multiply Eq. (1.26) by (1 − a)y−a , it becomes Solution
dy
− y = 2xe−x · y2
−a dy
(1−a) y +P (1 − a) y 1−a
= Q (1 − a) (1.27) dx
dx
dz This is Bernoulli’s equation.
or + P1 z = Q 1 (1.28)
dx
P = −1, Q = 2xe−x , a=2
P1 = (1 − a)P, Q1 = (1 − a)Q Multiplying the equation by − y12 , we have
z = y1−a , dz = (1 − a)y−a
1 dy 1
which is linear in z. Its general solution is − + = −2xe−x

y2 dx y
(1−a)Pdx
ze = y1−a e (1−a)Pdx 1 dy 1

− 2 + = −2xe−x
y dx y
= c + (1 − a) Qe (1−a)Pdx (1.29)
1
Note This is a linear equation in =z
y
There are also equations that are not of the Bernoulli Integrating factor = e 1·dx = ex
type shown in Eq. (1.26). These equations are The general solution is
reducible to linear form by appropriate substitution.
(See Example 1.67) 1 x
zex = e +c
y
Example 1.64
dy ⇒ 2xe−x ex dx = c − x2
Solve − y tan x = −y2 sec x.
dx
ex
Solution This is Bernoulli’s type equation. ⇒ = c − x2 .
y
P = − tan x, Q = − sec x, a = 2.
Example 1.66
Multiplying the equation by − y12 , we have dy
Solve xy 1 + xy2 = 1.
dx
1 dy 1
− + (tan x) = sec x Solution
y2 dx y dx
1 1 − y · x = y3 x2
Put = z ⇒ − 2 dy = dz dy
y y
dz This is linear in 1x .
+ (tan x)z = sec x
dx Multiplying by − x12 , we have
(This is a linear equation.)
Integrating factor 1 dx 1
− 2
+ y · = −y3
e Pdx = e tan x = elog sec x = sec x x dy x
1 dz
Put =z ⇒ + yz = −y3
The general solution is x dy
1 y2
z sec x = sec x = c + sec2 xdx = c + tan x. e ydy
= e+ 2
y
The general solution is EXERCISE 1.8

y2 1 y2 Solve the following:
ze 2 = e2
x dy
y2 1. x + y = x3 y6 .
= c− 3
y e dy 2 dx
[JNTU 1995, 2002, 2004]
y2 Ans: (5 + cx2 )x3 y5 = 2
Put = u; ydy = du
2 dy
2. xy 1 + xy2 = 1.
= c − 2ueu du dx
y2
= c − 2 (u − 1) eu Ans: 1x = 2 − y2 + ce− 2
dy y y 2
1 y2 y2 3. + log y = log y .
e 2 = c − y2 − 2 e 2 dx x x
x Ans: (log y)−1 = 1 + cx
1 y2
= ce− 2 + 2 − y2 . dy 2
x 4. 2xy = x + y2 + 1 .
dx
Example 1.67 Ans: y2 − x2 = cx − 1
dy
Solve + x sin 2y = x3 cos2 y.
dx 5. x x − y dy + y2 dx = 0.
Ans: x = log y + c
y
Solution This is not strictly in the form of
Bernoulli equation (1.26). But we can write it as
dy
6. tan y + tan x = cos y cos2 x.
dy dx
sec2 y + 2x tan y = x3 Ans: cos y = cos x (sin x + c)
dx
Put tan y = z ⇒ sec2 ydy = dz dy
dz 7. y = cos x + y2 (1 − sin x) cos x,
+ 2xz = x3 , dx
dx y (0) = 2.
Ans: 2 (tan x + sec x) = y (2 sin x + 1)
which is a Bernoulli type Eq. (1.26).

x2 8. y 2xy + ex dx = ex dy.
Integrating factor = e 2xdx
=e
Ans: ex = y c − x2
x2 x2
= tan ye
ze

2 dy
= c + x3 ex dx 9. tan y + tan x = cos y cos2 x.
dx
Ans: sec y = (c + sin x) cos x
1
= c+ tet dt
2 dy tan y
Put x2 = t ⇒ 2xdx = dt 10. − = (1 + x) ex sec y.
dx 1 + x
1 Ans: sin y = (1 + x) (ex + c)
= c + (t − 1) et
2
The general solution is 1.4 Applications of Ordinary
Differential Equations
2 1 2 2 Newton’s law of cooling
tan y ex = c + x − 1 ex
2
x2 1 2 Physical experiments show that the rate of change of
or tan y = ce− 2 + x −1 . temperature T with respect to time t, dT /dt, of a body
2
is proportional to the difference between the temper- If k is the constant of proportionality, then the
ature of the body (T ) and that of the surrounding required differential equation is
medium (T0 ).
dy
This principle is known as Newton’s Law of Cool- = ky (1.36)
ing and is expressed through the following first order dt
and first degree differential equation: where k is a real constant.
For growth, k > 0 and the differential equation is
dT
= −k(T − T0 ) (k > 0) (1.30)
dt dy
= ky (k > 0) (1.37)
Separating the variables dt
dT For decay, the differential equation is
= −kdt (1.31)
(T − T0 ) dy
= −ky (k > 0). (1.38)
Integrating, we get dt
log(T − T0 ) = −kt + log c Example 1.68
⇒ (T − T0 ) = ce−kt (1.32) The temperature of a body initially at 80°C reduces to
60°C in 12 min. If the temperature of the surrounding
If Ti is the initial temperature of the body when air is 30°C, find the temperature of the body after 24
t = 0, we get from Eq. (1.32) min.
Ti − T0 = c (1.33) Solution Let T be the temperature of the body
Eliminating c between Eqs. (1.32) and (1.33), we at time t.
get By Newton’s law of cooling
dT dT
(T − T0 ) = (Ti − T0 ) e−kt = −k (T − T0 ) ⇒ = −kdt
dt T − T0
or T (t) = T0 + (Ti − T0 ) e−kt . (1.34)
⇒ log (T − T0 ) = −kt + log c
Method of solving the problem of Newton’s ⇒ T − T0 = ce−kt (1)
law of cooling
Temperature of the surrounding medium T0 = 30
(1) Identify T0 , the temperature of the surrounding
medium. Then the general solution is given by Eq. T = T0 + ce−kt = 30 + ce−kt (2)
(1.34).
Initial temperature T = Ti = 80 when t = 0
(2) Use two given conditions and find the constant of
Ti − T0 = 80 − 30 = ce0
integration c and the proportionality constant k.
⇒ c = 50 (3)
(3) Substitute c and k obtained from step 2 in Eq. ∴ T = −kt
T0 + ce = 30 + 50e −kt
(1.34). We can determine (i) the value of T for
When t = 12, T = 60 ⇒ 60 = 30 + 50e−k·12
a given time t or (ii) the value of t for a given
temperature T from Eq. (1.34). 1 5
⇒k = log
12 3
Law of natural growth or decay − 1 log 5 ·24
When t = 24, T = 30 + 50e 12 3
If the rate of change of a quantity y at any time t is 2

3
proportional to y, then = 30 + 50
5
dy 9
∝y (1.35) = 30 + 50 × = 48.
dt 25
Example 1.69 be solved is

A body is heated to 105°C and placed in air at dN dN
15°C. After 1 hr its temperature is 60°C. How much = kN ⇒ = kdt
dt N
additional time is required for it to cool to 37 12 °C?
⇒ log N = kt + log c ⇒ N = cekt (1)
Solution Let T be the temperature of the body When t = 0, N = 100
at time t.
⇒ 100 = ce0 ⇒ N = 100 ekt
By Newton’s law of cooling
t = 1hr = 3600 sec, N = 332 = 100ekt (2)
dT 332
= −k (T − T0 ) ⇒ ek.1 = (3)
dt 100
dT 3
⇒ = −kdt 3 3 332 2
T − T0 When t = hr N = 100 ek 2 = 100
⇒ log (T − T0 ) = −kt + log c 2 100
⇒ T − T0 = ce−kt N2 332 3
(1) ⇒ =
10000 100
Temperature of the surrounding medium T0 = 15 ⇒ (10N ) = 3323 ⇒ N = 604.9.
2
T = T0 + ce−kt = 15 + ce−kt (2) Example 1.71

A radioactive substance disintegrates at a rate pro-
Initial temperature (t = 0) portional to its mass. When the mass is 10 mg the
rate of disintegration is 0.051 mg per day. How long
T = Ti = 105 ⇒ 105 = 15 + ce0 will it take for the mass of 10 mg to reduce to its half.
⇒ c = 90 (3) [JNTU 1995]
When t = 1, T = 60 ⇒ 60 = 15 + 90e−k·1 Solution The governing differential equation is
1 dm
⇒ e−k = = −km (k > 0) (1)
2 dt
1 1
When T = 37 , 37 = 15 + 90e−kt where m is the mass of the substance.
2 2 Separating variables and integrating
t
= 15 + 90
1 log m = kt + log c ⇒ m = ce−kt (2)
2
t When mass m = 10 mg
22 12 1 dm
⇒ = ⇒ t=2 = −0.051
90 2 dt
dm
Additional time required = 2 hr −1 hr = 1 hr. negative sign is to be taken since is decreasing
dt
rate.
Example 1.70
The number N of bacteria in a culture grew at a rate ∴ −0.051 = −k × 10
proportional to N . The value of N was initially 100 0.051
⇒k = = 0.0051 (3)
and increased to 332 in 1 hr What was the value of N 10
3
after hrs. [JNTU 1996S, 2003] Eq. (2) becomes
2
m = ce−(0.0051)t (4)
Solution Here N is a natural number and is there-
When t = 0 m = 10 mg
fore discrete. But in view of its largeness N will be
treated as a continuous variable which is a differen- ∴ 10 = ce−(0.0051)10 ⇒ c = 10 (5)
tiable function of time t. The differential equation to ∴ m = 10e−(0.0051)t
We have to find t when m = 5 mg A family of curves which are orthogonal to them-

5 = 10e−(0.0051)t selves are called self-orthogonal.
⇒ e(0.0051)t = 2 ⇒ 0.0051 t = ln 2 (1) In the electrical field, the paths along which the
0.6931 current flows are the orthogonal trajectories of
⇒ t= = 135.9 days.
0.0051 the equipotential curves.
(2) In fluid mechanics, the stream lines and the
EXERCISE 1.9 equipotential lines are orthogonal trajectories of
1. A body initially at 80°C cools down to 60ºC in 20 one another.
min. The temperature of the air is 40°C. Find the
temperature of the body after 40 min. (3) In thermodynamics, the lines of heat flow are
Ans: 50°C perpendicular to isothermal curves.
2. The air temperature is 20°C. A body cools from Method for f inding orthogonal trajectories
140°C to 80°C in 20 min. How much time will it
take to reach a temperature of 35°C? (1) Cartesian coordinates
Ans: 60 min
Let f (x, y, c) = 0 (1.39)
3. If the temperature of the air is 20°C and the tem-
perature of a body drops from 100°C to 75°C in represent the equation of a given family of curves
10 min. what will be its temperature after 30 min? with single parameter ‘c’.
When will the temperature be 30°C? Differentiating Eq. (1.39) with respect to ‘x’ and
Ans: 46°C, 55.5 min eliminating ‘c’ from the equation thus obtained and
Eq. (1.39) we obtain a differential equation of the
4. Uranium disintegrates at a rate proportional to the form
amount present at any instant. If m1 and m2 gms of dy
φ x, y, =0 (1.40)
uranium are present at time t1 and t2 , respec- dx
tively. Show that the half-life of uranium is
[(t2 − t1 ) log 2] / log (m1 /m2 ) . for the given family of curves.
Suppose there passes a curve of the given family
5. The rate at which bacteria multiply is proportional and a member of the orthogonal trajectories through
to the instantaneous number present. If the original a point P(x, y). Let m1 be the slope of the curve of
number doubles in 2 hrs, in how many hours will the given family and m2 the slope of a orthogonal
it triple? [JNTU 1987, 2000] trajectory. Since the curves cut at right angles we
2 log 3 dy
Ans: hrs = 3.17 hrs have m1 m2 = −1 and m1 = so that
log 2 dx
6. The rate of decay of radium varies as its mass 1 dx
at a given time. Given that half-life of radium is m2 = − dy = −
dx
dy
1600 yrs, find out the percentage of the mass of
radium it will disintegrate in 200 yrs. dy
Therefore, if we replace in Eq. (1.40) by
Ans: 8.3% approximately dx
dx
− , then we get the differential equation for the
1.4.1 Geometrical Applications dy
orthogonal trajectories
as
Orthogonal trajectories of a family of curves dx
φ x, y, − =0 (1.41)
dy
Definition A curve which cuts every member of
a given family of curves at a right angle is called an Integrating Eq. (1.41), we obtain the equation for
orthogonal trajectory of the given family of curves. the orthogonal trajectories.

Example 1.72 dx dy
x − = 2y or 2y +x =0 (4)
Find the orthogonal trajectories of the rectangular dy dx
hyperbolas x2 − y2 = c where c is a parameter.
Separating the variables and integrating, we get
Solution Equation of the given curves the equation for orthogonal trajectories as
x2 − y2 = c (1) 2y2 + x2 = c
Differentiating Eq. (1) with respect to x, we get the where ‘c’ is a constant.
differential equation of the given curves as
Example 1.74
dy Find the orthogonal trajectories of the family of semi-
x=y (2) cubical parabolas ay2 = x3 where ‘a’ is a parameter.
dx
dy dx Solution The equation of the given curves is
Replacing in Eq. (2) by − the differential
dx dy ay2 = x3 (1)
equation for the orthogonal trajectories is obtained
as Differentiating Eq. (1) with respect to ‘x’, we get

dx dx dy
x−y − =0⇒x+y =0 (3) 2ay = 3x2 (2)
dy dy dx
Separating the variables, we get Eliminating ‘a’ between Eqs. (1) and (2), the
dx dy differential equation is obtained as
+ =0 dy
x y 2x = 3y (3)
dx
Integrating, the equation for orthogonal trajecto-
dy dx
ries is Replacing in Eq. (3) by − , the differential
dx dy
log x + log y = log k or xy = k. (4) equation for the orthogonal trajectories is obtained
as
dx dy
Example 1.73 3y = 2x − or 3y + 2x = 0 (4)
dy dx
Find the orthogonal trajectories of the family of
parabolas through the origin and focii on the y-axis. Separating the variables and integrating, we obtain
the equation for orthogonal trajectories as
Solution Equation of the family of parabolas
3y2 − 2x2 = c
x = 4ay
2
(1)
Differentiating Eq. (1) with respect to x, where ‘c’ is an arbitrary constant.
dy Example 1.75
2x = 4a (2)
dx Show that the system of confocal ellipses
x2 y2
Eliminating ‘a’ between Eqs. (1) and (2), the + 2 = 1 (λ, parameter) is
differential equation for the given curves is a +λ b +λ
2
dy dy self-orthogonal.
x2 = 2xy or x = 2y (3)
dx dx
Solution The equation of the system of ellipses
dy dx is
Replacing in Eq. (3) by − , the differential
dx dy x2 y2
equation for orthogonal trajectories is + =1 (1)
a2 + λ b2 + λ
Differentiating Eq. (1) with respect to ‘x’ (2) Polar coordinates
2x 2yy Let f (r, θ, a) = 0 (1.42)

+ =0 (2)
a2 + λ b2 + λ
be the equation of the given system of curves in polar
⇒ λ x + yy + a2 yy + b2 x = 0 coordinates where ‘a’ is a parameter.

− a2 yy + b2 x dy Differentiating Eq. (1.42) with respect to ‘θ’ we get
⇒λ= y = another equation. Between these two equations, we
x + yy dx
eliminate the parameter ‘a’and obtain the differential
2 ⎫ equation for the given system of curves as
a2 yy + b2 x a − b2 x ⎪
⎪
a + λ=a −
2 2
= ⎪
⎬ dr
x + yy x + yy F r, θ, =0 (1.43)
2 (3) dθ
a2 yy + b2 x a − b2 yy ⎪
⎪
b + λ=b −
2 2
= ⎪
⎭
x + yy x + yy dr dθ
If we replace in Eq. (1.43) by −r 2 (which
dθ dr
Eliminating λ from Eq. (1) using Eq.(3) amounts to interchanging the role of the polar
subnormal and the polar sub-tangent), we get the
x + yy x2 x + yy y2 differential equation for the orthogonal trajectories
− =1 as
a2 − b2 x a2 − b2 yy

dθ
y F r, θ, −r 2 =0 (1.44)
⇒ x + yy x − = a2 − b2 dr
y
Solving this differential equation, we get the

1 equation for the orthogonal trajectories.
It is clear that if we replace y by − , the same
y The following solved examples illustrate the proce-
equation is obtained. dure.

Show that the system of confocal and coaxial Find the orthogonal trajectories of the family of
cardioids r = a(1 − cos θ) where ‘a’ is a parameter.
parabolas y2 = 4a(x + a) is self-orthogonal.
Solution The equation of the given cardioids is
Solution The equation of the given parabolas
r = a (1 − cos θ) (1)
y2 = 4a(x + a) (1)
Differentiating Eq. (1) with respect to ‘θ’, we have
Differentiating Eq. (1) with respect to ‘x’, we have
dr
= a sin θ (2)
dy dθ
2yy1 = 4a y1 = (2)
dx Eliminating ‘a’ between Eqs. (1) and (2), we get
r 1 − cos θ
Eliminating ‘a’ from Eqs. (1) and (2), we get dr =
dθ
sin θ
yy1
y2 = 2yy1 x + or y = 2xy1 + yy12 (3) 2 sin2 θ
θ
2 = θ
2
θ
= tan
2 sin cos
2 2
2
1
Replacing y1 by − in the above differential The differential equation for the given curves
y1
equation for the given curves, we obtain the same dr θ
equation for the orthogonal trajectories. = r cot
dθ 2
dr −r 2 dθ
Replacing by we obtain from Eq. (3) n dr n cos nθ
dθ dr + =0
the differential equation for orthogonal trajectories r dθ sin nθ
as
−r 2 dθ θ 1 dr
= r cot (3) ⇒ + cot nθ = 0 (2)
dr 2 r dθ
dr θ
⇒ = − tan dθ which is the differential equation for the given curves.
r 2
Integrating we get (4) dr dθ
Replacing by −r 2 in Eq. (2), we get the
θ dθ dr
log r = 2 log cos + log 2c differential equation for the orthogonal system as
2
θ 1 dθ cos nθ
⇒ r = 2c cos2 = c (1 + cos θ) (5) − · r2 + =0
2 r dr sin nθ
which is the equation for the orthogonal trajectories. n sin nθ dr
⇒ =n (3)
cos nθ r
Example 1.78
By integrating, we obtain the equation for
Find the equation of the system of orthogonal
orthogonal trajectories as
trajectories of the family of curves r n sin nθ = an
where an is a parameter. log r n = − log cos nθ + n log c (4)
Solution The given curves are ⇒ r cos nθ = cn .
n
(5)
r n sin nθ = an (1)
EXERCISE 1.10
By logarithmic differentiation of Eq. (1) with Find the orthogonal trajectories of the family
respect to ‘θ’ of curves given in Table 1.2
Table 1.2
S. No. Curves Parameter Orthogonal trajectories
Cartesian form
1. y2 = 4ax (parabolas) a 2x2 + y2 = c (ellipses)
2. x2 + y2 + 2gx + c = 0 g x2 + y2 + 2fy − c = 0
(coaxial circles) (circles)
2 2 2 4 4 4
3. x3 + y3 = a3 (astroids) a x3 − y3 = c3
4. x2 − y2 = cx c y(y2 + 3x2 ) = a
Polar form

θ2
5. rθ = a a r 2 = c exp
2
2a
6. r= a r(1 − cos θ) = 2c
(1 + cos θ)
7. r n = an cos nθ a r n = cn sin nθ
8. r n cos nθ = an a r n sin nθ = cn
Linear Differential
Equations of Second
and Higher Order
2
2.1 Introduction Example 2.1
dy
In the previous chapter, we have studied the linear Solve + y = 2ex (1)
differential equation of the first order dx
Solution Clearly u(x) = ce−x satisfies the
dy dy
+ P(x)y = Q(x) (2.1) homogeneous equation + y = 0 derived from
dx dx
Eq. (1).
It is called a homogeneous (or reduced) equation
The general (complete) solution of Eq. (1) is
if Q(x) ≡ 0 and a non-homogeneous equation if
Q(x) ≡ 0. y = ce−x + ex
If we substitute y = u + v in Eq. (2.1), we have Thus, the general solution y of a linear equation
d comprises of two parts: y = yc + yp where yc , which
(u + v) + P(x)(u + v) = Q(x) satisfies the homogeneous equation and contains
dx as many arbitrary constants as the order of the
du dv
⇒ + P(x)u + + P(x)v − Q(x) = 0 differential equation, and yp , which satisfies the non-
dx dx homogeneous equation and contains no arbitrary
which is satisfied if constants.
du 2.1.1 Linear Differential Equations

+ P(x)u = 0 (2.2) of the Second Order
dx
dv Second order linear differential equations have many
+ P(x)v = Q(x) (2.3)
dx applications in physics and engineering such as
in motion of mechanical systems and fundamental
This shows that y = u + v is the general solution
electric circuits. They also lead to the study
(or complete solution) of Eq. (2.1) if u and v satisfy
of higher functions such as Bessel’s, Legendre’s,
Eqs. (2.2) and (2.3), respectively, i.e., if
and hypergeometric functions, which have wide

u(x) = ce− Pdx
(2.4) applications.

A second order differential equation is called
v(x) = e− Pdx
Qe Pdx
dx (2.5) linear if it can be written in the standard form
d 2y dy
The function u(x) which contains an arbitrary 2
+ P(x) + Q (x) y = R (x)
dx dx
constant c is called the complementary function
or D2 + P (x) D + Q (x) y = R (x) (2.6)
(C.F.) and v(x) which contains no such arbitrary
constant is called the particular integral (P. I.) of d d2
Eq. (2.1). There is one arbitrary constant in the C.F., where D ≡ , D2 ≡ are the differential
dx dx2
because the order of the differential equation is 1. operators. When applied on a differentiable function,
“C
‘D’ does the operation of differentiation and so it 5. y − y = 0.

must always be written on the left-hand side of the
function upon which it is operating. 6. y + y = tan x + sec2 x.
1. ‘D’ is a linear operator: A solution of Eq. (2.6) is a twice differential

function y on some open interval I , which satisfies
D c1 y1 (x) + c2 y2 (x) = c1 Dy1 (x)+c2 Dy2 (x) Eq. (2.6) identically.
As in the case of the first order linear differential
2. ‘D’ satisfies index laws: equation, the general solution of Eq. (2.6) consists of
two parts yc and yp .
D2 = D · D, D3 = D · D · D If u(x) and v(x) satisfy Eqs. (2.7) and (2.6),
D · Dn = Dn Dm = Dm+n
m
respectively,
0
D = 1; , D0 f (x)
u (x) + P(x)u (x) + Q(x)u(x) = 0 (2.8)
= 1 · f (x) = f (x)
v (x) + P(x)v (x) + Q(x)v(x) = R(x) (2.9)
Note
dy then their sum y = u(x) + v(x) satisfies Eq. (2.6).
Equation + P (x) y = Q (x) , in operator notation, u(x) is called the complementary function (C.F.) and
dx
is v(x) is called the particular integral (P. I.) of Eq. (2.6).
(D + P) y = Q (x)

Example 2.2
− Pdx
⇔y=e Qe Pdx
dx
Solve y + y = 2ex (1)
This gives a meaning to the operator D + P. Solution Clearly ex is a solution of the complete
A differential equation that cannot be written in Eq. (1). Now consider the reduced equation
the form Eq. (2.6) is called non-linear (see Examples
y + y = 0 (2)
2.75 and 2.76)
If R(x) ≡ 0 then Eq. (2.6) becomes
1 and e−x are two solutions of Eq. (2). A linear
d 2y dy combination of these solutions namely y = c1 +c2 e−x
2
+ P (x) + Q (x) y = 0 is also a solution of Eq. (2). The general solution of
dx
2 dx Eq. (2) is
(or) D + P (x) D + Q (x) y = 0 (2.7)
y = c1 + c2 e−x + ex (3)
and is called a homogeneous (or reduced)
equation (illustrative Examples (2) (3)
and (5) below) and if R(x) ≡ 0 then 2.1.2 Homogeneous Equations—
Eq. (2.6) is called a non-homogeneous equation Superposition or Linearity
(Illustrative examples (1), (4) and (6) below.) Principle
Here P, Q, R are continuous functions of x on some We begin our discussion of solution of a second order
interval I , which is to be understood even if we do linear equation with the homogeneous equation. Let
not mention it in each case, hereafter. us consider an example.
1. y + 4y = e−x sin x.
Example 2.3
2.
(1 − x )y − 2xy + 6y = 0.
2 Solve y − y = 0 or (D2 − 1)y = 0 (1)
3. y + y2 = 0. Solution y1 = ex and y2 = e−x are solutions of

the homogeneous linear differential equation for all
4. y = (y2 + 1). x since
“C
Linear Differential Equations of Second and Higher Order 2-3
(D2 − 1)ex = D2 ex − 1 · ex Non-homogeneous linear differential

= ex − ex = 0 equation
(D − 1)e = D2 e−x − 1 · e−x

2 −x
Consider the non-homogeneous differential equa-
= (−1)2 e−x − e−x = 0 tion
In fact, a linear combination of y1 and y2 , i.e., (D2 − 1)y = 1 (2.10)

y = c1 y1 + c2 y2 is also a solution of Eq. (1) since −x
y1 = e − 1 and y2 = e − 1
x
are solutions of Eq. (2.10) since

(D2 − 1)y = (D2 − 1)(c1 y1 + c2 y2 )
(D2 − 1)y1 = (D2 − 1)(ex − 1)
= (D2 − 1) c1 ex + c2 e−x
= D2 (ex − 1) − 1(ex − 1)
= D2 c1 ex + c2 e−x − c1 ex + c2 e−x
= ex − ex + 1 = 1
= c1 (D2 ex ) + c2 (D2 e−x )
(D2 − 1)y2 = (D2 − 1)(e−x − 1)
− c1 ex + c2 e−x
= D2 (e−x − 1) − 1(e−x − 1)
= c1 ex + c2 (−1)2 e−x
−x
= e−x − e−x + 1 = 1
−c1 e − c2 e
x
=0 (2)
But their sum y1 + y2 = (ex − 1) + (e−x − 1) =
This is known as the superposition or linearity e + e−x − 2 is not a solution.
x
principle.
(D2 − 1)(y1 + y2 ) = D2 (ex + e−x − 2)
−1(ex + e−x − 2)
2.1.3 Fundamental Theorem for
the Homogeneous Equation = ex + e−x
−(ex + e−x − 2) = 1
Theorem 2.1 For a homogeneous linear differen- Also, 2y1 = 2(ex − 1)
tial Eq. (2.7), any linear combination of two solutions
= 2ex − 2 is not a solution
is again a solution.
(D2 − 1)(2y1 ) = 2(D2 − 1)(ex − 1)
Proof: Let y1 and y2 be solutions of Eq. (2.7). Then = 2[D2 (ex − 1) − 1(ex − 1)]
substituting = 2[ex − ex + 1]
y = c1 y1 + c2 y2
= 2 = 1.
(D2 + P(x)D + Q(x))y Non-linear differential equation

= (D2 + P(x)D + Q(x))(c1 y1 + c2 y2 )
y1 = x2 , y2 = 1 are solutions of the non-linear
= D2 (c1 y1 + c1 y2 ) + P(x) differential equation
D(c1 y1 + c2 y2 ) + Q(x)(c1 y1 + c2 y2 )
= c1 [D2 y1 + P(x)Dy1 + Q(x)y1 ] yy − xy = 0 (2.11)
+c2 [D2 y2 + P(x)y2 + Q(x)y2 ] but (−1)y2 = −x2 and their sum y1 + y2 = x2 + 1
= c1 .0 + c2 .0 = 0 are solutions of Eq. (2.11).
[∵ y1 , y2 are solutions of Eq.(2)]
Note 2.1.4 Initial Value Problem (IVP)

The above principle does not hold for non- For a first-order differential equation, a general solu-
homogeneous equations and non-linear equations. tion contains one arbitrary constant c. If an initial
“C
Solution
condition y(x0 ) = y0 is given, then a particular
cos x, sin x are solutions. We take y = c1 cos x +
solution in which c will have a definite value is
c2 sin x
obtained.
Differentiating with respect to x
Example 2.4 y = −c1 sin x + c2 cos x (2)
Solve the initial value problem
y (0) = c1 · 1 + c2 · 0 = 2 ⇒ c1 = 2
1 1 y (0) = −c1 · 0 + c2 · 1 = −1 ⇒ c2 = −1
y + y = 2 , y (1) = 0 (1)
x x
Now the required particular solution is
Solution This is a linear equation of the first
order with y = 2 cos x − sin x (3)
1 1
P= ; Q = 2;
x
x
1 2.1.5 Linear Dependence and
Pdx = dx = log x Linear Independence of
x
Solutions
Integrating factor = elog x = x
Two solutions y1 (x) and y2 (x) defined on some inter-
Multiplying Eq. (1) by the integrating factor ‘x’ val I are said to be linearly independent (L.I.) on I
if
1 1
xy + y = ⇒ d xy = dx k1 y1 (x) + k2 y2 (x) = 0 (2.12)
x x
⇒ k1 = k2 = 0
Integrating and applying the initial condition They are said to be linearly dependent (L.D.) if Eq.
xy = log x + c (2.12) holds, for some non-zero constants k1 and k2 .
1.0 = log 1 + c y1 and y2 are linearly independent if one is not
⇒c=0 proportional to the other.
The required solution is
xy = log x Criterion for linear independence of two
functions y1 and y2 on I
For a second-order homogeneous linear Eq. (2.7),
Two functions y1 and y2 defined on some interval
a general solution is of the form
I are linearly independent on I if the Wronskian
y = c 1 y1 + c 2 y2 (determinant)
involving two arbitrary constants c1 and c2 . An ini- y1 y2

tial value problem now consists of Eq. (2.7) and two W y1 , y2 (x) = = 0 (2.13)
y1 y2
initial conditions
y (x0 ) = y0 , y (x0 ) = y0
Example 2.6
We have to find the particular solution satisfying Show that y1 = cos x and y2 = sin x are linearly
these conditions as illustrated below. independent
Example 2.5 cos x sin x

Solve the initial value problem W (cos x, sin x) =
− sin x cos x

y + y = 0, y (0) = 2, y (0) = −1 (1) = cos x + sin2 x = 1 = 0.
2
“C
Example 2.7 We have Dy = memx , D2 y = m2 emx .

π Substituting these in Eq. (2.14), we have
Show that y1 = cos (π + x) and y2 = sin +x
2
are linearly dependent (am2 + bm + c)emx = 0

cos (π + x) sin π2 + x ⇒ f (m) = am2 + bm + c = 0 (2.17)
W y1 , y2 = ∵ emx = 0. (2.18)
− sin (π + x) cos π2 + x
− cos x cos x Hence Eq. (2.15) is a solution of Eq. (2.16) if m is
= = 0. a solution of the algebraic Eq. (2.16) which is called
sin x − sin x
the auxiliary equation or characteristic equation of
Eq. (2.14).
2.1.6 General Solution, Basis and It has, in general, two roots
Particular Solution √ √
A general solution of a second-order linear −b+ b2 − 4ac −b− b2 − 4ac
(α, β) = ,
homogeneous (L.H.) equation 2a 2a
y + P (x) y + Q (x) y = 0 (2.19)
on an open interval I is of the form y = c1 y1 + c2 y2 Note

where y1 and y2 are linearly independent on I (they Auxiliary equation f (m) = am2 + bm + c = 0
are not proportional to each other) and c1 and c2 are is obtained from the homogeneous differential Eq.
arbitrary constants. Then, y1 and y2 are called the (2.14) by replacing ‘D’ by m. The general solution
basis (fundamental system) of solution for the L.H. of Eq. (2.14) is obtained by combining two linearly
equation. independent solutions of Eq. (2.14) on I , which
A particular solution of L.H. equation is obtained constitute a basis of solutions of Eq. (2.14) on I .
by giving specific values to c1 and c2 in the general Case (i) Two distinct real roots α, β.
solution.
In this case, y1 = eαx , y2 = eβx constitute a basis of
solutions of Eq. (2.14) on any interval I . Since these
2.1.7 Second Order Linear are linearly independent, the general solution is
Homogeneous Equations with
Constant Coef f icients y = c1 eαx + c2 eβx . (2.20)
We will now discuss the method of solution of a Case (ii) Two real repeated (double) roots α, α
homogeneous linear differential equation of the form Let β = α. The solution is
ay + by + cy = 0 (2.14)
y = c1 eαx + c2 eαx = (c1 + c2 ) eαx .
or f (D) y = 0 (2.15)
where f (D) = aD2 + bD + c So, we have to find another independent solution
for basis. We proceed as follows:
To solve Eq. (2.14), we recall that a first-order
linear equation Method 1
Let β = α + h. The solution is
y − ay = (D − a) y = 0
y = c1 eαx + c2 e(α+h)x
has an exponential function as solution: y = eax . This
hx (hx)2
gives us the idea to try as a solution of Eq. (2.14), the = eαx c1 + c2 1+ + + ···
function 1! 2!
y = emx (2.16) expanding ehx by exponential theorem.

= eαx [(c1 + c2 ) + (c2 h) x] Example 2.8

d 2y dy
assuming that terms with h2 and higher powers of h Solve 2
+ 3 + 2y = 0.
dx dx
tend to zero.
We can choose c2 to be sufficiently large so as Solution The given differential equation is
to make hc2 finite as h → 0 and c1 large with a (D2 + 3D + 2)y = 0
sign opposite to that of c2 so that c1 + c2 is finite.
If c1 + c2 = A and hc2 = B, the general solution The auxiliary equation is
corresponding to two equal roots is y = (A + Bx)eαx . m2 + 3m + 2 = (m + 1) (m + 2)
= 0 ⇒ m = −1, −2
Method 2
If the auxiliary equation am2 + bm + c = 0 has roots The general solution is y = c1 e−x + c2 e−2x .
equal to α Example 2.9
am + bm + c = a (m − α)
2 2 d 2y dy
Solve 2
+ 1.5 + 0.5y = 0.
dx dx
The differential equation becomes (D − α)2 y = 0
Put (D − α) y =z (2D2 + 3D + 1)y = 0
Then, (D − α) z =0
The auxiliary equation is
dz αx
(i.e.) − αz = 0 ⇒ z = Be 2m2 + 3m + 1 = (2m + 1) (m + 1) = 0
dx
Hence, (D − α) y = Beαx ⇒ m = −1/2, −1
dy The general solution is
or − αy = Be−αx
dx y = c1 e−x/2 + c2 e− x.
which is a linear equation
Example 2.10
and integrating factor = e−αx
Solve (9D2 + 6D + 1)y = 0.
−αx
Its solution ye = Beαx · e−αx dx
Solution The auxiliary equation is
= Bx + A
9m2 + 6m + 1 = (3m + 1)2 = 0
⇒ y = (A + Bx) eαx .
⇒ m = −1/3. − 1/3
Case (iii) Since a, b, c are real, complex roots occur
in conjugate pairs if the discriminant of the auxiliary The general solution is
equation y = (c1 + c2 x) e−x/3 .
f (m) = am2 + bm + c = 0 is negative. Example 2.11
d 2x dx
The general solution in this case is Solve 2
+k + k 2 x = 0.
dt dt
y = c1 e(α+iβ)x + c2 e(α−iβx)
= eαx c1 eiβx + c2 e−iβx
d
= eαx [c1 (cos βx + i sin βx) (D + kD + k )x = 0
2 2
D≡
dt
+c2 (cos βx − i sin βx)]
αx
= e (A cos βx + B sin βx)
m2 + km + k 2 = 0
where A = c1 + c2 ; B = (c1 − c2 ) i √
k 3ki
are arbitrary constants. ⇒m=− ±
2 2
The general solution is 2.1.8 Higher Order Linear

√ √ Equations
3kt 3kt −kt/2 The method of solution of second order linear
x = c1 cos + c2 sin e .
2 2 equations can be extended to higher order equations.
Example 2.12
The differential equation

d d ny d n−1y
Solve (D2 + a2 )y = 0 D= P0 n + P1 n−1 + · · · + Pn y = X (2.21)
dx dx dx
Solution The auxiliary equation is is called an nth order differential equation with
variable coefficients if at least one of the coefficients
m2 + a2 = 0 ⇒ m = ±ia Pi (i = 0, 1, · · · , n) is a function of x and Eq. (2.19)
is called an equation with constant coefficients if all
The general solution is Pi (i = 0, 1, · · · , n) are constants. The right-hand
side member X in Eq. (2.19) is a function of x.
y = c1 cos ax + c2 sin ax. Equation (2.19) is homogeneous if X ≡ 0 and
EXERCISE 2.1 non-homogeneous if X ≡ 0 (Exercises 1, 3 and
4 below). In each term of Eq. (2.19), y and its
Solve the following: derivatives occur in the first degree without being
multiplied together.
1. y + y − 6y = 0.
Ans: y = c1 e2x + c3 e−3x Table 2.1 Classification of Linear Differential
Equations
2. 8y − 2y − y = 0. S.No. Equation Type Coefficients
Ans: y = c1 ex/2 + c2 e−x/4
1. yv + y = ex NH constants

3. 9D2 + 12D + 4 y = 0.
2. y −6y +11y−6y = 0 H constants
Ans: y = (c1 + c2 x) e−2x/3
1 1
3. y + y + 2 y = xex NH variables
x x
4. 4y + 4y + y = 0.
Ans: y = (c1 + c2 x) e−x/2 4. (x+1)2 y −3(x+1)y NH variables
+4y = x2 + x + 1
5. y + π2 y = 0.
1
Ans: y = c1 cos πx + c2 sin πx 5. y − y=0 H variables
x2

6. y + 2ky + k 2 + 4 y = 0. Bessel Equation

Ans: y = e−kx (c1 cos 2x + c2 sin 2x) 1 x2
6. (xy )+ 1− 2 y = 0 H variables
2
x x
d y dy
7. + (a + b) + aby = 0. Legendre’s Equation
dx2 dx
Ans: y = c1 e−ax + c2 e−bx 7. [(1 − x2 )y ] H variables
+x(x + 1)y = 0
d 2y dy
8. + 8 + 16y = 0.
dx2 dx As in the case of linear equations of the first
Ans: y = (c1 + c2 x) e−4x order, the general solution (or complete solution)
of Eq. (2.19) comprises of two parts: y = yc +
9. D2 − 3D + 4 y =√0. [JNTU
√ 2003]
yp where yp , called the particular integral (P. I.)
Ans: y = e3x/2 c1 cos 7x/2 + c2 sin 7x/2 satisfies the complete non-homogeneous (N.H.)
Eq. (2.19) and contains no arbitrary constants, When applied on a differentiable function, D does
while yc , called the complementary function (C.F.) the operation of differentiation; so it must be always
(a linear combination of n linearly independent on the left-hand side of the function upon which we
solutions containing n arbitrary constants) satisfies are applying it. It is a linear operator.
the homogeneous equation.
D c1 y1 + c2 y2 = c1 Dy1 + c2 Dy2
2.1.9 Linearly Independent (L.I.) It satisfies index laws
Solutions Dm · Dn = Dn · Dm = Dm+n
We now extend the definition of the two linearly D1 = D, D2 = D · D, · · · , D0

independent functions given above to a set of = I D0 f (x) = 1.f (x) = f (x) .
linearly independent solutions of an nth order linear
differential equation.
A set of solutions {yi |i = 1, 2, · · · , n} defined on 2.1.10 Exponential Shift
some interval I is said to be linearly independent Theorem 2.2 If f (D) = Dn +a1 Dn−1 +a2 Dn−2 +
(L.I.) on I if · · · + an where ai are real constants, then
c1 y1 + c2 y2 + · · · + cn yn = 0
eαx f (D) y = f (D − α) eαx y
⇒ c1 = c2 = · · · = cn = 0 where y is a function of x.
It is linearly dependent (L.D.) on I if it is not Proof:

linearly independent on I .
The set of solutions {y; |i = 1, 2, · · · n} is linearly d αx
(D − α) eαx y = e y − αeαx y
independent on I if the Wronskian determinant dx
dy
W (y1 , y2, · · · , yn )(x0 ) = eαx + αeαx y − αeαx y
dx

y1 y2 ··· yn = eαx Dy
y1 y2 ··· yn
= (x0 ) = 0 (x0 ∈ I ) Apply (D − α) again
··· ··· ··· ···

y1(n−1) y2(n−1) ··· yn(n−1) (D − α)2 eαx y = (D − α) eαx Dy = eαx D2 y
Repeating this r times, we have

The differential operator D

We can conveniently denote (D − α)r eαx y = eαx Dr
dy d 2 y d xy % n
% n

, , ··· , ⇒ ar (D − α)r eαx y = ar eαx Dr y
dx dx 2 dxn r=0 r=0
as Dy, D y, · · · , D y, respectively, and write Eq.

2 n %
n
= eαx (ar Dr )y
(2.19) in the symbolic form
r=0

f (D)y = P0 Dn + P1 Dn−1 ⇒ f (D − α) eαx y = eαx f (D) y

+ · · · + Pn−1 D + Pn ) y = X or eαx f (D) y = f (D − α) eαx y
n
$ This relation shows that the effect of shifting an
where f (D) = Pr D n−r
y is a polynomial in the
r=0 exponential factor from the right-hand side of the
symbol D. operator to its left side is to replace D by (D + α).
Note If we put y = emx as a trial solution we have the

By the theorem on the exponential shift, we have auxiliary equation

f (D − α) (eαx V ) = eαx f (D)V (2.22) f (m) = mn + a1 mn−1 + · · · + an−1 m + an = 0
where V is a function of x. Changing D to D + α in [∵ emx = 0]

Eq. (2.20) we get Thus, the solution of the homogeneous differential
equation now reduces to that of an algebraic equation
f (D) (eαx V ) = eαx f (D + α)V (2.23) called the auxiliary equation obtained replacing D
with m. Solving the auxiliary equation we get n roots
Corollary 1 mi (i = 1, 2 · · · x), each giving rise to a solution.
The general solution containing n arbitrary con-
(D − α)n xk eαx = eαx Dn xk stants is obtained depending on the nature of roots in
⎧
⎪ 0 if k<n Table 2.2.
⎨
αx
= e n! if k=n Example 2.13
⎪
⎩ αx
e k (k − 1) · · · (k − n + 1) xk−n if k > n. Solve 4y + 4y + y = 0. [JNTU 2003]
Corollary 2 Solution The differential equation is

%
n
Let φ D 2
= ar D2r (4D3 + 4D2 + D)y = 0
r=0

φ D2 sin αx = φ −α2 sin αx
4m3 + 4m2 + m = m (2m + 1)2 = 0
φ D2 cos αx = φ −α2 cos αx
1 1
⇒ m = 0, m = − , −
Rules 2 2
(1) f (D) eαx = f (α) eαx
(Replace every D by α.) y = (c1 + c2 x) e−x/2 + c3 .
2 cos αx cos αx
(2) f D = f −α2 (2.24)
sin αx sin αx Example 2.14
Solve (D3 + 27)y = 0.
Replace every D2 by − α2 .
Higher order
m3 + 27 = (m + 3) m2 − 3m + 9 = 0
Linear differential equation with constant √
coefficients 3 3 3
⇒ m = −3, ±
We consider from now on, that the coefficients in Eq. 2 2
(2.19) are all constants, i.e., Pi = ai constants.
Method of finding the complementary function √ √
Consider the homogeneous equation −3x 3 3 3 3
y = c1 e + c2 cos x + c3 sin x e3x/2 .
2 2
f (D) y = 0

or Dn + a1 Dn−1 + · · · + an−1 D + an y = 0 Example 2.15
(a0 = 1) Solve y − y + 100y − 100y = 0.
Table 2.2
Case Roots Linearly independent General solution
solutions
1. n distinct, real roots {em1 x , em2 x , . . . , emn x } [c1 em1 x + c2 em2 x

m1 ,m2 , . . . , mn + · · · + cn emn x ]
2. Two equal, and (n − 2) {em1 x , xem1 x , (c1 + c2 x) em1 x + c3 em3 x
distinct real roots, m1 , m1 , em3 x , . . . , emn x } + · · · + cn emn x
m3 , . . . , mn

3. Three equal, and {em1 x , xem1 x , x2 em1 x , c1 + c2 x + c3 x2 em1 x
(n − 3) distinct real roots em4 x , . . . , emn x } +c4 em4 x + · · · + cn emn x
m1 , m1 , m1 , m4 , . . . , mn

4. r equal, (n − r) distinct {em1 x , xem1 x , xr−1 em1 x c1 + c2 x + · · · + cr xr−1 em1 x
real roots m1 , m1 , . . . , m1 emr+1 x , . . . emn x } +cr+1 emr+1 x + · · · + cn emn x
(r roots) mr+1 , . . . , mn
5. Two complex conjugate {eαx cos βx, eαx sin βx, eαx (c1 cos βx + c2 sin βx)
roots (n − 2) distinct real roots, em3 x , . . . emn x } +c3 em3 x + · · · + cn emn x
m1 = α + iβ, m2 = α − iβ
6. Two equal complex conjugate eαx cos βx, xeαx cos βx eαx [(c1 + c2 x) cos βx
roots, (n − 2) distinct real roots eαx sin βx, xeαx sin βx + (c3 + c4 x) sin βx]
m1 = m2 = α + iβ, em5 x · · · emn x +c5 em5 x + · · · + cn emn x
m3 = m4 = α − iβ
Solution The auxiliary equation is m3 − m2 + ⎧

⎪
⎪ cos π4 + i sin π
= 1+i
√
100m − 100 = 0. Clearly m = 1 is a root. By ⎪ 4 2
⎪
⎪
synthetic division ⎨ cos 3π + i sin 3π
= −1+i
√
= 4 4 2
1 1 −1 100 −100 ⎪
⎪
⎪
⎪
5π
cos 4 + i sin 5π
= −1−i
√
⎪
⎩ 4 2
1 0 100 cos 7π + i sin 7π
= 1−i
√
4 4 2
1 0 100 0
√ √ √
∴ m3 − m2 + 100 m − 100 y = eax/ 2 c1 cos(ax/ 2) + c2 sin(ax/ 2)
√ √ √
= (m − 1) m2 + 102 = 0
+e−ax/ 2 c3 cos(ax/ 2) + c4 sin(ax/ 2) .
⇒ m = 1, ±10i
Example 2.17
Solve (4D3 + 16D2 + 21D + 9)y = 0.
y = c1 ex + (c2 cos 10x + c3 sin 10x) .
Example 2.16
4
d y 4m3 + 16m2 + 21m + 9 = 0
Solve + a4 y = 0.
dx4 ⇒ (m + 1) (2m + 3)2 = 0
Solution The auxiliary equation is ⇒ m = −1, −3/2, −3/2
m4 + a4 = 0 m4 = −a4 = a4 e(2k+1)πi The general solution is
m
⇒ = e(2k+1)πi/4 k = 0, 1, 2, 3 y = c1 e−x + (c2 + c3 x) e−3x/2
a

By synthetic division 9. D3 + D2 + 4D + 4 y = 0 given that
y = 0, y = −1, y = 5 when x = 0.
−1 4 16 21 9 Ans: y = e−x − cos 2x
−4 −12 −9
4 12 9 0 2.1.11 Inverse Operator D−1 or 1
D
The general solution is Method of f inding the particular integral
y = c1 e−x + (c2 + c3 x) e−3x/2
If Q is a differentiable function of x on an interval
Example 2.18 d
2 I and D ≡ is a differential operator such that
Solve D2 + 3 y = 0. dx
dQ 1
DQ = , D−1 or is then called the inverse
Solution The auxiliary equation is dx D
2 2 √ 1 1
m + 3 = 0 ⇒ m = ±i 3 repeated operator such that (DQ) = D Q = 1·Q =
D D
Q.
√ √
1
y = (c1 + c2 x) cos 3x + (c3 + c4 x) sin 3x. Q = Qdx is the integral of Q(x)
D
EXERCISE 2.2
If Q is a continuous function on an interval I , then
Solve:
1 % n
1. (D3 + 3D2 − 10D)y = 0. Q, where f (D) = ar Dn−r ,

f (D)
Ans: y = c1 + c2 e2x + c3 e−5x r=0
2 is a function of x containing no arbitrary constant

2. D2 − 4 y = 0.
Ans: y = (c1 + c2 x) e2x + (c3 + c4 x) e−2x ] such that

d 3x d 2x dx 1
3. − 2 − 3 = 0. f (D) Q = Q.
dt 3 dt 2 dt f (D)
Ans: x (t) = c1 + c2 e3t + c3 e−t ]
Theorem 2.3 If Q is a continuous function of
4. D4 − 2D3 − 13D2 + 38D − 24 y = 0. x on I and α is a real or complex constant, then a
Ans: y = c1 ex + c2 e2x + c3 e3x + c4 e−4x ] particular value of

5. D3 + 3D2 − 4 y = 0.
1 αx
Ans: y = c1 ex + (c2 + c3 x) e−2x ] Q is e Qe−αx dx.
D−α

6. 2D4 − 3D3 − 2D2 y = 0.
Ans: y = c1 + c2 x + c3 e2x + c4 e−x/2 ]
1
7. D2 + D + 1 y = 0. Proof: Let y = Q
√ D−α
Ans: y = e−x/2 c1 cos x 3/2 Operating (D − α) on both sides, we get
√
+c2 sin x 3/2
1
(D − α) y = (D − α) =Q
D−α
8. D4 + 8D2 + 16 y = 0.
Ans: y = (c1 + c2 x) cos 2x + (c3 + c4 x) dy
sin 2x or − αy = Q
dx
which is a linear equation of the first order, whose Example 2.19

solution is Find the particular values of
1 1 1
(i) x3 (ii) 2 eax (iii) 3 cos ax.
ye−αx = c + Qe−αx dx D D D
Solution
Since we are interested here in a particular solution 1 3 1
1 (i) x = x3 dx = x4 .
Q, we take c = 0. Hence D 4
D−α
1 1 1 ax 1 eax
(ii) 2 eax = e =
αx
D D D D a
y=e Qe−αx dx
1 1 ax 1 ax
= e = 2e .
1 a D a
Thus, Q = eαx Qe−αx dx (2.23)
D−α
1 1
(iii) 3 cos ax = 2 cos axdx
1 1 D D
If and are two inverse operators
D−β D−α 1 sin ax 11 1
where α, β are given real/complex constants, then = 2 = sin ax
D a aD D

1 1 1 1 1
Q= Q (2.24) = sin axdx
(D − β) (D − α) D−β D−α a D

1 1 − cos ax
1 αx −αx =
= e Qe dx
D−β aD

a
−1
=e βx
e αx −αx
Qe dx dx = 2 cos axdx
a
1 sin ax 1
This can be extended to n factors. The following =− 2 = − 3 sin ax.
a a a
1
method of putting into partial fractions is of Example 2.20
f (D)
practical importance. Find the particular values of
1 1 1
(i) x (ii) e2x (iii) cos 2x.
D+2 D−2 D+5
2.1.12 General Method for Finding
the P. I. Solution
1 1
Theorem 2.4 If f (D) = (D − α1 ) (D − α2 ) · · · (i) x= x
D+2 D − (−2)
(D − αn ) where αr are real or complex constants,
then particular integral = e−2x xe2x dx
2x
1 1 e e2x
= Q= Q = e−2x x − 1 dx
f (D) (D − α1 ) (D − α2 ) · · · (D − αn ) 2 2

A1 A2 An x 1 2x
≡ + + ··· + Q e−2x − e
D − α1 D − α2 D − αn 2 4
%n 1
= Ar e αn x
Qe−αr x dx = (2x − 1) .
4
r=1

1
where Ar is the determinable constant corresponding (ii) e2x = e2x e2x · e−2x dx
1 D−2
to the partial fraction . = xe2x .
D − αr
1 1 Alternative method
(iii) cos 2x = cos 2x
D+3 D − (−3)
(i)
1
e2x
= e−3x cos 2xe3x dx (D − 1) (D − 2)

e3x 1 1
= e−3x = − e2x
+ 22
32 D−2 D−1
(3 cos 2x + 2 sin 2x) 1 1
= e2x − e2x
3 cos 2x + 2 sin 2x D− 2 D − 1
= .
13 = e2x e2x · e−2x dx − ex e2x e−x dx
Example 2.21 = xe2x − ex · ex = e2x (x − 1) .

Find the particular value of 1
(ii) x
(i)
1
e2x (ii)
1
x. (D + 1) (D − 1)
(D − 1) (D − 2) D −1
2
1 1 1
= − x
Solution 2 D−1 D+1
1 1
(i) e 2x 1 x −x 1 −x
D−1 D−2 = ·e xe dx − e xex dx
2 2

1 −x
= e2x e2x e−2x dx 1 x e e−x
D−1 = e x − 1· dx
2 (−1) (−1)
1 2x
= xe 1
D − 1 − e−x xex − 1ex dx
2
= ex xe2x · e−x dx 1 1
= ex (−x − 1) e−x − e−x · (x − 1) ex
2 2
=e x
xex dx 1 1
= (−x − 1) − (x − 1) = −x
2 2
= e xe − 1.e dx
x x x
= ex (x − 1) ex EXERCISE 2.3
= (x − 1) e2x . Find the particular value of each of the following:
1
1. 2 e3x .
1 D
(ii) x e3x
(D + 1) (D − 1) Ans:
9
1
= ex xe−x dx 1 3
D+1 2. x +1 .
D 2
1 e−x x5 x2
= ex −xe−x + Ans: +
D+1 (−1) 20 2
1
=− (x + 1) 1
D + 1 3. cos3 x.
D
= −e−x (x + 1) ex dx [Hint: cos3 x = 14 (cos 3x + cos x)]

1 sin 3x
= −e−x [(x + 1) ex − 1 · ex ] = −x Ans: + sin x
4 3
1 3x To find the particular integral,

4. e −2 .
−2
D2 1
1 yp = 2 e2x (1 + x)
Ans: e3x + 1 D − 7D + 6
7
1 1 1
1 = − (1 + x) e2x
5. cos 3x. 5 D−6 D−1
D2 − 9
1 1
−1 = (1 + x) e2x
Ans: cos 3x 5 D−6
18
1
1 − (1 + x) e 2x
6. x. D−1
−4
D2 1
−x = (I1 − I2 )
Ans: 5
4
1
I1 = (1 + x) e2x
2.2 General Solution of Linear D− 6
Equation f (D)y = Q(x) = e6x (1 + x) e2x .e−6x dx
We have seen that if y = yp is a particular solution
(containing no arbitrary constants) of the linear = e6x (1 + x) e−4x dx
equation f (D)y = Q(x) and y = yc is the general
solution of the homogenous equation f (D)y = 0 e−4x e−4x
= e6x (1 + x) −1
with as many arbitrary constants as the order of the (−4) (−4)2
equation then y = yc + yp is the general solution e2x
(G.S.) of the non-homogenous linear equation: =− (5 + 4x)
16
1
f (D)y = Q(x) I2 = (1 + x) e2x
D − 1
As we have studied methods of finding yc and the
= ex (1 + x) e2x · e−x dx
general method of finding yp earlier, we will work
out some examples and give some exercises.
The method of finding yp discussed above is par- =e x
(1 + x) ex dx
ticularly useful when Q(x) = tan ax, cot ax, sec ax
= ex [(1 + x) ex − 1 · ex ] = xe2x
or csc ax and also when we cannot use special short 2x
methods (discussed in section 2.2.1) when Q(x) = 1 e
yp = − (5 + 4x) − xe2x
eax , sin ax, cos ax, xm , eax V and xV . 5 16
e2x
Example 2.22 =− (1 + 4x)
16
Solve (D2 − 7D + 6)y = e2x (1 + x) .
[JNTU 2003] The general solution is
Solution To find the complementary function, e2x

y = yc + yp = c1 ex + c2 e6x − (1 + 4x) .
we have to solve 16
(D2 − 7D + 6)y = 0 Example 2.23

Solve (D3 + 2D2 − 3D)y = xe3x .
m2 − 7m + 6 = (m − 1) (m − 6) = 0 Solution To find the complementary function,
⇒ m = 1, 6 we have to solve
yc = c1 ex + c2 e6x (D3 + 2D2 − 3D)y = 0
6x
The auxiliary equation is e 1.e6x
= e−3x x − 2
m3 + 2m2 − 3m = m m2 + 2m − 3 6 6
e3x
= m (m − 1) (m + 3) = 0 = (6x − 1)
⇒ m = 0, 1, −3 36
1 1 1
yc = c1 + c2 ex + c3 e−3x =e x − + +
3x
9 8 72

To find the particular integral, 1 1 1
+ − −
1 1 27 6 432
yp = xe3x = xe3x . (x − 1) 3x
f (D) D (D − 1) (D + 3) = e
36
Partial fractions The general solution is
x − 1 3x
1 A B C y = c1 + c2 ex + c3 e−3x + e .
xe3x = + + xe3x 36
f (D) D D−1 D+3
Example 2.24
1 3x 1 Solve (D2 + a2 )y = sec ax. [JNTU 2000]
= A· xe +B xe3x
D D−1
Solution To find the complementary function,
1
+C xe3x we have to solve
D+3
(D2 + a2 )y = 0
1 e3x 1
= − (3x − 1) + ·(2x − 1)
3 9 4 The auxiliary equation is
e3x 1 e3x
+ (6x − 1) m2 + a2 = 0 ⇒ m = ±ia
4 12 36
D 1 yc = c1 cos ax + c2 sin ax
A= =−
f (D) D=0 3
To find the particular integral,
D−1 1
B= = 1
f (D) D=1 4 yp = sec ax
D+3 (D + ia) (D − ia)
1
C= = 1 1 1
f (D) D=−3 12 = − sec ax
2ia D − ai D − ia
1
= (I1 − I2 )
1 3x e3x e3x 2ia
xe = xe3x dx = x −1 2 1
D 3 3 I1 = sec ax
e3x D − ai
= (3x − 1)
9 = eiax e−iax sec ax dx
1 3x −x
xe3x =e x
xe e dx cos ax − i sin ax
D−1 =e iax
dx
2x cos ax
e e2x
= ex x − 2 i
2 2 = eiax x + log cos ax
a
e3x
= (2x − 1) i
4 I2 = eiax x − log cos ax ,
1 a
xe3x = e−3x xe3x e3x dx Replacing i by − i in the above result, we have
D+3
1 iax eiax
∴ yp = x e − e−iax =− [(cos ax − i sin ax)
2ia a
i +i log (sec ax + tan ax)]
+ log cos ax eiax + e−iax
a 1
I1 = − 1 + ieiax log (sec ax + tan ax)
x 1 a
= sin ax + 2 cos ax log (cos ax)
a a Replacing i by −i in the above result, we have
1
The general solution is I2 = − 1 − ieiax log (sec ax + tan ax)
a
y = yc + yp = c1 cos ax + c2 sin ax 1
yp = − 2 1 + ieiax log (sec ax + tan ax) − 1
x 1 2ia
+ sin ax + 2 cos ax log cos ax. +ieiax log (sec ax + tan ax)
a a
1
= − 2 cos ax log (sec ax + tan ax)
Example 2.25 a
d 2y The general solution is
Solve + a2 y = tan ax. y = yc + yp = c1 cos ax + c2 sin ax
dx2
1
Solution To find the complementary function, − 2 cos ax log (sec ax + tan ax) .
a
we have to solve
Example
2.26

(D + a )y = 0
2 2 Solve D2 − 1 y = 2ex + 3x. [JNTU 1996S]

The auxiliary equation is we have to solve
m2 + a2 = 0 ⇒ m = ±ia (D2 − 1)y = 0
y = c1 cos ax + c2 sin ax The auxiliary equation is
To find the particular integral, m2 − 1 = 0 ⇒ m = ±1

yc = c1 ex + c2 e−x
1
yp = tan ax To find the particular integral,
D2 + a2
1 1 1
1 1 1 yp = − 2ex
yp = − tan ax 2 D−1 D+1
2ia D − ia D + ia
1 1 1 1
= (I1 − I2 ) + − 3x
2ia 2 D−1 D+1

1
I1 = tan ax = eiax eiax tan ax dx = ex e−x ex dx − e−x ex ex dx
D − ia

cos ax − i sin ax 3 x
= eiax sin ax dx + e xe−x dx − e−x xex dx
cos ax 2

− cos ax −x
= eiax − i (sec ax − cos ax) dx e2x 3 e e−x
a = xex − e−x . + ex x −
2 2 (−1) (−1)2
∵ sin2 ax = 1 − cos2 ax x
− cos ax 3 e
i − e−x x − ex
= eiax − {log (sec ax + tan ax) 2 1
a a 1
− sin ax} = xex − ex − 3x
2
The general solution is 2. (D2 − 1)y = 3 + 7x.

Ans: y = c1 cosh x + c2 sinh x − 3 − 7x
y = y c + yP
1
= c1 ex + c2 e−x + xex − ex − 3x. 3. (D2 + 9)y = sec 3x.
2 x
Ans: y = c1 cos 3x + c2 sin 3x + sin 3x
Example 2.27 3
Solve (D2 + 1)y = cosec x. cos 3x
+ log cos 3x
Solution To find the complementary function, 9
we have to solve
4. (D2 + a2 )y = cos bx.
(D + 1)y = 0
2 cos bx
Ans: y = c1 cos ax + c2 sin ax + ,
a2 − b2
The auxiliary equation is a2 = b2
m2 + 1 = 0 ⇒ m = ±i
yc = c1 cos x + c2 sin x 5. (D2 + 3D + 2)y = x2 .
1 2 7
Ans: y = c1 e−x + c2 e−2x + x − 3x +
To find the particular integral, 2 2

1 1 1
yp = − cosec x
2i D − i D − i 2.2.1 Short Methods for Finding

1 1 1 the Particular Integrals in
= cosec x − cosec x
2i D − i D+i Special Cases

1
cosec x = eix e−ix cosec x dx n
D−i %
Let f (D) y = ar D n−r
y = Q (x) (2.25)
cos x − i sin x
= eix dx r=o
sin x

cos x be a non-homogenous linear equation, where ar are
= eix dx − i dx
sin x real constants.
We will now consider short methods for finding the
= e log sin x − ix
ix
P.Is in special cases when Q(x) is of the form

1
cosec x = e−ix log sin x + ix (1) eaX (2) sin ax or cos ax (3) xm (4) eaX V
D+i and (5) xV , where V is a function of x.
Replacing i by −i in the above result, we have
eix − e−ix eix + e−ix
yp = log sin x − x (1) Q(x) = eax where ‘a’ is a constant
2i 2
= sin x log sin x − x cos x Case (i) f (a) = 0
y = yc + yp = c1 cos x + c2 sin x Since Deax = aeax , D2 eax = a2 eax , · · ·
+ sin x log sin x − x cos x. and generally, Dr eax = ar eax
n
%
EXERCISE 2.4 f (D) e =
ax
ar D n−r
eax
r=0
Solve:
%
n
1. (D2 − 3D + 2)y = e3x . = ar a n−r

eax
e3x r=0
Ans: y = c1 ex + c2 e2x + = f (a) eax (2.26)
2
1 where (D − a) is not a factor of φ(D).

Applying on both sides of Eq. (2.23)
f (D)
We have f (r) (D) = r!φ(D) + terms containing
1 1 (D − a) as a factor
f (D) eax = f (a) eax
f (D) f (D) (r)
⇒ f (a) = r!φ(a)

1 ax 1 ax 1
= f (a) e e = eax
f (D) f (D) (D − a)r φ(D)

1 ax 1 1
⇒ e = f (a)
ax
e = e ax
f (D) φ(a) (D − a)r
1 ax 1 ax 1 ax xr eax
⇒ e = e , (2.27) Thus, e = (F3 )
f (D) f (a) f (D) r!φ(a)
provided f (a) = 0 1 ax xr eax
or e = (r) (F4 )
f (D) f (a)
Rule Example 2.28

Replace D in f (D) with a if f (a) = 0, i.e., (D − a) D + 1 4x
Evaluate e .
is not a factor of f (D). D2 − 5
Solution We have to evaluate
Case (ii) f (a) = 0, f (a) = 0, i.e., (D − a) is a
factor of f (D) but (D − a)2 is not a factor of f (D). D + 1 4x
e
Let f (D) = (D − a)φ(D), where (D − a) is not a D2 − 5
factor of φ(D). D + 1 4x 4 + 1 4x
e = 2 e
D2 − 5 4 −5
We have f (D) = φ(D) + (D − a)φ (D) =
5 4x
e
⇒ f (a) = φ(a) 11

∵ 4 is not a factor of D2 − 5 .
1 ax 1 Example 2.29
e = eax
f (D) (D − a)φ(D) 1
Evaluate e2x .
1 1 D2 −4
= eax , φ(a) = 0
φ(a) D − a
Solution Here f (D) = D2 − 4 = (D − 2)(D + 2)
1
= eax eax · e−ax dx
φ(a) 1 1
∴ e2x = e2x
xeax D2 −4 (D − 2)(D + 2)
= (F1 )
1!φ(a) 1
= e2x
xeax (F2 ) (D − 2)(2 + 2)
or
f (a) f (a) = 0 1 1
= e 2x
4 D−2
Case (iii) f (a) = 0, f (a) = 0, · · · f (r) (a) = 0, 1 x 2x x
= e = e2x .
f (r+1) (a) = 0, 4 1! 4
i.e., (D − a)r is a factor of f (D) but (D − a)r+1 is not
a factor of f (D). Example 2.30
1
Let f (D) = (D − a)r φ(D) Evaluate e2x .
(D − 2)2 (D + 3)
Solution Alternative method

Here, f (D) = (D − 2)2 (D + 3) 3
f (D) = D2 − 3 = D6 − 9D4 + 27D2 − 27,
= (D − 2)2 φ(D)
f (D) = 6.5.4D3 − 9.4.3.2D
where φ(D) = D + 3
1 √3x 1 √
φ(2) = 2 + 3 = 5 yp = e = √ e 3x
f (D) f 3
1 x2 e2x
e 2x
= √
(D − 2)2 (D + 3) 2! 5 x3 e 3x
= √ √
x2 e2x 3
= , by F3 6.5.4 3 − 9.4.3.2 3
10 √
x3 e 3x
Alternative method = √
144 3
f (D) = (D − 2)2 (D + 3)
Example 2.32
= D3 − D2 − 8D + 12 Solve y − y − 2y = 3e2x given that

f (D) = 3D2 − 2D − 8
f (D) = 3.2D − 2.1 y(0) = 0, y (0) = −2 [JNTU 2004] (1)

f (2) = 3.2.(2) − 2.1 = 10 Solution To find the complementary function,
1 1 we have to solve
∴ e2x = 3 e2x
(D − 2) (D + 3)
2 D − D − 8D + 12
2
x2 e2x x2 e2x y − y − 2y = 0 (2)

= = , by F4 .
f (2) 10
Example 2.31 √
Find the particular integral yp of (D2 − 3)3 y = e 3x
. m2 − m − 2 = (m − 2) (m + 1) = 0
⇒ m = 2, −1
Solution
yc = c1 e2x + c2 e−x (3)
3
f (D) = D2 − 3
√ 3 To find the particular integral,
= D − 3 φ(D),
1 1
√ 3 yp = Q (x) = 2 3e2x
φ(D) = D + 3 f (D) D −D−2
x
1 √ = 3e2x = xe2x
yp = e 3x (2D − 1)D=2
√ 3 √ 3
D− 3 D+ 3
1 1 √
= √ √ 3 √ e 3x
, y (x) = yc + yp = c1 e2x + c2 e−x + xe2x (4)
3+ 3 D− 3
√ √ √ Differentiating Eq. (5) with respect to x
∵ φ 3 = 3+ 3
y (x) = 2c1 e2x − c2 e−x + 1e2x + 2e2x (5)
1 x3 √
= √ 3
· e 3x y (0) = 0 = c1 · 1 + c2 · 1 + 0
x!
2 3
⇒ c 1 + c2 = 0 (6)
x3 √
y (0) = −2 = 2c1 · 1 + c2 · 1 + 1 + 0
= √ e 3x , by F3 .
144 3 ⇒ 2c1 − c2 + 1 = −2 (7)
Adding Eqs. (6) and (7) To find the particular integral,

3c1 + 1 = −2 ⇒ c1 = −1 (∵ c2 = 1) 1
yp = Q (x)
f (D)
Substituting these values in Eq. (5), the required
1 1 1
solution is obtained as = 3 e2x + 2 3 ex + 3 e0.x
D −1 D −1 D −1
y (x) = −e2x + e−x + xe2x . (8) 1 2x
= e2x + ex − 1 (4)
Example 2.33 7 3
Solve (D3 + 2D2 + D)y = e2x . (1)
y (x) = yc + yp
Solution To find the complementary function, √ √
we have to solve 3 3
= c1 e + c2 cos
x
x + c3 sin x
(D3 + 2D2 + D)y = 0 (2) 2 2
1 2x
The auxiliary equation is e−x/2 + e2x + ex − 1. (5)
7 3
m3 + 2m2 + m = m (m + 1)2 = 0
Example 2.35
⇒ m = 0, −1, −1 (3)
Solve (D2 + 6D + 9)y = e−3x . (1)
yc = c1 + (c2 + c3 x) e−x (4)
1 1 we have to solve
yp = Q (x) = 3 e2x
f (D) D + 2D2 + D
(D2 + 6D + 9)y = 0
1 e2x
= 3 e 2x
= The auxiliary equation is
2 + 2.22 + 2 18
The general solution is m2 + 6m + 9 = (m + 3)2 = 0
e2x ⇒ m = −3, −3
y (x) = yc + yP = c1 + (c2 + c3 x) e−x + . (5)
18 yc = (c1 + c2 x) e−3x (2)
Example 2.34 To find the particular integral,
Solve (D − 1)y = (e + 1) .
3 x 2
(1) 1 1
yp = Q (x) = 2 e−3x
Solution To find the complementary function, f (D) D + 6D + 9
we have to solve x2 x2
= e−3x = e−3x ;
(D3 − 1)y = 0 (2) 2 2

D + 6D + 9
2
=2
The auxiliary equation is D=−3
m3 − 1 = (m − 1) m2 + m + 1 = 0
√ y (x) = yc + yp = (c1 + c2 x) e−3x + x2 −3x
e
2
1 i 3
⇒ m = 1, − ± Example 2.36
2 2
√
3 Solve (D2 − k 2 )y = cosh kx. (1)
yc = c1 e + c2 cos
x
x
2 Solution To find the complementary function,
√ we have to solve
3
+c3 sin x e−x/2 (3)
2 (D2 − k 2 )y = 0 (2)
The auxiliary equation is 6. (2D + 1)2 y = 4e−x/2 .

x2 −x/2
m2 − k 2 = 0 ⇒ m = ±k Ans: y = (c1 + c2 x) e−x/2 + e
2
yc = c1 ekx + c2 e−kx (3)
7. D2 − D − 6 y = ex cosh 2x.
To find the particular integral, x 3x 1 −x
Ans: y = c1 e3x + c2 e−2x + e − e
1 10 8
yp = Q (x)
f (D)
kx 8. D2 − 2pD + p2 y = ex .
1 1 1 e + e−kx
= − Ans: ⎧
2k D − k D+k 2 ⎨ (c1 + c2 x) epx + (p−1)ex
(p = 1)
2
1 1 1 1 y=
= ekx + e−kx − ekx ⎩ 2
4k D − k D−k D−k (c1 + c2 x) ex + x2! ex (p = 1)

1 −kx
− e 9. D4 − 1 y = ex .
D+k
Ans: y = c1 ex + c2 e−x + c3 cos x + c4 sin x
1 1 −kx 1 kx −kx x
= xe − e − e − xe
kx
+ ex
4k 2k 2k 4
x 1
= sinh kx − 2 cosh kx (2) Q (x) = sin ax or cos ax where ‘a’ is a constant
2k 4k
The general solution is Case (i) f (−a2 ) = 0
We have D(sin ax) = a cos ax
y (x) = yc + yp = c1 ekx + c2 e−kx
x 1
+ sinh kx − 2 cosh kx. (4) D2 sin ax = D (a cos ax) = −a2 sin ax
2k 4k 2 2 2
D sin ax = −a2 sin ax
EXERCISE 2.5
and generally,
Solve the following: 2 r r
D sin ax = −a2 sin ax
1. (D2 + 1)y = cosh x.
1 for any integer ‘r’.
Ans: y = c1 cos x + c2 sin x + cosh x
2 ∴ f D2 sin ax = f −a2 sin ax

2. D2 − 1 y = sinh x. 1
x Operating on both sides by , we have
Ans: y = c1 cosh x + c2 sinh x + cosh x f D2
2
1 1
3. (D + 1)3 y = e−x . f D2 sin ax = f −a2 sin ax
f D 2 f D2
x3
Ans: y = c1 + c2 x + c3 x2 e−x + e−x
3! 2 1
3 ⇒ sin ax = f −a sin ax
4. D − 5D2 + 7D − 3 y = e2x cosh x. f D2
Ans: y = (c1 + c2 x) ex + c3 e3x 1 1
x 3x ⇒ sin ax = sin ax (2.28)
+ e − 8ex f D 2 f −a2
8
3 if f −a2 = 0.
5. D + 1 y = 3 + 5ex . [JNTU 2002]
√ √
3 3 Rule
Ans: y = c1 + c2 cos x + c3 sinh x Replace D2 by −a2 if f (−a2 ) = 0
2 2
The same rule applies if Q(x) = cos ax General case: If f (D) contains odd powers of D,
then we can arrange the even and odd powers of D
1 1
cos ax = cos ax (2.29) so that
f D 2 f −a2

f (D) = f1 D2 + Df2 D2
if f −a2 = 0 2 2 2 2
Let D = f1 D − D 2 f2 D 2
This rule holds if sin ax and cos ax are replaced by
sin(ax + b) and cos(ax + b), where b is a constant. Then
1 1
sin ax = 2 sin ax
Case (ii) f −a2 = 0, f (D) f1 D + Df2 D2
i.e., D2 + a2 |f D2 , D2 + a2 f D2
The above rule fails if f (−a2 ) = 0. In this case,
f1 D2 − Df2 D2
(D2 + a2 )|f (D2 ) so that we can write = sin ax
D2
f D2 = D2 + a2 φ D2 where φ −a2 = 0.
sin ax
Let us evaluate = f1 D2 − Df2 D2
−a2
2
1 1 1 if −a = 0
e iax
= e iax
2 2
D 2 + a2 D − ia D + ia
f1 −a sin ax − f2 −a a cos ax
1 1 = , (2.32)
= eiax −a2
2ia D − ia 2
1 iax if −a = 0
= xe
2ia Similarly,
1 x
(cos ax + i sin ax)= (cos ax + i sin ax)
D 2 + a2 2ia 1 f1 −a2 cos ax + f2 −a2 a sin ax
cos ax = .
Equating the real and imaginary parts, f (D) −a2
2
1 x if −a = 0 (2.33)
cos ax = sin ax
D 2 + a2 2a
Example 2.37
1 x
sin ax = − cos ax 1
D +a
2 2 2a Evaluate
D3 + D2 + D + 1
sin 2x.
2
We have, if φ −a = 0,
Solution
1 1 1
sin ax = 2 sin ax 1
f D2 D + a2 φ D 2 sin 2x
D3 + D2+D+1
1 1
= 2 sin ax 1
φ −a D 2 + a2 = 2 sin 2x
D + 1 + D D2 + 1
1 x
= 2 − cos ax (2.30) 1
φ −a 2a = sin 2x
−3 (D + 1)
1 1 1 D−1
cos ax = 2 cos ax = sin 2x
f D2 D + a2 φ D 2 −3 D2 − 1

1 1 D (sin 2x) − 1 · sin 2x
= 2 cos ax =
φ −a D 2 + a2 −3 −22 − 1
1 x 2 cos 2x − sin 2x
= 2 sin ax . (2.31) = .
φ −a 2a 15
Example 2.38 The general solution is

1
Evaluate 2 sin2 x. y (x) = yc + yP
D −4
9 cos 3x − 7 sin 3x
= c 1 e x + c2 e x + .
Solution 130

1 1 1 − cos 2x Example 2.40
sin x = 2
2
Solve (D2 − 1)y = sin x cos x.
D2 − 4 D −4 2
1 1 1 1
= e0.x − · 2 cos 2x Solution To find the complementary function,
2 D2 − 4 2 D −4 we have to solve
1 1 1 1
= ·1− · cos 2x 2
20−4 2 −2 − 4
2 D −1 y =0
1 1
=− + cos 2x. The auxiliary equation is
8 16
Example 2.39 m2 − 1 = (m − 1) (m + 1) = 0 ⇒ m = 1, −1
Solve (D − 3D + 2)y = sin 3x.
2
(1) yc = c1 e + c2 e
x −x
Solution To find the complementary function, To find the particular integral,

we have to solve 1 1
yp = Q (x) = 2 sin x cos x
(D2 − 3D + 2)y = 0 (2) f (D) D −1
1 1
The auxiliary equation is = sin 2x
2 D2 − 1
1 1 1
m2 − 3m + 2 = 0 ⇒ (m − 1) (m − 2) = 0 = sin 2x = − sin 2x
2 −22 − 1 10
⇒ m = 1, 2
The general solution is,
yc = c1 ex + c2 e2x (3) 1
y (x) = yc + yp = c1 ex + c2 e−x − sin 2x.
To find the particular integral, 10
1 1 Example 2.41
yp = Q (x) = 2 sin 3x Solve (D2 − 4D)y = ex + sin 3x cos 2x.
f (D) D − 3D + 2
1
= sin 3x Solution To find the complementary function,
−3 − 3D + 2
2
we have to solve
1 3D − 7
=− · sin 3x.
3D + 7 3D − 7 (D2 − 4D)y = 0
To get D2 in the denominator The auxiliary equation is

m2 − 4m = 0 ⇒ m = 0, m = 4
3D − 7
=− sin 3x yc = c1 + c2 e4x
9D2 − 72
3D − 7 To find the particular integral,
= − 2 sin 3x
9 −3 − 72
1 1 1
3 (D sin 3x) − 7 · sin 3x yp = ex + 2 · (sin 5x + sin x)
= D2 − 4D D − 4D 2
130
9 cos 3x − 7 sin 3x 1
= ∵ sin 3x cos 2x = [sin(3x + 2x) + sin(3x − 2x)]
130 2

1
P.I1 = ex yp =
1
e2x + 2
1 1
cos (3x + 7)
D − 4D
2
D3 + 2D D +2D
1 1
= ex = − ex P.I1 . =
1
e2x = 3
1
e2x =
1 2x
12 − 4.1 3 D + 2D
3 2 + 2.2 12
e
1
P.I2 = sin 5x P.I2 . =
1 1
sin (3x + 7)
D − 4D
2
3 D2 + 2
1
= sin 5x =
1 1 1
(sin 3x + 7) = − sin (3x + 7)
−52 − 4D .
3 −32 + 7 6
4D − 25
= − sin 5x e2x 1
16D2 − 252 yp = − sin (3x + 7)
12 6
4.5 cos 5x − 25 sin 5x
= −
16 −52 − 625 The general solution is
4 cos 5x − 5 sin 5x y (x) = yc + yp
=
105 √ √
= c1 + c2 cos 2x + c3 sin 2x
1 1
P.I3 = sin x = sin x e2x 1
D2 − 4D −12 − 4D + − sin (3x + 7) .
12 6
4D − 1
=− sin x
16D2 − 12 EXERCISE 2.6
1 Solve the following:
= (4 cos x − sin x)
17
1 4 cos 5x − 5 sin 5x 1. y + 4y + 47 = 4 cos x + 3 sin x;
yp = − e x + y (0) = 1, y (0) = 0.
3 210
1 Ans: y = (1 + x) e−2x + sin x
+ (4 cos x − sin x)
34 2. (D3 − 1)y = ex + sin 3x√ + 2. [JNTU
√ 2004]
Ans: y = c1 ex + c2 cos x +c3 sin x
2 2
y(x) = yc + yp 27 cos 3x − sin 3x
1 4 cos 5x − 5 sin 5x +
= c1 + c2 e4x − ex + 730
3 210 2
1 3. D − 4 y = 2 cos2 x.
+ (4 cos x − sin x) . 1
34 Ans: y = c1 e2x + c2 e−2x − (2 + cos 2x)
8
Example
2.42
Solve D3 + 2D y = e2x + cos (3x + 7) . 4. (D2 + 1)y = sin x sin 2x. [JNTU 2003]
x cos x
Ans: y = c1 cos x + c2 sin x + sin x +
Solution To find the complementary function, 4 16
we have to solve 2
3 5. D + 9 y = cos 3x + sin 2x.
D + 2D y = 0
Ans: y = c1 cos 3x + c2 sin 3x
The auxiliary equation is x sin 2x
+ sin 3x +
m3 + 2m = 0 ⇒ m m2 + 2 = 0 6 5
√ 2
⇒ m = 0, m = ± 2i 6. D − 1 y = sin (x + 5)
√ √ 1
yc = c1 + c2 cos 2x + c3 sin 2x Ans: y = c1 ex + c2 e−x − sin (x + 5)
2

7. D2 + 4 y = ex + sin 2x + cos 2x. 1 D D2
= 1− + − ··· x
1 3D 3 9
Ans: y = (c1 cos 2x + c2 sin 2x) + ex
5 11 1 1
x = x − x + Dx
+ (− cos 2x + sin 2x) 3D 9 27
4
1 2 1 1
= x − x+ .
6 9 27
8. D2 − 4D + 3 y = sin 3x cos 2x.
1 Example 2.45
Ans: y = c1 ex + c2 e3x + (10 cos 5x
884 Solve D3 − 1 y = x3 .
1
−11 sin 5x) + (sin 2x + 2 cos x)
20 Solution To find the complementary function,
we have to solve
(3) Q(x) = xm where m is a positive integer
(D3 − 1)y = 0
1 m 1
To evaluate x , we expand in increasing The auxiliary equation is
f (D) f (D)
m m
powers of D up to D and apply on x .
m3 − 1 = 0 ⇒ m = 1,
Example 2.43 √
1 2 1 3
Evaluate x +x+1 . 1− ± i
1+D+D 2 2 2
√ √
Solution 3 3
yc = c1 ex + c2 cos x + c3 sin x e−x/2
2 2
1 2
x +x+1 To find the particular integral,
1+D+D 2
1
1−D 2 yp = 3 x3
= x +x+1 D −1
1−D 3
−1 3
−1 2 yp = − 1 − D3 x
= (1 − D) 1 − D3 x +x+1
= − 1 + D + D6 + · · · x 3
3
= (1 − D) 1 + D3 + D6 + · · ·
2 = −x3 − 6
x +x+1

= 1 − D + D3 − · · · x 2 + x + 1 The general solution is
= x + x + 1 − (2x + 1)
2
y(x) = yc + yc
= x2 − x. √ √
3 3
= c1 e + c2 cos
x
x + c3 sin x
2 2
Example 2.44
Evaluate (D2 + 3D)−1 x. ×e−x/2 − x3 − 6.
Solution Example 2.46

Solve (D3 − D2 )y = 1 + x3 .
−1 1 1 1
D2 + 3D x= x= x Solution To find the complementary function,
3D + D2 3D 1 + D
3
we have to solve
1 D −1
= 1+ x (D3 − D2 )y = 0
3D 3
The auxiliary equation is The general solution is

m3 − m2 = m2 (m − 1) = 0 ⇒ m = 0, 0, 1 x3
y(x) = c1 + c2 e−x/2 + − 2x2 + 8x − 16.
3
yc = c1 + c2 x + c3 ex
EXERCISE 2.7
Solve the following:
1
yp = 1 + x3 1. y − 2y − y − 2y = 1 − 4x3 . [JNTU 2004(3)]
D3 − D2
Ans: (5 + cx2 )x3 y5 = 2
1
= − 2 (1 − D)−1 (1 + x3 )
D 2. D2 + D + 1 y = x3 . [JNTU
2004]
1 √ √
= − 2 (1 + D + D2 + D3 + D4 + D5 + · · · ) Ans: y = c1 cos 2 + c2 sin 2
3x 3x
D
×(1 + x3 ) e−x/2 + x3 − 3x2 + 6

1 1 3. (D3 − 3D − 2)y = x2 .
= − + + 1 + D + D 2
+ D 3
+ · · ·
D2 D Ans: y = c1 e2x + (c2 + c3 x)e−x
×(1 + x )
3 − 14 (2x2 + 6x + 9)
2
x x5 x4
= − +x+1+ + + x3 4. (D3 − D2 − D + 1)y = 1 + x2 .
2 4.5 4
[JNTU 1997S]
+3x2 + 6x + 6 Ans: y = c1 e−x + (c2 + c3 x) ex + x2 + 2x + 5
The general solution is 5. (D3 + 2D2 + D)y = e2x + x2 + x + sin 2x.
y(x) = c1 + c2 x + c3 ex [JNTU 2003S(1)]
5
x x4 7 e2x
− + + x3 + x2 + 7x + 7 . Ans: y = c1 + (c2 + c3 x) e−x +
20 4 2 18
x
+ (2x2 − 9x + 24)
Example 2.47 6
Solve (2D2 + D)y = x2 . 1
+ (3 cos 2x − 4 sin 2x)
50
we have to solve 6. (D2 + 5D + 4)y = x2 . [JNTU 2003S(2)]
1
(2D2 + D)y = 0 Ans: y = c1 e −x
+ c2 e −4x
+ (8x2 − 20x + 21)
The auxiliary equation is 32
2m2 + m = 0
1 (4) Q(x) = eax V(x) where ‘a’ is a constant
m(2m + 1) = 0 ⇒ m = 0, − and V(x) is a function of x
2
yc = c1 + c2 ex/2 By the theorem on exponential shift, we have
To find the particular integral, f (D) (eax V1 ) = eax f (D + a) V1 (2.34)
1 1
yp = x2 = (1 + 2D)−1 x2 where V1 is a function of x.
D(1 + 2D) D 1
1 Applying on both sides, we have
= 1 − 2D + 4D2 − 8D3 + · · · x2 f (D)
D
1 1
= (x2 ) − 2x2 + 4D(x2 ) − 8D2 (x2 ) eax V1 = [eax f (D + a) V1 ] (2.35)
D f (D)
x3 1
= − 2x2 + 8x − 16 ⇔ [eax f (D + a)V1 ] = eax V1 (2.36)
3 f (D)
Let f (D + a) V1 = V Example 2.51

1 Solve (D4 − 1)y = ex cos x.
⇔ V1 = V (2.37)
f (D + a) Solution To find the complementary function,
From Eqs. (2.36) and (2.37), we get we have to solve (D4 − 1)y = 0
1 1
(eax V ) = eax V
f (D) f (D + a) m4 − 1 = (m − 1) (m + 1) m2 + 1 = 0
The effect of shifting the exponential factor eax ⇒ m = ±1, ±i
from the right side of the operator to the left side is yc = c1 ex + c2 e−x + c3 cos x + c4 sin x
to replace D by (D + a) in f .
Example 2.48
1
1 yp = ex cos x
Evaluate ex x. D4 − 1
(D − 1)3 1
= ex cos x
Solution (D + 1)4 − 1
1 1
ex x = ex = ex 4 cos x
(D − 1)3 D + 4D3 + 6D2 + 4D
1 1 1
x = ex 3 x = ex 2 cos x
(D + 1 − 1) 3
D −12 + 4 −12 D + 6 −12 + 4D
x4 x 4 ex ex cos x
= ex · = . =−
2.3.4 24 5
Example 2.49 The general solution is
1
Evaluate 2 xe2x . y(x) = yc + yp
D − 4D + 4
ex cos x
Solution = c1 ex + c2 e−x + c3 cos x + c4 sin x − .
5
1 1
e2x · x = e2x x Example 2.52
D2 − 4D + 4 (D − 2)2 Solve (D2 + 4D + 4)y = e−2x cos x.
1
= e2x x
(D + 2 − 2)2 Solution To find the complementary function,
1 we have to solve (D2 + 4D + 4)y = 0
= e2x 2 x
D
x3 x 3 ex The auxiliary equation is
= e2x = .
2·3 6 m2 + 4m + 4 = (m + 2)2 = 0
Example 2.50 ⇒ m = −2, −2
1
Evaluate e5x sin x. yc = (c1 + c2 x) e−2x
(D − 5)3
Solution
1 1 1
e5x sin x = e5x sin x yp = e−2x cos x
(D − 5)3 (D + 5 − 5)3 (D + 2)2
1 e−2x
= e5x 3 sin x = −e5x cos x. = cos x
D (D − 2 + 2)2
(5) Q(x) = xV (x) where V(x) is a function of x

1
= e−2x 2 cos x Theorem 2.5 Prove that
D
= −e−2x cos x D (xV1 ) = xDn V1 + nDn−1 V1
n
(2.38)
where V1 is a function of x.
y(x) = y (x) = yc + yp Proof: For n = 1
= (c1 + c2 x) e−2x − e−2x cos x. D (xV1 ) = x (DV1 )+V1
1
(true)
Assume that Eq. (2.38) is true for n = m
EXERCISE 2.8
Dm (xV1 ) = xDm V1 + mDm−1 V1 (2.39)
Applying D on both sides,
1. (D2 − 7D + 6)y = e2x (1 + x).
[JNTU 2003] Dm+1 (xV1 ) = D (xDm V1 ) + mDm V1
e2x = xDm+1 V1 + (m + 1) Dm V1 (2.40)
Ans: y = c1 e−x + c2 e−6x − (4x + 1)
16 If Eq. (2.38) is true for n = m, then it is true for
2. (D − 2D + 1)y = xe .
2 x n = m + 1. Hence by the principle of mathematical
Ans: y = (c1 + c2 x) induction, Eq. (2.38) is true for all n ∈ N
%
ex − ex (x sin x + 2 cos x) Let f (D) = Cr D r
3. (D3 − 3D2 + 3D − 1)y = x2 ex . %
f (D) (xV1 ) = Cr Dr (xV1 )
[JNTU 2002S]
5 x %
x e =x Cr Dr V1
Ans: y = (c1 + c2 x + c3 x2 )ex +
60 %
4. (D2 + 4D + 3)y = e−x sin x + x. + rCr Dr−1 V1
[JNTU 1996, 1999] = xf (D) V1 + f (D) V1
Ans: y = c1 e−x + c2 e−3x 1
e−x Take V1 = V in the above result.
− (2 cos x + sin x) f (D)
5
1 4
+ x− 1 1
3 3 We have f (D) x V = xV + f (D) V
f (D) f (D)
5. (D2 − 4)y = x sinh x. [JNTU 2000]
1 1 1
Ans: y = c1 e2x + c2 e−2x ⇒ x V = (xV ) + f
x 2 f (D) f (D) f (D)
− sinh x − cosh x 1
3 9 (D) V
6. (D3 − 7D2 + 14D − 8)y = ex cos 2x. f (D)
[JNTU 2003S(3)] 1 1 1
⇒ (xV ) = x V− f
Ans: y = c1 ex + c2 e2x + c3 e4x f (D) f (D) f (D)
2ex 1
+ (8 cos 2x − sin 2x) (D) V
229 f (D)

7. (D2 − 4D + 1)y = e2x cos 3x. 1
= x− f (D)
Ans: y = e2x (c1 cos 3x + c2 sin 3x) f (D)
xe2x 1
+ sin 3x V.
6 f (D)
Example 2.53 yc = c1 cos 3x + c2 sin 3x

Solve (D + 1)2 y = x cos x. To find the particular integral,
Solution To find the complementary function, 1

yp = x sin x
we have to solve (D + 1)2 y = 0 D2 +9
Here f (D) = D2 + 9 f (D) = 2D
(m + 1)2 = 0 ⇒ m = −1, −1

1 1
yc = (c1 + c2 x) e−x = x− f (D) sin 2x
f (D) f (D)
2D 1
= x− 2 sin 2x
1 D +9 D +9 2
yp = x cos x
(D + 1)2 2D 1
= x− 2 sin 2x
D + 9 −4 + 9
Here f (D) = (D + 1)2 f (D) = 2(D + 1)
x 2.2 cos 2x
1 1 = sin 2x −
= x− f (D) (x cos x) 5 5 (−4 + 9)
f (D) f (D) x sin 2x 4 cos 2x x sin 2x 4
= − = − cos 2x
2 (D + 1) 1 5 5 (−4 + 9) 5 25
= x− cos x
(D + 1)2 −12 + 2D + 1
2 sin x x D−1
= x− = sin x − 2 sin x
D+1 2 2 D −1 y(x) = yc + yp
x cos x − sin x x sin 2x 4
= sin x + = c1 cos 3x + c2 sin 3x + − cos 2x.
2 2 5 25
The general solution is Example 2.55
Solve (D − 2)2 y = x2 sin x + e2x + 3.
y (x) = yc + yp = (c1 + c2 x)
[JNTU 2001S]
x 1
e−x + sin x + (cos x − sin x) . Solution yc = (c1 + c2 x)e2x
2 2
Example 2.54 1 1
P.I1 = x2 sin x = x(x sin x)
Solve (D2 + 9)y = x sin 2x. (D − 2) 2 (D − 2)2

1
Solution To find the complementary function, = x· (x sin x)
we have to solve D2 + 9 y = 0 (D − 2)2

The auxiliary equation is D−2
−2 (sin x) = xI1 − 2I2
m2 + 9 = 0 ⇒ m = ±3i (D − 2)4
1
I1 = (x sin x)
(D − 2)2 (D − 2)3 = D3 − 3 · 2 · D2 + 3 · 22 D − 23
1 (D − 2) = 11D − 2, Putting D2 = −1
= x· (sin x) − 2 · (sin x)
(D − 2) 2 (D − 2)4
1 1
= x· (sin x) − 2 · (sin x) (D − 2)4 = D4 − 4 · 2D3 + 6 · 22 D2 − 4 · 23
3 − 4D (D − 2)3
= −7 − 24D, Putting D2 = −1
3 + 4D 11D + 2
= x· 2 (sin x) − 2 · 2 2 (sin x)
3 −4 D2 2 11 D − 22
x
= (3 sin x + 4 cos x) Alternatively, P.I1 can also be obtained as follows:
25
2 1 1
+ (2 sin x + 11 cos x) P.I1 = x2 sin x = x2 (Imeix )
125 (D − 2) 2 (D − 2)2
1 1
I2 = (x sin x) = Im eix (by exponential shift)
(D − 2)3 (D + i − 2)2
1 1 1
=x (sin x) − 3 · (sin x) = Im eix x2
(D − 2)3 (D − 2)4 [(2 − i) − D]2

=
x
(11D + 2)(sin x) 1 D −2 2
−125 = Im eix 1 − x
(2 − i)2 2−i

3 1 1
− (7 − 24D)(sin x) = Im eix 1+ 2D
625 (2 − i)2 (2 − i)
x
=− (2 sin x + 11 cos x) 1
125 + 3D 2
+ · · · x2
3 (2 − i)2
+ (7 sin x − 24 cos x) (by Binomial theorem)
625
x2 3 + 4i 2 2 + 11i
P.I1 = xI1 − 2I2 = (3 sin x + 4 cos x) = Im eix x + 4x
25 25 125

2x −7 + 24i
+ (2 sin x + 11 cos x) + .6
125 625
2x
+ (2 sin x + 11 cos x)
125 1 2+i
=
6 2−i 5
− (7 sin x − 24 cos x) 1 1 3 + 4i
625 = =
1 (2 − i)2 3 − 4i 25
= (75x2 + 40x − 42) sin x 1 2 + 11i
625 =
(2 − i)3 125
+(100x2 + 220x + 144) cos x
1 −7 + 24i
=
(2 − i)4 625
2
1 2x x
P.I2 · e 2x
= e ·
(D − 2)2 2
1 3 = Im(cos x + i sin x)[(75 + 100i)x2
P.I3 · 3= +(40 + 220i) + (−42 + 144i)]
(D − 2)2 4
1
yp = P.I1 + P.I2 + P.I3 = [(75x2 + 40x − 42) sin x
625
+(100x2 + 220x + 144) cos x].
EXERCISE 2.9
y = y c + yp
x2 2x 3
= (c1 + c2 x)e2x + e +
2 4
1. (D + 1)2 y = x cos x. [JNTU 2003]
1
+ (75x2 + 40x − 42) sin x Ans: = (c1 + c2 x) e−x +
1
65 2
+(100x2 + 220x + 144) cos x (x sin x + cos x − sin x)
2. (D2 + 4)y = x sin x. [JNTU 2003] 12. y + 4y + 20y = 23 sin t − 15 cos;
Ans: y = c1 cos 2x + c2sin 2x y(0) = 0, y (0) = 1. [JNTU 2004(4)]
1 2 e−2t
+ x sin x − cos x Ans: y± = (2 cos 4t + sin 4t) + sin t − cos t
3 3 2
3. (D2 − 1)y = x sin 3x + cos

x. [JNTU 1997] 13. D3 − 1 y = ex + sin 3x + 2.
1 3
Ans: y = c1 ex +c2 e−x − cos 3x + x sin 3x [JNTU 2004(1)]
10 5 Ans: y =c1 ex +e−x/2
√ √
3 3
4. D2 + 1 y = x2 sin x. c2 cos x + c3 sin x
1 2 2
Ans: y = − cos x
2 x 1
1 3 + ex + (27 cos 3x − sin 3x) − 2
− 2x cos x − 3x cos x − 3x2 sin x 3 730
12

5. (D − 1)2 y = xex sin x. [JNTU 2002] 14. D2 + 1 x = t cos 2t x = x = 0 at t = 0.
Ans: y = (c1 + c2 x) ex − ex (2 cos x + x sin x) [JNTU 2002]
5 t 4
Ans: x (t) = − sin t − cos 2t + sin 2t
6. D2 − 1 y = x sin x + x2 ex . 9 3 9
1
Ans: y = c1 ex + c2 e−x − (x sin x + cos x)
2 15. D2 − 4D + 4 y = 8x2 e2x sin 2x.
x
+ ex 2x2 − 3x + 3 [JNTU 2001S]
12
2 Ans: y = (c1 + c2 x) e2x
7. D + 1 y = x sin x.
2
− 2e2x · 2x cos 2x + (x2 − 3/2) sin 2x
Ans: y = c1 cos x + c2 sin x
1 3
− 2x cos x−3x cos x − 3x2 sin x
12 2.2.2 Linear Equations with
2 Variable Coef f icients—
8. D − 1 y = x sin x + ex 1 + x2 .
[JNTU 1999S]
Euler–Cauchy Equations
1 (Equidimensional Equations)
−x
Ans: y = c1 e + c2 e − (cos x + x sin x)
x
2 An equation of the form
xex ex x 3 x2 x
+ + − +
2 2 3 2 2 f (xD)y ≡ (a0 xn Dn + a1 xn−1 Dn−1
2 + · · · + an−1 xD + an )y
9. D + 3D + 2 y = xex sin x.
ex = Q(x) (2.41)
Ans: y = c1 e−x + c2 e−2x + n
5 %
[5x(sin x − cos x) − 2 sin x + 5 cos x] or ar xn−r Dn−r y = Q(x)
r=0
10. D2 + 1 y = e−x + x3 + ex sin x.
[JNTU 1998] where ai are all constants and Q(x) is a function of x is
1 −x −ex called an Euler–Cauchy equation. Some authors call
Ans: y = c1 cos x + c2 sin x + e
2 5 it a homogeneous equation. Since the phrase is used
(2 cos x − sin x) + x3 − 6x for an equation with Q(x) ≡ 0, it leads to confusion
if we use it for Eq. (2.41). Equation (2.41) is rightly
e−x
11. D2 + 2D + 1 y = − 2 . called an equidimensional equation since the dimen-
x sion of each term with respect to the independent
Ans: y = (c1 + c2 x) e−x − e−x log x
variable is same.
Change the independent variable to z by putting The general solution is

⎫ c x
x = ez or z = log x so that ⎪
⎪ y= + . (8)
⎪
⎪ x 2
dy dy dz 1 dy ⎪
⎪
= = ⎪
⎪
dx dz dx x dz ⎪
⎪ Example 2.57
⇒ xDy = θy ⎪
⎪
⎪
⎪
⎪
⎪ d 2y dy
2
d y 2
1 d y dz 1 dy ⎪
⎬ Solve 2x2 + 3x − y = x. (1)
2
= − dx 2
dx
2
dx x 2dz dx x dz 2 (2.42)
1 d y dy ⎪
⎪
= 2 − ⎪
⎪ Solution
⎪
⎪
x dz 2 dz ⎪
⎪
⇒ x D y = θ(θ − 1)y
2 2 ⎪
⎪ Putx = ez or z = log x (2)
⎪
⎪
⎪
⎪ 2
By induction ⎪
⎪ dy d y
⎪
⎭ so that x = θy, x2 2 = θ(θ − 1)y (3)
x D y = θ(θ − 1) · · · (θ − r + 1)y
r r dx dx
d
where θ =
d d dz
where D = and θ =
dx dz
The given equation becomes
Substituting these into Eq. (2.41) the equation
reduces to a linear differential equation with constant [2θ(θ − 1) + 3θ − 1]y = ez
coefficients, which can be solved by the methods
or (2θ 2 + θ − 1)y = ez (4)
discussed above.
which is a linear equation with constant coefficients.
Example 2.56
To find the complementary function, we have to
dy
Solve x + y = x. (1) solve
dx
Solution (2θ 2 + θ − 1)y = 0 (5)
Put x = ez , xDy = θy (2) The auxiliary equation is

d d
where D = , θ= (3) 2m2 + m − 1 = 2m2 + 2m − m − 1 (6)
dx dz
= 0(m + 1)(2m − 1) = 0
The equation reduces to 1
m = −1 or yc = c1 e−z (7)
2
(θ + 1)y = e z
(4) 1 √
+c2 ez/2 = c1 + c2 x (8)
To find the complementary function, we solve x
(θ + 1)y = 0 (5)
1
The auxiliary equation is yp = ez
2θ 2 + θ − 1
1 ez x
m+1=0 ⇒ m = −1 = e z
= =
c 2×1 +1−1
2 2 2
yc = ce−z = (6)
x 1 √ x
The general solution is y = c1 + c2 x + .
To find the particular integral, x 2
Example 2.58
1 z 1 z x d 2y dy
yp = e = e = (7) Solve x2 + 3x + y = log x. (1)
θ+1 1+1 2 dx 2 dx
Solution which is a linear equation with constant coefficients.

Put x = ez or z = log x (2)
dy d 2y solve
so that x = θy, x2 2 = θ(θ − 1)y (3)
dx dx
(θ 2 − 2θ + 2)y = 0 (4)
d
where θ = .
dz The auxiliary equation is
m2 − 2m + 2 = (m − 1)2 + 1 = 0
[θ(θ − 1) + 3θ + 1]y = z or (θ + 1)2 y = z (4)
⇒ m − 1 = ±i ⇒ m = 1 ± i
which is a linear equation with constant coefficients. yc = (c1 cos z + c2 sin z)ez (5)
To find the complementary function, we have to = [c1 cos(log x) + c2 sin(log x)]x (6)
solve
(θ + 1)2 y = 0 (5) 1
yp = zez
The auxiliary equation is (θ − 1)2 + 1
1
(m + 1)2 = 0 ⇒ m = −1, −1 = ez z (exponential shift)
(θ + 1 − 1)2 + 1
1 −1
yc = (c1 + c2 z)e−z = (c1 + c2 log x) (6) = ez (1 + θ 2 ) z
x
= ez (1 − θ 2 + · · · )z
1 = zez = x log x
yp = z (7)
(θ + 1)2 The general solution is
= (1 + θ)−2 z
y = [c1 cos(log x) + c2 sin(log x)]x + x log x.
= (1 − 2θ + 3θ 2 − · · · )z
= z − 2 = log x − 2 Example 2.60
The general solution is Solve (x2 D2 + xD + 1)y = log x sin(log x). (1)
1
y = (c1 + c2 log x) + log x − 2. Solution
x
Example 2.59 Put x = ez or z = log x (2)
2
d y dy so that xD = θ, x2 D2 = θ(θ − 1)
Solve x2 2
− x + 2y = x log x. (1)
dx dx
Solution Put
⎫ [θ(θ − 1) + θ + 1]y = z sin z
x = ez or z = log x ⎪
⎪ (θ 2 + 1)y = z sin z (3)
dy d 2y ⎪
⎬
so that x = θy, x2 2 = θ(θ − 1)y θ (2)
dx dx ⎪ which is a linear equation with constant coefficients.
d ⎪
⎪
= ⎭ To find the complementary function, we have to
dz solve
The given equation becomes (θ 2 + 1)y = 0
[θ(θ − 1) − θ + 2]y = zez or
(θ 2 − 2θ + 2)y = zez (3) m2 + 1 = 0 ⇒ m = ±i
yc = c1 cos z + c2 sin z (4) Solution Put x = ez or z = log x

= c1 cos(log x) + c2 sin(log x) (5) so that
dy d 2y
x = θy, x2 2 = θ(θ − 1)y, (2)
To find the particular integral, dx dx
3
d y
x3 3 = θ(θ − 1)(θ − 2)y (3)
1 1 dx
yp = z sin z = Im 2 zeiz
θ +1
2 θ +1 The given equation becomes
1
= Im eiz 2 z (exponential shift) [θ(θ − 1)(θ − 2) + 2θ(θ − 1) + 2]y
(θ + 1)2 + 1 = 10 ez + e−z (4)

1 eiz θ −1 (θ 3 − θ 2 + 2)y = 10 eZ + e−z
= iz
Im e 2 z = Im 1+ z
θ + 2iθ 2iθ 2i which is a linear equation with constant coefficients.

eiz θ θ2 To find the complementary function, we have to
= Im 1 − + 2 − ··· z
2iθ 2i 4i solve 3
θ − θ2 + 2 y = 0
−i 1 1 θ
= Im e iz
− − ··· z
2 θ 2i 4 The auxiliary equation is
2
i z z 1 m3 − m2 + 2 = 0 ⇒ m = −1 is a root
= Im eiz − − −
2 2 2i 4
By synthetic division
1 z2 z
= Im(cos z + i sin z) i − +
8 4 4 m2 − 2m + 2 = (m − 1)2 + 1 = 0

1 z2 z ⇒m=1±i
= − cos z + sin z
8 4 4
yc = c1 e−z + (c2 cos z + c3 sin z)ez
1 1
= − (log x)2 cos(log x) 1
8 4 = c1 + [c2 cos(log x) + c3 sin(log x)] x (5)
x
1
+ (log x) sin(log x) To find the particular integral,
4
1
yp = 10 3 ez
θ − θ2 + 2
1
+10 3 e−z (6)
θ − θ2 + 2
y = c1 cos(log x) + c2 sin(log x)
= 10 3
1
ez
1 1 − 12 + 2
+ − (log x)2 cos(log x)
4 1
1 +10z 2 e−z
+ (log x) sin(log x). 3θ − 2θ
4 = 5ez + 2ze−z
1 2
8
cos(log x)is merged in the = 5x + log x (7)
x
complementary function. The general solution is
1
Example 2.61 y = c1 + [c2 cos(log x)
x
+c3 sin(log x)]x
d 3y 2
2d y 1 2
Solve x3 + 2x + 2y = 10 x + . (1) +5x + log x. (8)
dx3 dx2 x x
EXERCISE 2.10 d 2y dy
11. x2 2
+ x + y = x + log x.
Solve the following: dx dx
Ans: y = c1 cos(log x) + c2 sin(log x) + x
2
+ log x
2
d y 1
1. x2 2 − 2y = x2 + . d 2y dy
dx c
x 12. x2 − 4x + 6y = x.
Ans: y = c1 x2 + x2 + 13 x2 − 1x log x dx2 dx
Ans: y = c1 x2 + c2 x3 + 2x
d 3y d 2y dy
2. x2 − 4x 2 + 6 = 4.
dx 3 dx dx d 2y dy
13. x2 − x + 2y = x log x. [JNTU 1993]
[ Hint: Multiply by x] dx2 dx
2 Ans: y = x(c1 cos(log x) + c2 sin(log x)) + x log x
Ans: y = c1 + c2 x3 + c3 x4 + x
3
d 2y dy
d y 2
dy 14. x2 2
− 2x − 4y = x2 + 2 log x.
3. x2 + 2x − 20y = (x + 1)2 . dx dx
c 2
dx2 dx Ans: y = c1 x4 + x2 − x6 − 12 log x + 34
2
Ans: y = c1 x4 + c2 x15 − x14 − 9x − 20
1
d 2y dy
d 2y dy 15. x2 − x + y = log x. [JNTU 1995]
4. x2 2 − 4x + 6y = x2 . dx2 dx
dx dx Ans: y = (c1 + c2 log x)x + log x + 4
Ans: y = c1 x2 + c2 x3 − x2 log x
d 2y dy
d 2y dy 1 16. (1+x)2 +(1+x) +y = sin[2 log(1+x)].
5. x2 2 − x + y = [log x sin(log x) + 1]. dx2 dx
dx dx x [JNTU 1996]
Ans: y = x(c1 + c2 log x) + 100x1
Ans: y = c1 cos log(1 + x) + c2 sin log(1 + x) −
+ 125x cos(log x) + 125x sin(log x)
6 4 1
3
sin[2 log(1 + x)]
+ 25x
1
log x[4 cos(log x) + 3 sin(log x)]
d 3y d 2y dy 1
d 2y dy 1 17. x4 3
+ 2x3 2 − x2 + xy = ·
6. x2 2 + 3x + y = · dx dx dx x
dx dx (1 − x)2 [JNTU 2004]
Ans: y = x (c1 + c2 log x) + x log (1−x)
1 1 x
c1 1
Ans: y = + (c2 + c3 log x)x − 2
x 8x
d 2y dy
7. x2 2
+ x − 3y = x2 .
dx √dx √
Ans: y = c1 x 3 + c2 x− 3 + x2 2.2.3 Legendre’s Linear Equation
2
An equation of the form
2dy dy
8. x + 8x + 12y = x4 . n
dx2 dx % d n−r
4
Ans: y = c1 x−3 + c2 x−4 + x16 ar (a + bx) n−r
y
r=0
dxn−r
d 2y dy d ny n−1 d
n−1
y
9. x2 − 3x − 5y = sin(log x). = a0 (a + bx)n n
+ a 1 (a + bx) n−1
2
dx c dx dx dx
Ans: y = x1 + c2 x5 dy
+ · · · + an−1 (a + bx) + an y
+ 261 [2 cos(log x) − 3 sin(log x)] dx
= Q(x) (2.43)
d 2y dy
10. x2 + 7x + 13y = log x.
dx 2 dx where ar (r = 0, 1, · · · , n), a, b are constants and
Ans: y = x13 [c1 cos(2 log x) + c2 sin(2 log x)] Q(x) is a function of x is called a Legendre’s linear

+ 131 log x − 136 equation.
Put a + bx = ez or z = log(a + bx) so that To find the particular integral,
dy dy dz b 1 1
= = yp = · (e2z − 4)
dx dz dx a + bx 27 θ 2 − 4

dy 1 1 1
⇒ (a + bx) = bθy = z · e2z + · 4
dx 27 2θ 4
2 2z
d 2y b2 d y dy 1 ze
= − = −1
dx2 (a + bx)2 dz 2 dz 27 4
2
d y (3x + 2)2 log(3x + 2) 1
⇒ (a + bx)2 2 = b2 θ(θ − 1)y = +
dx 108 27
By induction, The general solution is
r
d y y = c1 (3x + 2)2 + c2 (3x + 2)−2
(a + bx)r = br θ(θ − 1) · · · (θ − r + 1)y
dxr (3x + 2)2 1
+ log(3x + 2) + .
Equation (2.33) now reduces to a linear equation 108 27
with constant coefficients which can be solved by
Example 2.63
the methods discussed above.
Solve
d 2y dy
Example 2.62 (1 + 2x)2 − 6(1 + 2x) + 16y = 8(1 + 2x)2 .
Solve dx2 dx
d 2y dy Solution This is a Legendre’s linear equation
(3x + 2)2 2
+3(3x + 2) −36y = 3x2 + 4x. (1)
dx dx Put (1 + 2x) = ez or z = log(1 + 2x) so that
Solution This is a Legendre’s linear equation
dy
Put 3x + 2 = ez or z = log(3x + 2) (1 + 2x) = 2θy,
so that dx
d 2y
dy d 2y (1 + 2x)2 2 = 22 θ(θ − 1)y
(3x + 2) = 3θy (3x + 2)2 2 dx
dx dx
= 32 θ(θ − 1)y (2)
[22 θ(θ − 1) − 6 · 2θ + 16] y = 8e2z
(θ 2 − 4θ + 4)y = 2x2z
[32 θ(θ − 1) + 3 · 3θ − 36] y ⇒ (θ − 2)2 y = 2e2z
z z
e −2 e −2
=3 +4 (3) To find the complementary function, we have to
3 3 solve (θ − 2)2 y = 0
1 2z The auxiliary equation is
(θ 2 − 4)y = (e − 4)
27 (m − 2)2 = 0 ⇒ m = 2, 2
yc = (c1 + c2 z)e2z
solve (θ 2 − 4)y = 0
The auxiliary equation is = [c1 + c2 log(1 + 2x)](1 + 2x)2
m2 − 4 = 0 ⇒ m = ±2
−2z
yc = c1 e + c2 e
2z
1
= c1 (3x + 2)2 + c2 (3x + 2)−2 yp = 2e2z
(θ − 2)2
1 be the complementary function. Since u and v are

= e2z = e2z z 2
(θ + 2 − 2)2 solutions of the homogeneous equation
= (1 + 2x)2 [log(1 + 2x)]2
y + Py + Qy = 0 (2.46)
We have
y = [c1 + c2 log(1 + 2x)](1 + 2x) 2
u + Pu + Qu = 0 (2.47)

+(1 + 2x)2 [log(1 + 2x)]2 .
v + Pv + Qv = 0 (2.48)
EXERCISE 2.11
We assume that a particular integral yp of Eq.
Solve the following: (2.44) is given by
1. [(5 + 2x)2 D2 −√6(5 + 2x)D + 8] y = 0. yp = Au + Bv (2.49)
Ans: y = [c1 cosh{ √
2 log(5 + 2x)}
+c2 sinh[ 2 log(5 + 2x)]] · (5 + 2x)2 where A and B are now considered as functions of x
to be determined.
2. [(2x + 1)2 D2 − 2(2x + 1)D − 12] y = 6x. Differentiating Eq. (2.49) with respect to x
c2 3 1
Ans: y = c1 (2x + 1)2 + − (2x + 1) +
2x + 1 16 4 dA dB
yp = Au + Bv + u +v
dx dx
3. [(x + 2)2 D2 − (x + 2)D + 1] y = 3x + 4.
3 Now we choose A and B such that
Ans: y = c1 + c2 log(x + 2) · (x + 2) + (x + 2)
2 dA dB
u +v =0 (2.50)
4. [(1 + x) D + (1 + x)D = 1] y
2 2 dx dx
= 4 cos [log(1 + x)] .
Ans: y = c1 cos log(1 + x) + c2 sin log(1 + x) Then yp = A u + B v (2.51)
+2 log(1 + x) sin log(1 + x)
Differentiating Eq. (2.51) with respect to x again
5. [(1 + x)2 D2 + (1 + x)D + 1]y
= sin 2[log(1 + x)]. [JNTU 1996] dA dB
yp = Au + Bv + u + v (2.52)
Ans: y = c1 cos[log(1 + x)] + c2 sin[log(1 + x)] dx dx
− 13 sin[2(log(1 + x))]
Since yp satisfies Eq. (2.44) we have from Eqs.
(2.46), (2.47) and (2.48)
2.2.4 Method of Variation of A(u + Pu + Qu) + B(v + Pv + Qv)
Parameters
dA dB
We will now consider the method of variation of +u + v =R
dx dx
parameters for finding the particular integral from
the complementary function and hence the general dA dB
⇒ u + v =R
solution of a linear non-homogeneous equation dx dx
by Eqs. (2.47) and (2.48).
y + Py + Qy = R (2.44) Since u(x) and v(x) are linearly independent
solutions over the interval I we have
where P, Q, R are functions of x.
u v
yc = C1 u(x) + C2 v(x) W (u, v)(x) = = 0 (2.53)
Let (2.45) u v

Hence, solving Eqs. (2.50) and (2.53) we have, vR
A=− dx (4)
W
dA/dx dB/dx −1
sin x cosec x
= = (2.54) = − dx = −x (5)
vR −uR
W 1
vR uR
⇒ A=− dx, B = dx (2.55) uR
W W A= dx
W

cos x cosec x
Substituting these values in Eq. (2.49) we get the = dx
particular integral yp . From Eqs. (2.45) and (2.49) 1
we obtain the general solution or complete solution. = log sin x (6)
Note 1 The complete solution is

Since the arbitrary constants C1 and C2 are replaced
by A and B in the complementary function and y = c1 cos x + c2 sin x − x cos x
are considered as functions of x for finding the + sin x log sin x. (7)
particular integral this method is called "variation of
parameters." (2) The given equation is
Note 2
The above method is applicable when the coeffi- (D2 + 1)y = x sin x (1)
cients are functions of x or constants.
P = 0, Q = 1, R = x sin x
Example 2.64
Solve the following equations by the method of To find the complementary function, we have to
variation of parameters. solve
d 2y (D2 + 1)y = 0 (2)
(1) + y = cosec x; (2) (D2 + 1)y = x sin x;
dx2
(3) (D2 + a2 )y = tan ax. [JNTU 2003] The auxiliary equation is
Solution
m2 + 1 = 0 ⇒ m = ±i
(1) The given equation is
(D2 + 1)y = cosec x (1) yc = c1 cos x + c2 sin x (3)

P = 0, Q = 1, R = cosec x
We take y = A cos x + B sin x, where A, B are
To find the complementary function, we have to functions of x and
solve (D2 + 1)y = 0
The auxiliary equation is u(x) = cos x, v(x) = sin x
cos x sin x
m2 + 1 = 0 ⇒ m = ±i W (u, v)(x) = = 1 ± 0 (4)
− sin x cos x
yc = c1 cos x + c2 sin x (2)

vR sin x(x sin x)
We take y = A cos x + B sin x, where A, B are A=− dx = − dx
W 1
functions of x and
−1 1
= xdx+ x cos 2xdx
u(x) = cos x, v(x) = sin x 2 2
cos x sin x x2 sin 2x cos 2x
W (u, v)(x) = = 1 = 0 (3) = − +x +
− sin x cos x 4 4 8

uR cos x(x sin x) vR sin ax tan ax
B= dx = − dx A= − dx = − dx
W 1 W a

1 sin2 ax
= x sin 2xdx = − dx
2 a cos ax

x cos 2x 1 sin 2x 1 − cos2 ax
= − − − = − dx
2 2 2 4 a cos ax
−x cos 2x 1
1 1
= + sin 2x = cos axdx − sec axdx
4 8 a a
1 1
The complete solution is = 2
sin ax − 2 log (sec ax + tan ax)
a a
y = c1 cos x + c2 sin x (5)
2 uR cos ax tan ax
x x 1 B= dx = − dx
+ − + sin 2x + cos 2x cos x W a

4 4 8 1
= sin ax dx
x 1 a
+ − cos 2x + sin 2x sin x
4 8 1
= − 2 cos ax
y = c1 cos x + c2 sin x (6) a
2
x x The complete solution is
+ − cos x − sin x . (7)
4 4 1
y = c1 cos ax + c2 sin ax −
1 a2
cos x is merged with the (8)
8 cos ax log(sec ax + tan ax). (4)

complementary function.
Example 2.65
d 2y
(3) The given equation is Solve the equation + y = sec x
dx2
[JNTU 1995, 1996, 1998, 1999]
(D2 + a2 )y = tan ax (1)
Solution The given equation is
P = 0, Q = a2 , R = tan ax
(D2 + 1)y = sec x (1)
solve (D2 + a2 )y = 0 The complementary function is
The auxiliary equation is yc = c1 cos x + c2 sin x; P = 0, Q = 1, R = sec x
m2 + a2 = 0 ⇒ m = ±ia Let y = A cos x + B sin x
where A, B are functions of x and
yc = c1 cos ax + c2 sin ax (2)
u(x) = cos x, v(x) = sin x
Let y = A cos ax + B sin ax where A, B are W (u, v)(x) =
cos x sin x
= 1 = 0
functions of x and − sin x cos x

u(x) = cos ax, v(x) = sin ax vR sin x sec x
A=− dx = − dx
W 1
cos ax sin ax = log cos x
W (u, v)(x) = = a = 0
−a sin ax a cos ax uR cos x sec x
(3) B= dx = − dx = x
W 1
The complete solution is The auxiliary equation is

y = c1 cos x + c2 sin x + cos x log x + x sin x.
m2 + 3m + 2 = 0 ⇒ m = −1, −2
Example 2.66
The complementary function is
Solve the equation (D2 − 2D)y = ex sin x.
yc = c1 e−x + c2 e−2x ; u(x) = e−x , v(x) = e−2x
we have to solve (D2 − 2D)y = 0 Let y = Ae−x + Be−2x where A, B are functions of
The auxiliary equation is x.
m2 − 2m = 0 ⇒ m = 0, m = 2; e−x e−2x
W (u, v)(x) = = −e−3x = 0
−e−x −2e−2x
P = −2, Q = 0, R = ex sin x
yc = c1 + c2 e2x −2x x
vR e (e + x2 )
u(x) = 1, v(x) = e 2x A=− dx = − dx
W −e−3x

Let y = A + Be where A, B are functions of x.
2x
= e2x dx + x2 ex dx
1 e2x 1 2x
W (u, v)(x) = = 2e2x = 0 =
e + ex (x2 − 2x + 2)
0 2e2x
2
−x x
uR e (e + x2 )
B= dx = dx = − e3x dx
vR e2x ex sin x W −e−3x
A=− dx = − dx
W 2e2x − x2 e2x dx

1
=− ex sin x dx −1 3x
2 x 1 2 2 2
= e −e x − x+
ex 3 2 4 8
= − (sin x − cos x)
4 The complete solution is
uR
B= dx

W
−x −2x 1 x2 3
1.ex sin x 1 y = c1 e + c2 e + ex + e−x
− x .
= dx = e−x sin x dx 6 2 2
2e2x 2
7 −x
e−x e is merged in the complementary function.
= (− sin x − cos x) 4
4
The complete solution is EXERCISE 2.12
x Solve the equations by the method of variation of
e e−x
y = c1 + c2 e2x + − (sin x − cos x) · 1 + parameters:
4 4
(− sin x − cos x)e 2x
1. (D2 + 1)y = cosec x cot x.
ex Ans: y = c1 cos x + c2 sin x
y = c1 + c2 e2x − sin x. − cos x log(sin x) − x sin x
2
Example 2.67 2. (D + 1)2 y = e−x log x.
Solve the equation D + 3D + 2)y = e + x .
2 x 2
Ans: y = c1 e−x + c2 xe−x
x2 e−x
Solution Here P = 3, Q = 2, R = ex + x2 + (1 − 2 log x) + xe−2x (log x − 1)
4
3. (D2 + a2 )y = a2 sec ax. The system Eq. (2.57) is solved by elimination and
Ans: y = c1 cos ax + c2 sin ax + ax sin ax we get
+ cos ax log (cos ax) g2 (D)f (t) − f2 (D)g(t)
x= ,
4. (D2 − 3D + 2)y = xex + 2x. f1 (D)g(t) − g1 (D)f (t)
x2 3 y=
Ans: y = c1 ex + c2 e2x − ex − xe−x + x +
2 2 where = f1 (D)g2 (D) − f2 (D)g1 (D) = 0.
5. (D + 4)y = tan 2x.
2
[JNTU 2003, 2002, 1997, 1994]

Example 2.68
Ans: y = c1 cos 2x + c2 sin 2x dx dy
cos 2x Solve = 3x + 2y, = −5x − 3y.
− log (sec 2x + tan 2x) dt dt
4
Solution In operator notation, the equations are
6. (x − 1)D2 − xD + 1 y = (x − 1)2 .
x (D − 3)x − 2y = 0 (1)
[Hint: e , x are basis solutions.]
Ans: y = c1 ex + c2 x − (1 + x + x2 ) 5x + (D + 3)y = 0 (2)
Multiplying Eq. (1) by D + 3 and Eq. (2) by −2,

e3x
7. (D2 − 6D + 9)y = . we get
x2
Ans: y = c1 e3x + c2 xe3x − e3x log x (D + 3)(D − 3)x − (D − 3)2y = 0
−10x − (D + 3)2y = 0
8. (D3 + D)y = cosec x. On subtraction, (D2 + 1)x = 0
Ans: y = c1 + c2 cos x + c3 sin x The auxiliary equation is
− log (cosecx + cot x) − cos x log sin x − x sin x
m2 + 1 = 0 ⇒ m = ±i
9. (D − 2D + 2)y = e tan x.
2 x
x = c1 cos t + c2 sin t
Ans: y = ex (c1 cos x + c2 sin x)
− ex cos x log (sec x + tan x) Substituting in Eq. (1)
1
y= (D − 3)(c1 cos t + c2 sin t)
2.2.5 Systems of Simultaneous 2
Linear Differential Equations 1
= [−(c1 sin t + c2 cos t) − 3c1 cos t − 3c2 sin t]
with Constant Coef f icients 2
In many applied mathematical problems, we c2 − 3c1
= A cos t + B sin t A= ,
encounter systems of simultaneous linear differen- 2
tial equations involving two (or more) dependent c1 + 3c2
B = 1− ·
variables x and y and one independent variable t, as 2
follows: Example 2.69
) dx dy
f1 (D)x + f2 (D)y = f (t) Solve + 5x − 2y = t, + 2x + y = 0.
(2.57) dt dt
g1 (D)x + g2 (D)y = g(t) x = 0, y = 0 when t = 0
where f1 , f2 , g1 , g2 are rational integral functions of Solution The given system of equations is
d
D= with constant coefficients and f , g are given (D + 5)x − 2y = t (1)
dt
functions of t. 2x + (D + 1)y = 0 (2)
D + 5 −2 c2 1
= = D2 + 6D + 9 (3) 0 = c1 − +
2 D+1 2 27
= (D + 3)2 (4) 4
0 = c1 +
27
To obtain x, y respectively replace c1 , c2 in 4
by c1 = −
27
2 8 2 2
f (t) t c2 = 2c1 + =− + =−
= 27 27 27 9
g(t) 0
t −2
x = =1+t (5)
0 D+1 1 1
x=− (1 + 6t)e−3t + (1 + 3t)
27 27
D+5 t 2 1
y = = −2t (6) y = − (2 + 3t) + (2 − 3t).
2 0 27 3
(D + 3)2 x = 1 + t (7) Example 2.70
(D + 3)2 y = −2t (8) dx dy

Solve = 5x + y, = y − 4x.
dt dt
Solution The given system of equations is
(m + 3)2 = 0 ⇒ m = −3, −3 (repeated)
(D − 5)x − y = 0 (1)
4x + (D − 1)y = 0 (2)
yc = (c1 + c2 t)e−3t
Operate Eq. (1) by (D − 1) and add to Eq. (2)
The particular integral is
1 [(D − 1)(D − 5) + 4]x = 0
yp = −2
(D + 3)2
⇒ (D − 3)2 x = 0
2 D −2 x = (c1 + c2 t) e3t
t=− 1+ (3)
9 3

2 2D From Eq. (1)
t=− 1− + ··· t
9 3
y = −5x + Dx
2 4
=− t+ = −5(c1 + c2 t)e3t (4)
9 27
2 4 +(0 + c2 + 3c1 + 3c2 t)e3t
∴ y = (c1 + c2 t)e−3t − t +
9 27 y = (−2c1 + c2 − 2c2 t)e3t. (5)
Substituting in Eq. (2)
Example 2.71
2x = −(D + 1)y = −y − Dy
Solve D2 x + y = sin t (1)
2 4
= −(c1 + c2 t)e−3t + t − x + D2 y = cos t. (2)
9 27
2
−(0 + c2 · 1 − 3c1 − 3c2 t)e−3t + Solution Applying D2 on Eq. (1)
9
c2 1 1 D4 x + D2 y = − sin t
x = (c1 − + c2 t)e−3t + t +
2 9 27
t=0 ⇒ x=y=0 ⇒ (D4 − 1)x = − cos t + sin t (3)
The roots of the auxiliary equations are Ans: x = (c1 cos 6t + c2 sin 6t)e2t
e2t
m = 1, −1, i, −i [by Eq. (2)] y = [(6c1 − c2 ) sin 6t − (c1 + 6c2 ) cos 6t]
2
dx dy
xc = c1 et + c2 e−t + c3 cos t + c4 sin t 2. = −ay, = ax.
dt dt
1 Ans: x = c1 cos at − c2 sin at
xp = − 2 (cos t + sin t) y = c1 sin at − c2 cos at
(D − 1)(D2 + 1)
−1 t
=− · [cos(t + π/2) + sin(t + π/2)] 3. (D + 2)x + (D + 1)y = t; 5x + (D + 3)y = t 2 .
−2 2
t c1 − 3c2 3c1 + 3c2
= (sin t − cos t) Ans: x = sin t − cos t
4 5 5
y = sin t − D2 x −t +t+3
2
= sin t − c1 et − c2 e−t + c3 cos t + c4 sin t

y = c1 cos t + c2 sin t + 2t 2 − 3t − 4
1 1 t t
− cos t − sin t + sin t − cos t
2 2 4 4 dx dy
4. = 3x + 8y; = −x − 3y;
= −c1 et − c2 e−t + c3 cos t + c4 sin t dt dt
1 with x(0) = 6, y(0) = −2.
+ (2 + t)(sin t − cos t).
4
Ans: x = 4et + 2e−t ; y = −et − e−t
Example 2.72
dx dy
dx 5. = 2y − 1, = 1 + 2x.
Solve + y = et (1) dt dt
dt 1
Ans: x = c1 e2t + c2 e−2t − ;
dy 2
− x = e−t (2) −2t 1
dt y = c1 e − c2 e +
2t
2
Solution
6. (D + 6)y − Dx = 0; (3 − D)x − 2Dy = 0; with
d 2 x dy x = 2, y = 3 when t = 0.
+ = et ⇒ (D2 + 1)x = et − e−t (3)
dt 2 dt Ans: x = 4e2t − 2e−3t ; y = e2t + 2e−3t
)
xc = c1 cos t + c2 sin t
(4) 7. (D + 2)x + (D − 1)y = − sin t;
et −e−t
xp = 1
D2 +1
(et − e−t ) = 2 (D − 3)x + (D + 2)y = 4 cos t.
3 2 1
Ans: x = c1 e−t/s + sin t − cos t;
x = c1 cos t + c2 sin t + sinh t 5 5 5
y = c1 e−t/s + sin t + cos t
et + e−t
y = et − Dx = et + c1 sin t − c2 cos t −
2 8. D2 y = x − 2, D2 x = y + 2.
y = c1 sin t − c2 cos t + sinh t. (5) Ans: x = c1 sin t + c2 cos t + c3 et + c4 e−t + 2;
y = −c1 sin t − c2 cos t + c3 et + c4 e−t − 2
9. (D + 1)x + (D − 1)y = et ; (D2 + D + 1)x +

EXERCISE 2.13 (D2 − D + 1)y = t 2 .
Solve the following pairs of simultaneous linear
1 2 1 t
equations: Ans: x = t −t+ e;
2 2
dx dy 1 3 t
1. = x − 2y, = 5x + 3y. y = t2 + t − e
dt dt 2 2
2.3 Applications of Linear dx

= c1 μ sin μt + c2 μ cos μt (2.60)
Differential Equations— dt
Mechanical and Electrical
Suppose the particle P is initially at A where OA =
Oscillatory Circuits and
a. Then x = a at t =0 so that we have a = c1 from
Deflection of Beams
Eq. (2.59).
Linear differential equations find numerous appli- dx
cations in mechanical systems, electrical circuits, Also, = 0 at t = 0 ⇒ c2 = 0 from Eq.
dt
loaded beams, etc. We shall observe that many (2.60).
mechanical and electrical systems will be described Thus, we get the displacement of P given by
by the same differential equations. Hence, to study
the behaviour of a complicated mechanical sys- x(t) = a cos μt (2.61)
tem we can more conveniently study an equivalent
electrical system and obtain results. and the velocity
dx
v= = −aμ sin μt = a2 − x2 (2.62)
2.3.1 Free Oscillations in dt
Mechanical System Now, the maximum distance of P from O is a. It
(1) Simple Harmonic Motion of a Particle is called the amplitude of the oscillation. The time of
2π
When the acceleration of a particle P is proportional one complete oscillation is T = and it is called the
μ
to its displacement from a fixed point O, and is always period of oscillation. The reciprocal of this is called
directed towards O, the motion of the particle is the frequency of oscillation. Thus, the frequency =
called a simple harmonic motion (SHM) (Fig. 2.1). 1 μ
= .
period 2π
μ2x
(2) Free Oscillations of a Spring
A' x p Consider a mass m, suspended from an end A of a

O A
spring of negligible weight, whose other end O is
a fixed. Let e = AB be the elongation due to the mass
m and k be the restoring force per unit extension in
Figure 2.1 A particle executing simple harmonic the spring due to its elasticity. Then, for equilibrium
motion at B we have
mg = ke (2.63)
If the displacement of P from O is x, then the
differential equation governing the motion is where g is the acceleration due to gravity.
Let the mass m at P at any time t where BP = x be
d2 displaced. Then the motion starts and the acceleration
= −μ2 x (2.58) d 2x
dt 2 of the mass m is 2 . The forces acting on the mass
dt
where μ2 is the constant of proportionality. The are its own weight mg and the restoring force K(e+x)
general solution of Eq. (2.58) is directed upwards (Fig. 2.2). By Newton’s second law
we have
x(t) = c1 cos μt + c2 sin μt (2.59)
d 2x
m = mg − K(e + x) = −kx
Differentiating (2.59) w.r.t. ‘t’ we get the velocity dt 2
of P as by (2.63)
o c
A L
e
B Figure 2.3 Free oscillations in electrical circuits
x
p
2.3.3 Damped Oscillations in
mg Mechanical System
If the motion of the mass m is subjected to a resisting
Figure 2.2 Free oscillation of a spring
force then the oscillations are called damped. The
damping force may be either a constant or propor-
d 2x dx
or + μ2 x = 0 (2.64) tional to the velocity, r , where r is a constant.
dt 2 dt
where μ2 = k/m. Eq. (2.64) represents simple har- Now, the equation of motion becomes
monic free oscillations of a spring with its centre at
B. the period T of oscillation is given by o
2π
T = = 2π (m/k) = 2 (e/g). (2.65)
μ
2.3.2 Free Oscillations in Electrical

Circuits
e
Consider an electrical circuit containing a condenser
of capacity c and inductance L. If q is the electri- x
damper
damper
cal charge on the plate of the condenser c and I is

the current in the circuit at any time t then the volt-
di d 2q
age drop across the inductance L is L = L 2 and mg
dt dt
q
the voltage drop across c is . If there is no electro- Figure 2.4 Damped oscillations of mechanical
c
motive force (e.m.f) applied in the circuit, then by system
Kirchhoff ’s first law, we have
d 2x dx dx
d 2q q d 2q m = mg − K(e + x) − r = −Kx − r
L 2 + =0 or + μ2 q = 0 (2.66) dt 2 dt dt
dt c dt 2
(2.67)
where μ2 = LC 1
. Now, comparing Eq. (2.66) with Eq. (∵ mg = Ke)
(2.64) obtained for the free oscillations of a spring, d 2x dx 2
we observe that they are same except that the dis- or + 2λ μ x = 0 (2.68)
dt 2 dt
placement x is replaced by the change q. Thus, the
where μ2 = k/m (2.69)
discharge of a condenser through an inductance L is
same as the motion of a mass m at the end of a spring. and λ = r/2m (2.70)
2.3.4 Damped Oscillations in where a = λ2 − μ2 is a real number and c1

Electrical Circuits and c2 are arbitrary constants, whose values can be
Consider the discharge of a condenser c through determined using the initial conditions.
an inductance L and resistance R. The potential Let
dq
q di d 2q q = q0 and i = = 0 at t = 0 (2.79)
drops across C, L and R are , L = L 2 and dt
c dt dt
dq Differentiating Eq. (2.78) w.r.t. t we get
Ri = R , where q is the charge in the condenser
dt dq
at any time t. Hence, by Kirchhoff ’s law = c1 (a − λ)eat − c2 (a + λ)e−at (2.80)
dt
Substituting the initial conditions (2.79) into Eqs.
d 2q dq q (2.78) and (2.80), we get
L 2 +R + =0
dt dt c
d 2q dq c1 + c2 = q0 ;
or + 2λ + μ2 q = 0 (2.71) c1 (a − λ) − c2 (a + λ) = 0
dt 2 dt
c1 c2 q0
R 1 ⇒ = = (2.81)
λ= , μ2 = (2.72), (2.73) a+λ a−λ 2a

2L LC sum of numerators
Compare Eqs. (2.68) and (2.71). The two lin- sum of denominators
ear differential equations are same except that the
Therefore,
displacement x is replaced by the charge q. ⎫
q0 q0 λ ⎪
c1 = (a + λ) = 1+ ⎪
⎬
Solution of the Differential Equation
2a 2 λ2 − μ2
q0 q0 λ (2.82)
c2 = (a − λ) = 1− ⎪
⎪
Let us solve the linear differential equation ⎭
2a 2 λ2 − μ2
Now the solution becomes
d 2q dq √
2
+ 2λ + μ2 q = 0 q = 2 q0 e
1 −λt
1+ √ λ 2
e λ −μ t
2
dt dt λ2 −μ2
or (D2 + 2λD + μ2 )q = 0 (2.74) √
+ 1− √ λ
e − λ2 −μ2 t
(2.83)
d λ −μ
2 2
where D ≡ is the differential operator and λ, μ2
dt Case (ii) λ = μ
have been defined at (2.69), (2.70); and (2.72), (2.73).
In this case, μ1 = −λ is a repeated root of the
The auxiliary equation is m2 + 2λ + μ2 = 0 whose
auxiliary equation. Therefore, the general solution
roots are
is
m1 = −λ + λ2 − μ 2 (2.75) q = (c1 t + c2 ) e−λt (2.84)
m2 = −λ − λ2 − μ2 (2.76) Differentiating q w.r.t. t, we get
dq
Three cases arise: i= = c1 e−λt − λ (c1 t + c2 ) e−λt (2.85)
(i) λ > μ (ii) λ = μ (iii) λ < μ (2.77) dt
Let
Case (i) λ > μ dq
q = q0 and i= = 0 at t = 0 (2.86)
In this case, the roots are real and distinct; and the dt
general solution of Eq. (2.74) is We therefore obtain from Eqs. (2.84) and (2.85)
q = e−λt (c1 eat + c2 e−at ) (2.78) c2 = q0 and c1 − λc2 = 0 ⇒ c1 = λq0

The solution in this case is The presence of the trigonometric factor in the
above equation shows that the motion is oscillatory.
q = q0 (λt + 1) e−λt (2.87) The variable amplitude of oscillation is
This is a case of critical damping.
λ2 q0 μ
q0 1 + 2 e−λt = e−λt (2.97)
Case (iii) λ < μ μ μ −λ
2 2
In this case, the roots of the auxiliary equation are m1 ,

which decreases with time and the period of oscilla-
m2 = −λ±ib b = μ2 − λ2 which are a complex
tion is
conjugate pair; and so, the general solution of the
2π 2π
differential equation is T = = (2.98)
b μ2 − λ2
q = (c1 cos bt + c2 sin bt) e−λt (2.88)
But the period of free oscillations is
Differentiating w.r.t. t this gives
2π
dq T = (2.99)
i= = b (−c1 sin bt + c2 cos bt) e−λt μ
dt
−λ (c1 cos bt + c2 sin bt) e−λt (2.89) 2π 2π
Since b = μ2 − λ2 < μ, >
b μ
The initial conditions are 2π 2π
At t = 0, ⇒ > ⇒ T > T .
μ2 − λ2 μ
dq This shows that the effect of damping is to increase
q = q0 and i = =0 (2.90)(2.91)
dt the period of oscillation and the motion ultimately
dies down. Such a motion is called a damped oscil-
From the first condition, we get latory motion.
c1 = q0 (2.92)
2.3.5 Forced Oscillation (Without
The second condition gives Damping) in Mechanical
System
0 = b (−c1 .0 + c2 .1) − λ (c1 + c2 .0)
Suppose an external force is imposed on the system,
λ λq0
⇒ c2 = q 0 = (2.93) for instance, let the point of support of the spring is
b μ2 − λ 2 also vibrating with an external force Q cos nt, say,
The solution is then the resulting motion is called forced oscillatory
system (without damping). The equation of motion
λ is
q = q0 cos bt + sin bt e−λt (2.94)
b
d 2x
λ m = mg − K(e + x) + Q cos nt (2.100)
writing r cos α = 1, r sin α = so that dt 2
b d 2x K Q
* ⇒ 2 + x = cos nt
λ2 λ dt m m
r = 1 + 2 and α = tan−1 (2.95) (mg = ke)
b b 2
or D + μ2 x = E cos nt (2.101)
the above result can be written as
d
q = q0 r [cos bt · cos α + sin bt · sin α] e−λt D≡
* dt

λ2 λ −λt where μ2 = k/m and E = Q/m (2.102)
= q0 1 + 2 cos bt − tan−1 e (2.96)
b b
o where μ2 = LC 1
and E = QL . This equation is same
as the one obtained for mechanical system except
that x is now replaced by the charge q.
Solution of the differential equation
Since the auxiliary equation of the differential equa-

Q cos nt tion is m2 + μ2 = 0 whose are m = ±iμ (complex
e conjugate pair), the complementary function is
K (e + x)
yCF = c1 cos μt + c2 sin μt

mg = A cos(μt − B) (2.105)
Figure 2.5 Forced oscillation of mechanical where A = c12 + c22 and B = tan−1 c1 /c2 and the
system particular integral is
2.3.6 Forced Oscillation (Without

1
Damping) in Electrical yp1 = E cos nt
System D 2 + μ2
⎧
⎪
⎪ E
The electrical analogue of the above mechanical ⎨ 2 cos nt, if μ = n (2.106)
μ − n2
system is given by an L–C circuit with battery of =
⎪
⎪ Et
alternating e.m.f. Q cos nt. The governing equation is ⎩ sin nt, if μ = n (2.107)
2n
where y stands for x in the case of mechanical system
and q in the case of electrical systems.
When μ = n the general solution is
~
E
Q cos nt y = A cos(μt − B) + cos nt (2.108)
μ2 − n 2
Figure 2.6 Forced oscillation in an electrical
system This shows that the variation of q is obtained by the
superposition of two oscillatory motions. The term
A cos(μt−B) represents the free oscillation of natural
2π
di q period (when no external force is applied) and the
L + = Q cos nt μ
dt c E
d 2q q term 2 cos nt represents the forced oscillation
or L 2 + = Q cos nt (2.103) μ − n2
dt c 2π
due to an external force of period .
dq r
where we have replaced i by , q being the charge When μ = n, the frequency of the system is the
dt
in the condenser at any time t. The equation can be same as the natural frequency of the system. This
written as phenomenon is known as “resonance” and, in this
d 2q case, the solution becomes
+ μ2 q = E cos nt
dt 2 Et
or (D2 + μ2 )q = E cos nt (2.104) y = A cos(μt − B) + sin μt (2.109)
2μ
where A and B are arbitrary constants. This solution to forced oscillation then the equation of motion
can be put in the form becomes
Et
y = c1 cos μt + c2 sin μt + sin μt o
2μ

Et Q cos nt
= c1 cos μt + c2 + sin μt (2.110)
2μ
Et
If we write c1 = r sin α and c2 + = r cos α
2μ
then
e
y = r sin(μt + α) (2.111)
K (x + e)
where

x
Et 2 Damper Damper
r = c12 + c2 + r dx/dt
2μ
mg
Et
and α = tan−1 c1 / c2 + (2.112)
2μ Figure 2.8 Forced oscillation with damping of
Thus, in the case of resonance, the oscillations are of mechanical system
natural period 2π/μ while the amplitude is

Et 2 d 2x dx
c12 + c2 + m = mg − K(e + x) − r + Q cos nt (2.113)
2μ dt 2 dt
The amplitude increases with time and may become Since mg = Ke in the equilibrium position, we
very large causing over-strain; and the system may have
break-down. But, in practice, the occurrence of such
d 2x dx
a situation is rare. Although there is the possibility m = −Kx − r + Q cos nt
of the amplitude becoming dangerously large, there dt dt
is always some resistance present in the system the or (D2 + 2λD + μ2 )x = E cos nt (2.114)
variation q (or x) with time t is shown in Fig. 2.7.
where
y 2λ = r/m, μ2 = K/m and E = Q/m (2.115)
o 2.3.8 Forced Oscillation with

t
Damping in Electrical System
In the case of electrical system, the equation of
Figure 2.7 Showing the variation of q or x with variation of charge in the condenser is
time t
di q
L + Ri + = Q cos nt
dt c
2.3.7 Forced Oscillation with
Damping in Mechanical or (D2 + 2λD + μ2 )q = E cos nt (2.116)
System
where
If an additional damping force proportional to the
1
dx 2λ = R/L, μ2 = and E = Q/L
velocity, namely, r is added to the system subjected LC
dt
C R L Example 2.73
A mass weighing 4.9 kg is suspended from a spring.
A pull of 10 kg will stretch it to 5 cm. The mass
is pulled down 6 cm below the static equilibrium
position and then released. Find the displacement of
~
the mass from its equilibrium position at time t sec,
Q cos nt
the maximum velocity and the period of oscillation.
Figure 2.9 Forced oscillation with damping of an
Solution Let O be the fixed end and A, the free
electrical system
end of the spring. Suppose the stiffness of the spring
(spring constant) is K.
Solution of the differential equation
The auxiliary equation is m2 + 2λm + m2 = 0. It has o

roots m1 , m2 = −λ ± a where a = λ2 − μ2 .
yCF = (c1 eat + c2 e−at )e−λt (2.117)

A
The system is oscillatory or non-oscillatory e
according as λ < μ or λ ≥ μ. The complementary B
function represents free oscillations of the system x
because of the presence of the factor e−λt which die T
down as t → ∞. Now, the particular integral is p 0.06 m
1 w
yP1 = E cos nt
D2 + 2λD + μ2
E C
= cos nt
−n2 + 2λD + μ2
2λD − (μ2 − n2 ) Figure 2.10
= E 2 2 cos nt
4λ D − (μ2 − n2 )2 It is given that a pull of 10 kg.wt. stretches the
2λD − (μ2 − n2 ) spring by 0.05 m.
= E cos nt (2.118)
−4λ2 n2 − (μ2 − n2 )2
E 10 = 0.05 K ⇒ K − 200 kg/m (1)
= [(μ2 − n2 ) cos nt
(μ − n )2 + 4λ2 n2
2 2
If e(= AB) is the elongation in the mass weighing
+2nλ sin nt] 4.9 kg hanging in equilibrium, then
E
= cos(nt − θ)(2.119) 4.9
(μ − n2 )2 + 4λ2 n2
2 4.9 = Ke = 200e ⇒ e = = 0.0245 m (2)
200
where θ = tan−1 [2nλ/(μ2 − n2 )] Now, the mass is pulled down to c, where BC =
Thus, as t increases the free oscillations die out 0.06 m. After t sec of its release from c if the mass
while the forced oscillations persist which gives rise is at P where BP = xm, then the forces acting on the
to steady state of motion. We may remark that the mass are
impressed force need not necessarily be a periodic
(1) its weight w = 4.9 kg; downwards (3)
function as taken here. We can solve the differen-
tial equation even if the function is any other than (2) the restoring force T = K(e + x)
periodic. = 200(0.0245 + x) kg; upwards (4)
The equation of motion of the mass is Example 2.74

A mass m suspended from one end of a spring is
d 2x subjected to a periodic force f = f0 sin at in the
m 2 =w−T (5)
dt direction of its length. The force f is measured
positive vertically downwards and time t = 0 m
4.9 d 2 x is at rest. If the spring constant is K, prove that
⇒ = 4.9 − 200(0.0245 + x) the displacement of m at timet is given by x =
g dt 2
f0 a
4.9 d 2 x sin at − sin pt where p2 = x/m.
⇒ = −200 (∵ g = 9.8 m/sec2 ) m(p − a )
2 2 p
9.8 dt 2 Neglect the damping effects.
d 2x
⇒ 2 + a2 x = 0 (a2 = 400) (6)
dt Solution Let x be the displacement of the mass
m at any time t.
The auxiliary equation m2 + a2 = 0 gives two
The equation of motion is
complex roots:
d 2x
m1 = m2 = ±ai (7) m = −Kx + f0 sin at
dt 2
f0
The general solution is or (D2 + p2 )x = sin at (s/m = p2 ) (1)
m
x = c1 cos at + c2 sin at The auxiliary equation D2 + p2 = 0 has roots
= c1 cos 20t + c2 sin 20t (8) D = ±ip
The initial conditions are: when t = 0
f0 sin at
x(t) = c1 cos pt + c2 sin pt + (2)
(1) displacement x = 0.06 (9) m p2 − a2
dx The initial conditions are: when t = 0
(2) velocity =0 (10)
dt (i) x = 0 and (3)
Condition (1) yields the value of c1 as dx
(ii) =0 (4)
dt
c1 = 0.06 (11)
Condition (1) implies that c1 = 0.
Differentiating x, we get Differentiating (1) w.r.t t we get
dx dx f0 a
= 20 c1 sin 20t + 20 c2 = 0 = −pc1 sin pt + pc2 cos pt + cos pt
dt dt m p − a2
2
⇒ c2 = 0 (12) (5)
Condition (ii) gives
Substituting these value of c1 and c2 we get the f a
displacement at time t of the mass as pc2 + =0
m p − a2
2
x(t) = 0.06 cos 20t (13) a f0

⇒ c2 = − (6)
p m (p2 − a2 )
The maximum velocity of the system is
Substituting these values of c1 and c2 in (1) we get
μ(amplitude) = 20 × 0.06 = 1.2 m/sec (14)
a f0 f0 sin at
The period of oscillation x(t) = − sin pt +
p m (p2 − a2 ) m p2 − a 2

2π π f0 a
= = = 0.314 sec (15) or x(t) = sin at − sin pt (7)
μ 10 m(p2 − a2 ) p
Example 2.75 The equation of motion is

A mass of 200 gm is tied at the end of a spring which
extends to 4 cm under a force 196000 dynes. The 200 d 2 x 2000 dx
= w−T −
spring is pulled 5 cm and released. Find the displace- g dt 2 g dt
ment, t sec after the release if there is a damping force 2000 dx
of 2000 dynes/cm/sec. What is the damping force for = 200 − 50 (4 + x) −
g dt
the dead beat motion?
2000 dx
= 50x −
g dt
Solution Let O be the fixed end and A the free
end of the spring. It is given that a force 196000 = or (D2 + 10D + 245)x = 0 (g = 980 cm/sec2 )
980
200 gm.wt. stretches the spring by 4 cm.
The auxiliary equation (D2 + 10D + 245)x = 0
has the complex roots
o √
D = −5 ± i 220
The general solution of the problem is

√
A x(t) = (c1 cos at + c2 sin at)e−5t (a = 220)
4 cm
2000
dx/dt
B g The initial conditions are: when t = 0
x dx
T
(i) x = 5 (ii) =0
dt
p 5 cm The first condition yields c1 = 0. Differentiating
x w.r.t t we get
w
dx
= −5e−5t (c1 cos at + c2 sin at)
C dt
+ e−5t (−ac1 sin at + ac2 cos at)
Figure 2.11
The second condition yields 0 = −5c1 + ac2 ⇒
∴ 200 = 4K ⇒ K = 50 gm/cm. (K is the 25 25
c2 = =√
restoring force) a 220
The displacement after t sec is given by
Let B be the equilibrium position when a mass of
200 gm is attached to A. Then
−5t 25 √
200 = K AB ⇒ AB = 4 cm (∵ K = 50) x(t) = e 5 cos at + √ sin at (a = 220)
220
The 200 gm.wt. is pulled down to c where BC = 5
Let the damping force for the dead beat motion be
cm.
p dynes/sec, then the equation of motion is
If t sec after its release from c the weight is at p
where BP = x cm then the forces acting on it are 200 d 2 x p dx
= −50x −
(1) Its weight w = 200 gm downwards g dt 2 g dt
(2) The tension T = K AP = 50(4 + x) gm or (200D2 + pD + 49000)x = 0
upwards
2000 dx The auxiliary equation is (200D2 + pD +
(3) The damping force gm upwards 49000)x = 0
g dt
For dead beat motion, the roots of the auxiliary The particular integral is
equation must be equal.
1
√ yPI = E sin pt
∴ p = 4 × 200 × 49000 ⇒ p = 2800 5 = 6261
2 LD2 + RD + 1c

1
Example 2.76
=E sin pt
L(−p2 ) + RD + 1c
In an L–C–R circuit, the charge q on a plate of a
d 2q dq q E 1
condenser is given by L 2 + R + = E sin pt. = sin pt
dt dt c R D
The circuit is tuned to resonance so that p2 = 1/LC. E 1
If initially the current i and the charge q are zero, = cos pt ∵ +Lp2 =
R c
show that, for small values of R/L, the current in the
circuit at time t is given by (Et/2L) sin pt. Thus, the general solution of the given differential
equation is
R E
q = 1− t (c1 cos pt + c2 sin pt) − cos pt
1 2L Rp
LD2 + RD + q = E sin pt dq
c i=
dt
R
The auxiliary equation is = 1− t (−c1 p sin pt + c2 p cos pt)
2L

1 R E
LD + RD +
2
=0 − (c1 cos pt + c2 sin pt) + sin pt
c 2L Rp
+
R2
whose roots are − 2L
R
± 4L2
− 1
LC
The initial conditions are
dq
R
Since is small, we may neglect second and higher when t = 0, q = 0 and also i = =0
L dt
degree terms in RL . Thus, the roots of the equation From the first condition, we get
become
E E
0 = c1 − ⇒ c1 =
R 1 R Rp Rp
− ± i√ =− ± ip
2L LC 2L
From the second condition, we get
where p =
2 1
LC
. R R E E
The complementary function is 0 = c2 p − c1 ⇒ c2 = =
2Lp 2Lpp
R 2Lp2
yCF = (c1 cos pt + c2 sin pt) e(−R/2L)t Substituting the values of c1 and c2 , we get the
solution as
R R2
= (c1 cos pt + c2 sin pt) 1− t + 2 t 2 +· · ·
2L 4L
R E E
i = 1− t − sin pt + cos pt p
2L Rp 2Lp2
expanding e(−R/2L)t in powers of −R
2L
t, retaining R
2L
and
R E E E
neglecting higher powers we get =− − cos pt + 2
sin pt + sin pt
2L Rp 2Lp R
E R
R or i = sin pt since is small
yCF = (c1 cos pt + c2 sin pt) 1 − t 2L 2L
2L
2.3.9 Deflection of Beams If the deflection of the beam is small, the slope
Consider a uniform beam as made up of fibres run- of the elastic curve is also small so that may neglect
dy 2
ning lengthwise. We have to find its deflection under dx
in the radius of curvature formula
given loadings. 2 3/2
1 + dy
dx
R= 2 d y
dx2
A' 1
so that R may be taken as R = , for small
o p (x, y) d2y
A dx2
deflection. Under this assumption
d 2y
M =EI (2.121)
dx2
Now the shear force,
y dM d 3y
= EI 3 (2.122)
dx dx
Figure 2.12 Deflection of a beam
and the intensity of loading
In the bent form, the fibres of the lower half are
d 2M d 4y
stretched while those of the upper half are com-
2
= EI 4 (2.123)
pressed. In between these two, there is a layer of dx dx
unstrained fibres called neutral surface. Take the cen-
tral horizontal axis of the beam as the x-axis and the
N
line vertically downwards as the y-axis through the
point o such that oxy are in the vertical plane bisecting O x A
x
the cross-section of the beam, as shown in the figure. y
The fibre which was initially along the x-axis now lies R1 w
in the neutral surface, in the form of a curve called the R2
deflection curve or the elastic curve. While finding N'
y
the equation of the curve, we will be encountering
differential equations. Figure 2.13
Consider a cross-section of the beam cutting the
elastic curve at P and the neutral surface in the line We follow the following convention of signs. The
AA , called the neutral axis of this section (see Fig. sum of the moments about a section NN due to the
2.12). external forces on the left of the section is taken as
We find the bending moment about the axis AA , of positive if it is anticlockwise and negative if it is
all the forces acting on either side of the two portions clockwise (Fig. 2.13).
of the beam separated by this cross-section, given by The deflection y downwards and length x to the
Bernoulli–Euler law, as right are taken as positive. The slope dy/dx will be
positive if downwards in the direction of x-positive.
M = EI /R (2.120)
End conditions
where E is the modulus of elasticity of the beam, I is
the moment of inertia of the cross-section about AA The arbitrary constants that appear in the general
and R is the radius of curvature of the elastic curve solution of the differential Eq. (2.85) for a given prob-
at P(x, y). lem are found from the following end conditions:
O (4 ) Both the ends are freely supported or pin-

x
x jointed.
y
Example 2.77
A horizontal beam is uniformly loaded. Its one end is
y
fixed and the other end is subjected to a tensile force
2
Figure 2.14 p. The deflection of the beam is given by EI ddx2y =
py − 12 Wx2 . Given that dy
dx
= 0 at x = 0, show that the
deflection of the beam for a given x is y = pxW2 (1 −
1. At a freely supported end: There is no deflec-
Wx2
tion and no bending moment. So, we have cosh nx) + 2p
, where x2 = p/EI
d 2y Solution The equation governing the deflection

y=0 and =0 (2.124) of the beam is
dx2
d 2y Wn2 2
− n 2
y = − x
dx2 2p
O x
x Wn2 2 d
⇒ (D2 − x2 )y = − x D≡
2p dx
y A The auxiliary equation is D2 − n2 = 0 whose roots
are ±n. The general solution comprises yCF and yPI .
Figure 2.15
yCF = c1 cosh nx + c2 sinh nx
2. At a horizontal fixed end: The deflection and
the slope of the beam are both zero. So, we have and
Wn2 1
dy yPI = − x2
y=0 and =0 (2.125) 2p D − n2
2
dx −1
W D2
3. At a perfectly free end: There is no bending = 1− 2 x2
2p n
moment or shear force. So, we have
W D2 D4
= 1 + 2 + 4 + · · · x2
d 2y d 3y 2p n n
=0 and =0 (2.126)
dx2 dx4 W 2
= x2 + 2
A member of a structure or a machine when sub- 2p n
jected to end thrusts only is called a strut and a The general solution is
vertical strut is called a column. There are four pos-
sible ways of the end fixation of a strut: There are the W 2
y(x) = c1 cosh nx + c2 sinh nx + x + 2
2
following. 2p n
(1) Both the ends are fixed. This is called a built in Since y = 0 at x = 0
or encastre strut.
W 2 W
0 = c1 + 0 + 0 + 2 ⇒ c1 = − 2
(2) One end is fixed and the other end freely 2p n px
supported, hinged or pin-jointed. Also,
(3) One end is fixed and the other end free. It is dy Wx
called a cantilever. = c1 x sinh nx + c2 x cosh nx +
dx p
The second condition is and

dy R R
= 0 at x = 0 ⇒ c2 = 0 0 = 0 + a c2 − ⇒ c2 =
dx p pa
∴ The solution is ∴ The solution is

W W 2
y(x) = − 2 cosh nx + 0 + x2 + 2 Rl R R
pn 2p n y(x) = − cos ax + sin ax + (l − x)
p ap p
W W 2
or y(x) = − 2 (1 − cosh nx) + x
R 1
px 2p or y(x) = l (1 − cos ax) + sin ax − x
p a
which gives the equation of the elastic curve of the
beam under the given conditions. Example 2.79
A horizontal tie-rod of length l freely pinned at
Example 2.78 its ends carries a uniformly distributed load w per
The deflection y of a strut of length l with one end unit length and is subjected to a horizontal tension
built-in and the other end subjected to the end thrust T . Find the central deflection and show that the
2 2
p, satisfies ddx2y +a2 y = a pR (l −x). Find the deflection maximum bending moment is xw2 sech nl2 − l , where
y of the strut at a distance x from the built-in end. x2 = T /EI .
Solution The given equation is Solution At each of the two ends O and A of the

2 a2 R d tie-rod there will be a vertical reaction R = 12 wl. If
D +a y =
2
(l − x) D≡ P (x, y) is any point on the tie-rod, then wx is the
p dx
weight of the segment OP. Also, if T is the tension
The complementary function is then taking moment about p, we have
yCF = c1 cos at + c2 sin at
The particular integral is d 2y 1 x
EI 2
= Ty − wl x + wx ·
a2 R 1 dx 2 2
yPI = − (l − x) w 2
p D 2 + a2 ⇒ D2 − x 2 y = x − lx
2EI
−1
R D2
= 1+ 2 (l − x) where x2 = T /EI
p a

R D2 R
= 1 − 2 + · · · (l − x) = (l − x) l A
p a p O x
x
The general solution is y
pw
R= •
R 2 p(x, y)
y(x) = c1 cos ax + c2 sin ax + (l − x)
p y wl
R=
2
Also,
dy R Figure 2.16
= −a c1 sin ax + a c2 cos ax −
dx p
The complementary function and the particu-
dy
At the fixed end y = = 0 for x = 0 lar integral making up the general solution of the
dx equation are, respectively,
Rl Rl
∴ 0 = c1 + ⇒ c1 = − yCF = c1 cosh nx + c2 sinh nx
p p
Wn2 1 2 The maximum bending moment is

yPI = − x − lx
2T D − n
2 2 2
−1 d y
= Tymax −
Wl l
· +
Wl l
·
Wn2 D2 2 EI 2
=− 2
1 − 2
x − lx dx max 2 2 2 4
2Tn n
wl 2
W D 2
D4 = Tymax −
= 1 + 2 + 4 + · · · x2 − lx
8

2T n n W nl wl 2 wl 2
=T 2 sech −1 + −
W 2 nT 2 8T 8T
=− x2 − lx + 2
2T n w nl
= 2 sech − 1
The general solution is n 2
which had to be proved.
W 2
y(x) = c1 cosh nx + c2 sinh nx − x2 − lx +
2T n2 EXERCISE 2.14
where c1 and c2 are arbitrary constants to be deter-
1. An elastic string of natural length 2a and mod-
mined from the given conditions.
ulus λ is stretched between two points A and B
The ends are freely pinned.
distant 4a apart, on a smooth horizontal table.
∴ y = 0 at (1) x = 0 and (2) x = l A particle of mass m is attached to the middle
w of the string. Show that it can vibrate in line AB
(1) ⇒ 0 = c1 + 0 − 2 with period 2π/ω where ω2 = 2 λ/am.
Tx
w
⇒ c1 = 2 2. A spring of negligible weight which stretches 1
Tx
in. under tension of 2 lb is fixed at one end and is
(2) ⇒ 0 = c1 cosh nl + c2 sinh nl
attached to a weight of w lb at the other end. It is
W 2
− l −l + 2
2 2 found that resonance occurs when an axial peri-
2T n odic force 2 cos 2t lb acts on the weight. Show
W nl that when the free vibrations have died out, the
⇒ c2 = − 2 tanh
Tx 2 forced vibrations are given by x = ct sin 2t, and
Substituting for c1 and c2 , we get the solution as find the values of w and c.
Ans: w = 6 gm, c = 1/12
W nl
y (x) = 2 cosh nx − tanh sinh nx 3. Show that the frequency of free vibrations in a
nT 2
closed electrical circuit with
W 2 √ inductance L and
− x2 − lx + 2 capacity c in series is 30/π LC per minute.
2T n
4. An alternating e.m.f. E sin pt is applied to a
The maximum deflection occurs at the midpoint circuit at t = 0. The equation for the circuit is
where x = l/2 d 2i di 1
given as L 2 + R + = pE cos pt. Find the
W nl nl nl dt dt c
ymax (x) = 2 cosh − tanh sinh current i when (a) cR2 > 4L (b) cR2 < 44.
nT 2 2 2
2 2
Ans: (a) i = Ae−αt cosh (βt + γ)
W l l 2
− − + 2 (b) i = Ae−αt cos(βt + γ) + ER cos φ sin (pt + φ)
2T 4 2 n
where α = −R/2L, β = ±[(R/2L)2 − 1/L] and
W cosh2 nl2 − sinh2 nl2 Wl 2 Wl 2 φ = tan−1 [(1 − CLp2 )/CRp]
= 2 nl
+ −
nT cosh 2 8T n2 T 5. A light horizontal strut of length l is clamped at

W nl Wl 2 one end and carries a vertical load w at the free
= 2 sech − 1 + end. If the horizontal thrust at the free end is p,
nT 2 8T
show that the strut satisfies the differential equa- 6. A long column fixed at one end, x = 0 and

2
tion El ddx2y = δ − y p + w (l − x) where y is hinged at the other, x = l, is under the action of
the displacement of a point at a distance x from axial load p. If a force F is applied laterally at
the fixed end and δ, the deflection at the free the hinge to prevent lateral movement, show that
2 2
end. Prove that the deflection at the free end is it satisfies the equation ddx2y + n2 y = Enp (l − x)
given by np W
(tan nl − nl), where n2 EI = p. where Eln2 = p.
Rolle’s Theorem and
Mean Value Theorems 3
3.1 Introduction (2) f (x) attains every value between f (a) and f (b).
Continuous and differentiable functions have many
(3) If f (a) f (b) < 0, then there exists c ∈ (a, b) such
interesting properties, some of which are studied
that f (c) = 0 (see Fig. 3.1(b)).
through Rolle’s theorem and other mean value
theorems. Taylor’s and Maclaurin’s series, which
are generalisations of the mean value theorems are Differentiability
useful in approximating transcendental functions.
A function f : [a, b] → R is said to be differentiable
We briefly review the definitions and properties of f (x) − f (c)
continuity and derivability of functions. (derivable) at a point c ∈ (a, b) if lim
x→c x−c
exists and is finite. The limit is called the derivative
Continuity df
of f at c and is denoted by f (c) or .
dx x=c
A function f : [a, b] → R is said to be continuous at
a point c ∈ (a, b) if lim f (x) = f (c). f is derivable in (a, b) if f is derivable at each
x→c
f is continuous on the open interval (a, b) if it is point of (a, b). The function defined by these derived
continuous at every point c of (a, b). df
values is called the derivative of f on (a, b).
f is continuous on the closed interval [a, b] if dx
The derivability of f at a point c implies continuity
(1) f is continuous on the open interval (a, b), there. f is increasing or decreasing at c according as
(2) f is continuous from the right at the left f (c) ≷ 0.
end point ‘a’
i.e., f (a+) = lim f (x) = f (a) and
x→a+
3.1.1 Rolle’s Theorem
(3) f is continuous from the left at the right
end point ‘b’ Theorem 3.1 If f : [a, b] → R is
i.e., f (b−) = lim f (x) = f (b) (i) continuous in the closed interval [a, b]
x→b−
(ii) derivable in the open interval (a, b) and
Properties of continuous functions (iii) f (a) = f (b) then there exists at least
(1) If f is continuous in a closed interval [a, b], then one point c in (a, b) such that f (c) = 0.
f is bounded in that interval and further it attains
its bounds at least once in [a, b] i.e., there exist Proof: f (x) is continuous in [a, b]
c, d ∈ [a, b] such that ⇒ f (x) is bounded and attains its bounds
f (c) = Supf = M ; f (d) = lnf f = m (Fig. ⇒ There exist c, d in [a, b] such that
3.1(a)). f (c) = supf = M and f (d) = lnf f = m in [a, b].
y = f(x)
m = f(d) M = f(c) O
x
O a d c b
(a)
Figure 3.1b f is Continuous function;
Figure 3.1a Continuous function attains its
f (a)f (b) < 0. Then there exists c such that f (c) = 0
bounds: f (c) = M , f (c) = m
O
O
Figure 3.2b
Figure 3.2a
Similarly, if h > 0 such that (c+h) ∈ [a, b], we have

Case (i) M = m ⇒ f (x) is constant ⇒ f (x) = 0 for
all x. f (c + h) ≤ f (c)
In particular, f (c) = 0 (a < c < b) f (c + h) − f (c)
⇒ ≤0
Case (ii) M = m At least one of these numbers is h
different from the equal values f (a) and f (b). ⇒ Rf (c) ≤ 0 (3.4)
Suppose M = f (c), M = f (a), Since f (c) exists, we have

M = f (b) ⇒ c = a, c = b
Since f (c) = M = sup f in [a, b] f (c) = 0 by Eqs. (3.3) and (3.4)
∴ f (x) ≤ f (c) for all x ∈ [a, b] (3.1) Note

The conditions of Rolle’s theorem are sufficient but
In particular, if h > 0 such that (c−h) ∈ [a, b], we
are not necessary. That is to say, we may have the
have
conclusion of Rolle’s theorem, namely, the existence
f (c−h)−f (c) of c in [a, b] such that f (c) = 0 even if any of
f (c−h) ≤ f (c) ⇒ ≥0 (3.2)
−h the conditions are not satisfied. Illustrated below are
⇒ Lf (c) ≥ 0 (3.3) counter-examples.
Rolle’s Theorem and Mean Value Theorems 3-3
O π/ 2 3π/ 2 2π
Figure 3.3 Graph of f (x) = | sin x|
Figure 3.4 Geometrical illustration of Rolle’s

Example 3.1
theorem
1
f (x) = sin is discontinuous at ‘0’ in [−1, 1]
x
(Condition (1) fails to hold.) Solution f (x) = x2 −1 is a polynomial function
−1 1 which is continuous and derivable for all x.
But f (x) = 2 cos = 0
x x In particular, f is continuous in [−1, 1] and deriv-
2
at x = xn = ∈ (−1, 1). able in (−1, 1). Also, f (−1) = 0 = f (1)
(2n + 1)π
Thus, f satisfies all the conditions of Rolle’s the-
Example 3.2 orem and hence there exists at least one point x in
f (x) = | sin x| is not differentiable at x = π in (0, 2π) (−1, 1) such that
1
f (x) = sin x ∀ x ∈ [0, π] f (x) = 0 ⇒ 2x − 1 = 0 ⇒ x =
2
= − sin x ∀ x ∈ [π, 2π] ∴ c =
1
(Condition (2) fails to hold.) 2

Hence Rolle’s theorem is verified.
But f (x) = cos x in (0, π) and
f (x) = − cos x in (π, 2π) Example 3.5
Examine if Rolle’s theorem is applicable for
which vanish at x = π/2, in (0, π) and at x = 3π/2 in f (x) = |x| in [−1, 1].
(π, 2π), respectively (Fig. 3.3).
Solution
Example 3.3
f (x) = sin x in [0, 3π/4] x, 0 ≤ x ≤ 1
f (x) = |x| =
f is continuous in [0, 3π/4] and derivable in −x, −1 ≤ x < 0
(0, 3π/4). But√ f (0) = 0 while f (3π/4) = sin 3π/4 = ,
sin π/4 = 1/ 2 (Condition (3) is not satisfied.) f is continuous and derivable at (0,1)
But f is derivable and (1)
and (–1, 0) and continuous at x = ±1
f (x) = cos x = 0 at x = π/2 ∈ (0, 3π/4). Continuity at x = 0
f (0 +) = lim f (x) = lim x = 0

3.1.2 Geometrical Interpretation x→0+ x→0+
Rolle’s theorem states that the graph of f under the f (0 −) = lim f (x) = lim(−x) = −0
x→0− x→0
assumption has a horizontal tangent for at least one ⇒ lim f (x) = f (0) = 0
point c where c ∈ (a, b). x→0
⇒ f (x) is continuous at x = 0 (2)
Example 3.4
Verify Rolle’s theorem for f (x) = x2 −1 in [−1, 1]. and hence in [−1, 1], by Eqs. (1) and (2).
Example 3.7
Verify Rolle’s theorem for f (x) = x(x + 3)e−x/2 in
[−3, 0]. [JNTU 2001]
Solution Functions x(x + 3) and e−x/2 are con-

tinuous and derivable for all x. So, f (x) is continuous
and derivable. Also, f (−3) = f (0) = 0.
All the conditions of Rolle’s theorem are satisfied.
Figure 3.5 Modulus Function y = (x) There exists x ∈ (−3, 0) such that f (x) = 0
ex/2 2
⇒ f (x) = (x −x−6) = 0
2
Derivability at x = 0 ⇒ x = 3 or − 2
f (x) − f (0)
Rf (0) = f (0+) = lim f (3) = 0 and f (−2) = 0
x→0+ x−0
x−0 c of Rolle’s theorem is c = −2 ∈ (−3, 0)
= lim =1
x→0+ x ∵3∈ / (−3, 0).
f (x) − f (0)
Lf (0) = f (0−) = lim Example 3.8
x→0− x−0 If f (x), g(x) and h(x) are continuous in [a, b] and
−x − 0 are derivable in (a, b), then show that there exists
= lim = −1
x→0− x c ∈ (a, b) such that
Rf (0) = Lf (0) or f (0+) = f (0−)
f (a) g(a) h(a)
f (b) g(b) h(b) = 0.
⇒ f is not derivable at x = 0 and hence not
derivable at (−1, 1). Also, f (−1) = f (1) = 1. f (c) g (c) h (c)
Since the second condition of derivability is not
satisfied, Rolle’s theorem is not applicable. Solution Let φ : [a, b] → R be defined by
Example 3.6 f (x) g(x) h(x)

Verify Rolle’s theorem for f (x) = e−x sin x in [0, π]. φ(x) = f (a) g(a) h(a) (1)
[JNTU 2000S] f (b) g(b) h(b)
Solution Since sin x and e−x are continuous and ⇒ φ(x) = pf (x) + qg(x) + rh(x), where
derivable for all x, the function f (x) = e−x sin x is
continuous at [0, π] and derivable at (0, π), g(a) h(a) f (a) h(a)
p= , q=− ,
f (0) = e0 sin 0 = 0 and g(b) h(b) f (b) h(b)
f (π) = e−π sin π = e−π · 0 = 0
⇒ f (0) = f (π) = 0 f (a) g(a)
r=
f (b) g(b)
All the conditions of Rolle’s theorem are satisfied.
There exists x ∈ (0, π) such that f , g, h are continuous in [a, b] and p, q and r are real
f (x) = e−x (cos x−sin x) = 0 ⇒ cos x = sin x constants.
π ⇒ φ(x) is continuous in [a, b] and derivable in
⇒ x= .
4 (a, b). Also, φ(a) = 0 = φ(b) since two rows
π
For c = (0, π), we have f (c) = 0. become identical in the determinant in Eq. (1).
4
∴ φ satisfies all the conditions of Rolle’s theorem (e) f (x) = x2 − 5x + 6 in [2, 3] .

so that there exists c ∈ (a, b) such that Ans: c =
5
2
φ (c) = pf (c) + qg (c) + rh (c) = 0
(f) f (x) = ex sin x in [0, π].
f (a) g(a) h(a) 3π
Ans: c =
⇒ f (b) g(b) h(b) = 0. 4
f (c) g (c) h (c)
(g) f (x) = x3 (1 − x)2 in [0, 1].
Example 3.9 3
Ans: c =
a0 a1 a2 an−1 5
If + + +···+ + an = 0, prove
n+1 n n−1 2
that the equation a0 x +a1 x +· · ·+an−1 x +an = 0
n n−1
x2 + ab
has at least one root in (0, 1). (h) f (x) = log in [a, b].
√ (a + b)x
Ans: c = ab
a0 xn+1 a1 x n
Solution Let f (x) = + + ···
n+1 n 2. Examine if Rolle’s theorem is applicable for the
an−1 2
+ x + an x f is continuous on [0, 1] and function f (x) = tan x in [0, π].
2
derivable on (0, 1). Ans: Rolle’s theorem is not applicable since f is
discontinuous at x = π/2
Also, f (0) = 0 = f (1), (by hypothesis)
f satisfies all the conditions of Rolle’s theorem so 3.1.3 Lagrange’s Mean Value
that there exists c ∈ (0, 1) such that Theorem
f (x) = a0 xn + a1 xn−1 + · · · + an = 0 Theorem 3.2 If f : [a, b] → R is (i) continuous

⇒ f (c) = a0 cn + a1 cn−1 + · · · + an = 0 in the closed interval [a, b] and (ii) derivable in the
open interval (a, b) then there exists a real number c
⇒ f (x) = 0 has a root in (0, 1).
such that
EXERCISE 3.1 f (b)−f (a)
= f (c)
b−a
1. Verify Rolle’s theorem for each of the functions in
problems a–h. Proof: Let F(x) = f (x)+Ax where A is a constant
such that F(a) = F(b) ⇒ f (a)+Aa = f (b)+Ab
(a) f (x) = x2 − 2x in [0, 2]. [JNTU 1996S]
f (b)−f (a)
Ans: c = 1 ⇒ −A = (3.5)
b−a
(b) f (x) = sin x in [−π, π].
Clearly F is continuous in [a, b] and derivable
Ans: c = ±π/2
in (a, b) since f (x) and Ax satisfy these conditions.
Also, F(a) = F(b)
(c) f (x) = 4 − x2 in [−2, 2]. Since F satisfies all the conditions of Rolle’s
Ans: c = 0 theorem there exists c in (a, b) such that F (c) = 0.
(d) f (x) = (x − a)m (x − b)n in [a, b], ⇒ F (c) = f (c) + A = 0 (3.6)

m > 0, n > 0. f (b) − f (a)
mb + na ⇒ − A = f (c) = (3.7)
Ans: c = b−a
m+n
But f (x) = 0 for all x ∈ (a, b) ⇒ f (c) = 0

⇒ f (p) = f (a).
Since p is any point in (a, b), f is constant in (a, b).
Corollary 2 A function f is said to be monoton-

ically increasing in an interval if
x < y ⇒ f (x) < f (y) for all x, y ∈ I
and monotonically decreasing in I if
Figure 3.6 The Mean Value Theorem of x < y ⇒ f (x) > f (y) for all x, y ∈ I
Differential Calculus (Lagrange’s Mean Value If f is such that f (x) > 0 for all x ∈ (a, b), then
Theorem) f (x) is monotonically increasing in [a, b].
Proof: Let p be any point in (a, b).

By mean value theorem,
3.1.4 Geometrical Interpretation of
Lagrange’s Mean Value f (p)−f (a)
= f (c) > 0 (a < c < p)
Theorem p−a
Let y = f (x) be a curve whose graph is continuous p > a ⇒ f (p) > f (a)
in [a, b] and at each point in (a, b) a tangent can be for all p ∈ (a, b)
f (b)−f (a)
drawn. The difference quotient gives the ∴ f is monotonically increasing in (a, b).
b−a If f (x) < 0, then
slope of the secant (chord) to the curve through the
points (a, f (a)) and (b, f (b)). Then there must be at f (p)−f (a)
= f (c) < 0
least one point c in the interval (a, b) for which the p−a
tangent to the curve is parallel to the secant. Both
Therefore, we have
then have the same slope
p > a ⇒ f (p) < f (a) for all p ∈ (a, b)
f (b) − f (a)
f (c) = . ∴ f is monotonically decreasing in (a, b).
b−a
Another form Example 3.10

Verify Lagrange’s mean value theorem for
If the values of a and b are denoted by a = x and f (x) = (x − 1)(x − 2) in [1,3].
b = x+h, then c can be expressed in the form c = x+θh
(0 < θ < 1). Solution f (x) = (x − 1)(x − 2) = x2 − 3x + 2
The mean value theorem has the form f is a polynomial function. So, it is continuous and
derivable for all x.
f (x+h)−f (x)
= f (x+θh) (0 < θ < 1). a = 1, b = 3;
h f (1) = 0 and f (3) = 2; f (x) = 2x − 3
Corollary 1 If f is continuous in [a, b] and deriv- The conditions of Lagrange’s mean value theorem
able in (a, b), and f (x) = 0 for all x∈ (a, b) then f are satisfied.
is constant. ∴ There exists c in (1, 3) such that
f (b) − f (a)
Proof: Let p be any point in (a, b). f (c) =
b−a
f is continuous in [a, p] and derivable in (a, p). 2−0
⇒ f (c) = 2c − 3 = =1
By Lagrange mean value theorem, there exists a 3−1
point c in (a, p) such that f (p)−f (a) = (p−a)f (c). ⇒ c = 2 ∈ (1, 3)
Hence Lagrange’s mean value theorem is verified. Clearly f is continuous in [0, 1] and derivable in
(0, 1). So, all the conditions of Lagrange’s mean value
Example 3.11 theorem are satisfied.
Find c of Lagrange’s mean value theorem for
e−1
f (x) = x(x − 1)(x − 2) in [0, 1/2] . ∴ f (c) = ec = ⇒ c = log(e − 1) ∈ (0, 1).
1−0
Solution f (x) = x(x −1)(x −2) = x3 −3x2 +2x. Example 3.14
f is a polynomial function. So, it is continuous and Examine if Lagrange’s mean value theorem can be
derivable for all x. applied for
a = 0, b = 12 ;
x sin 1
x = 0
f (a) = f (0) = 0, f (b) = f 12 = 38 f (x) = x
in [−1, 1].
The conditions of Lagrange’s mean value theorem 0 x=0
are satisfied.
Solution For x = 0, f is clearly continuous and
Also, f (x) = 3x2 − 6x + 2 derivable
f (b) − f (a)
f (c) = Continuity at x = 0
b−a 1
3
− 0 3 lim f (x) = lim x sin = 0 = f (0)
⇒ 3c2 − 6c + 2 = 8 =
x→0 x→0 x
1/2 4 1
∵ sin ≤1
12c2 − 24c + 5 = 0 x
√
6 ± 21 So, f is continuous at x = 0 also.
⇒c=
6
√ f (0 + h)−f (0)
6 + 21 f (0) = lim
Since
∈ (0, 1/2)
h→0 h
6 √ = lim
h sin 1/h
= lim sin
1
6 − 21 h h
the required c = . h→0 h→0
6
Example 3.12 does not exist.
Verify Lagrange’s mean value theorem for So f is not derivable at x = 0.
f (x) = log x in [1, e]. Hence Lagrange’s mean value theorem cannot be
applied.
Solution f (x) = log x, a = 1, b = e;
f (a) = f (1) = 0, f (b) = f (e) = 1 Example 3.15
f is continuous and derivable for all x > 0. If f (x) = x2 find θ ∈ (0, 1) such that
1 f (x + h) = f (x) + hf (x + θh).
Also, f (x) = ; all the conditions of Lagrange’s
x
mean value theorem are satisfied. Solution f (x) = x2 is continuous and derivable

1 1−0 for all x. f (x) = 2x
∴ = ⇒ c = (e − 1) ∈ (1, e) By Lagrange’s mean value theorem,
c e−1 f (x + h)−f (x)
f (x + θh) =
Hence Lagrange’s mean value theorem is verified. h
1
Example 3.13 ⇒ 2(x + θh) = 2x + h ⇒ θ =
2
Find c of Lagrange’s mean value theorem if
Example 3.16
f (x) = ex in [0, 1].
Prove that
Solution f (x) = ex a = 0, b = 1; x2 x2
f (a) = f (0) = 1, f (b) = f (1) = e1 f (x) = ex x− < log(1+x) < x− for x > 0. (1)
2 2(1 + x)
x2
Solution Let f (x) = log(1 + x) − x − 2
. By Lagrange’s mean value theorem,
f is continuous and derivable for all x > 0.
f (a + h) f (a) + hf (a + θh)
1 x2 f (a)
f (x) = − (1 − x) = > 0 for x > 0 ⇒ h=−
1+x 1+x f (a)
−1 1
⇒ f is monotonically increasing for x > 0. ⇒ h=− = = 0.05
20 20
x2 An approximate root is a + h = 2 + 0.05 = 2.05.
⇒ x− < log(1 + x) (2)
2 Applying Lagrange’s mean value theorem again and
x2 taking a = 2.05
Let g(x) = x − − log(1 + x)
2(1 + x)
f (a) f (2.05) 0.061
h=− =− =− = −0.0027
g is continuous and derivable for all x > 0. f (a) f (2.05) 22.46
x x2 1 A second approximation to the root is
g (x) = 1 − + −
1 + x 2(1 + x)2 1+x
2.05 − 0.0027 = 2.0473.
x2
= >0
2(1 + x)2 Example 3.18 √
5
Calculate approximately 245 using Lagrange’s
⇒ g(x) is monotonically increasing for x > 0.
mean value theorem.
x2
⇒ log(1 + x) < x − (3) Solution Let f (x) = x1/5 , a = 243 = 35 ,
2(1 + x)
b = 245
From Eqs. (2) and (3), we obtain Eq. (1).
1 −4/5
f (x) = x
Note 5
Using Lagrange’s mean value, the approximate solu- f (a + h) = f (a) + hf (c)
tions of equations of the form f (x) = 0 can be
obtained by Newton’s method. c is a point in (a, b) = (243, 245).
Since we are interested in an approximate value of
Suppose (a + h) is an exact root of f (x) = 0, f (a + h), we may take c as 243 so that
0 = f (a + h) 1 5 −4/5
we have f (c) = f (243) = f (35 ) = (3 )
5
= f (a) + hf (a + θh) (0 < θ < 1)
1 −4 1 1
f (a) = ·3 = ·
⇒h− (Newton’s formula). 5 5 81
f (a) f (245) = f (243) + (245 − 243)f (243)
Example 3.17 √
5 1 1
245 = (243)1/5 + 2 · ·
Calculate the approximate root near 2 for the equa- 5 81
tion f (x) = x4 − 12x + 7 = 0. 1 1
= 3+2· ·
5 81
Solution = 3 + 0.0049 = 3.0049.
f (x) = x4 − 12x + 7 = 0 EXERCISE 3.2
f (x) = 4x3 − 12
Take a = 2 f (2) = 16 − 24 + 7 = −1 1. Verify Lagrange’s mean value theorem for each
f (2) = 32 − 12 = 20 of the functions in problems (a) to (e):
(a) f (x) = x − x3 in [−2, 1] [JNTU 2002] 6. Test if Lagrange’s mean value theorem holds for
Ans: c = −1 f (x) = x−x3 in [−2, 1]. If so, find an approximate
value of c. [JNTU 2002]
(b) f (x) = 5x2 + 7x + 6 in [3, 4]
7 Ans: c = 1 or −1
Ans: c =
2
1 3.1.5 Cauchy’s Mean Value
(c) f (x) = , a = 1, b = 4 Theorem
x
Ans: c = 2
Theorem 3.3 Let f : [a, b] → R, g : [a, b] → R
(d) f (x) = ex in [0, 1]
be
Ans: c = e − 1
(i) continuous in the closed interval [a, b]
(e) f (x) = log x in [e2 , e3 ] (ii) derivable in the open interval (a, b)
Ans: c = (e − 1)e2 and (iii) g (x) = 0 for all x in (a, b); then there
exists a point c in (a, b) such that
2. Show that
x2 f (c) f (b) − f (a)
(a) cos x > 1 − = .
2 g (c) g(b) − g(a)
π 1 3 x 1 Proof: Let
(b) − √ > cos−1 > −
3 5 3 5 3 8
[JNTU 1998S] φ(x) = f (x) + Ag(x) (3.7)
π 3 4 π 1 where A is a constant such that
(c) + < tan−1 < +
4 25 3 4 6
[JNTU 1996] φ(a) = φ(b) ⇒ f (a) + Ag(a) = f (b) + Ag(b)
√ f (b)−f (a)
3. Calculate an approximate value of 6 65 using ⇒ −A = (3.8)
Lagrange’s mean value theorem. g(b)−g(a)
[Hint: Take f (x) = x1/6 in (64, 65) If g(b) = g(a) then by Rolle’s theorem there exists
f (a + h) = f (a) + hf (a)] an x in (a, b) such that g (x) = 0, which contradicts
Ans: : 2.0052
condition (3) of the hypothesis.
4. Find an approximate value for the root of Now φ(x) satisfies all the conditions of Rolle’s the-
x3 − 2x − 5 in (2, 3). orem and hence there exists a point c in (a, b) such
that φ (c) = 0
[Hint: f (x) = x3 − 2x − 5
f (x) = 3x2 − 2 Take a = 2, at h = 2.1 f (c) f (b) − f (a)
⇒ −A = = .
f (a) f (2) g (c)
g(b) − g(a)
=− =− = 0.1
f (a) f (2)
Another form
f (2.1)
Take a = 2.1 h=− = −0.0053
f (2.1) Put b = a + h, c = a + θh where (0 < θ < 1),
root = 2.1 − 00543 = 2.0946] we have
f (a + θh) f (a + h) − f (a)
Ans: root=2.0946 = (0 < θ < 1).
g (a + θh)
g(a + h) − g(a)
5. Using Lagrange’s theorem show that
Note
x Cauchy’s mean value theorem is an extension (gen-
x > log(1 + x) > .
1+x eralisation) of Lagrange’s mean value theorem and
if we take g(x) = x, Cauchy’s mean value theorem f (b) = sin b, g(b) = cos b;
reduces to Lagrange’s mean value theorem. f (b) − f (a) sin b − sin a
= = − cot c
g(b) − g(a) cos b − cos a
Example 3.19
Verify Cauchy’s mean value theorem for the follow- 2 cos b+a sin b−a
⇒ 2 2
= − cot c
ing pairs of functions: −2 sin b+a
2
sin b−a
2
a+b a+b
(i) f (x) = x3 , g(x) = x2 in [1, 2] ⇒ cot = cot c ⇒ c = ·
2 2
1
(ii) f (x) = log x, g(x) = in [1, e] Example 3.20
x Let f : [a, b] → R be continuous in [a, b] and deriv-
(iii) f (x) = sin x, g(x) = cos x in [a, b] able in (a, b). Then show that there exists a number
c in (a, b) such that 2c[f (a)−f (b)] = f (c)(a2 − b2 ).
Solution
Solution By taking g(x) = x2 in Cauchy’s mean
(i) f (x) = x , g(x) = x in [1, 2]
3 2 value theorem, we have
f (x) = 3x2 , g (x) = 2x f (b)−f (a) f (c)
=
a = 1, b = 2 b −a
2 2 2c
f (a) = f (1) = 1, g(a) = g(1) = 1 ⇒ 2c(f (b)−f (a)) = f (c)(b2 − a2 ).
f (b) = f (2) = 8, g(b) = g(2) = 4 Example 3.21
f (c) 3c2 3 If f and g are differentiable on [0, 1] such that
= = c
g (c) 2c 2 f (0) = 2 and g(0) = 0; f (1) = 6 and g(1) = 2
f (b) − f (a) 8−1 7 3c2 3 then show that there exists c ∈ (0, 1) such that
= = = = c f (c) = 2g (c).
g(b) − g(a) 4−1 3 2c 2
14 f and g are derivable in [0, 1].
⇒c= . Solution
9 ⇒ f and g are continuous in [0, 1].
1 By Cauchy’s mean value theorem, there exists c
(ii) f (x) = log x, g(x) = in [1, e]
x such that
1 1
f (x) = , g (x) = − 2 [f (b) − f (a)] g (c) = [g(b) − g(a)] f (c)
x x
f (c) 1/c a = 0, b = 1 f (0) = 2, f (1) = 6
a = 1, b = e = = −c g(0) = 0, g(1) = 2
g (c)
−1/c2
f (a) = f (1) = 0, g(a) = g(1) = 1
∴ 4g (c) = 2f (c) ⇒ f (c) = 2g (c).
1
f (b) = f (e) = 1, g(b) = g(e) = Example 3.22
e
f (b) − f (a) 1−0 Find c of Cauchy’s mean value theorem for
= 1 = −c √
g(b) − g(a) e
−1 f (x) = x, g(x) = √
1
in [a, b].
e x
⇒ c= ∈ (1, e).
e−1
Solution xn is continuous for all x > 0 and n ∈ Q.
(iii) f (x) = sin x, g(x) = cos x
f and g are continuous on [a, b] ⊆ R.
f (x) = cos x, g (x) = − sin x Also, f (x) = 2√1 x , g (x) = − 2x1√x
f (a) = sin a, g(a) = cos a; exist for all x > 0.
f (c) ⇒ f and g are derivable on (a, b) ⊆ R+ and
= cot c
g (x) = 0.
g (c)
By Cauchy’s mean value theorem, there exists Differentiating (3.11) with respect to x,
c ∈ (a, b) such that
%
n−1
(a + h − x)r
f (b) − f (a) f (c) φ (x) = f (r+1) (x)
= r!
g(b) − g(a) g (c) r=0
√ √
b− a √ √1
%
n−1
(a + h − x)r−1
f (r) (x)
2 c
⇒ 1 = − ab = −
√ − √1
b a
− 2c1√c r=1
(r − 1)!
√
⇒ c = ab ∈ (a, b). −Ap(a + h − x)p−1
(Except the nth term all the terms of the first
3.1.6 Generalised Mean Value series cancel out with the corresponding terms of
Theorem—Taylor’s Theorem the second series.)
(a + h − x)n−1 (n)
Theorem 3.4 If f : [a, a + h] → R is a function ⇒ φ (x) = f (x)
(n − 1)!
whose (n − 1)th derivative f (n−1) is
(i) continuous in the closed interval [a, a+h] − Ap(a + h − x)p−1
(ii) derivable in the open interval (a, a + h) ⇒ 0 = φ (a + θh)
and (iii) p is a given positive integer, then there hn−1 (1 − θ)n−1 (n)
exists at least one number θ in (0,1) such that = f (a + θh)
(n − 1)!
%
n−1 r − Ap(1 − θ)p−1 hp−1
h
f (a + h) = f (r) (a) + Rn (3.9) hn−1 (1−θ)n−1 (n)
f (a + θh)
r! (n−1)!
r=0 ⇒A=
p(1 − θ)p−1 hp−1
(f (0) ≡ f )
hn−p (1 − θ)n−p (n)
hn (1 − θ)n−p (n) = f (a + θh); (3.13)
where Rn = f (a + θh) (3.10) (n − 1)!p
(n − 1)!p
h = 0, θ = 1
(0 < θ < 1)
Substituting the value of A from Eq. (3.13) into
Proof: f (r) (r = 0, 1, 2, · · · (n − 1)) are continuous Eq. (3.12)
in [a, a + h] and derivable in (a, a + h).
%
n−1 r
h
%
n−1
(a + h − x)r f (a + h) = f (r) (a)
Let φ(x) = f (r) (x) r=0
r!
r!
r=0
hn (1 − θ)n−p (n)
+ A(a + h − x)p (3.11) + f (a + θh) (3.14)
(n − 1)!p
where the constant A is to be determined such that (0 < θ < 1).
φ(a + h) = φ(a) Note

Taylor’s theorem is a generalisation of Cauchy’s
%
n−1 r
h (r)
⇒ f (a + h) = f (a) + hp A, (3.12) mean value theorem. Cauchy’s mean value theorem
r=0
r! is obtained from Taylor’s theorem for n = 1. Here, Rn
[using Eq. (3.11)] denotes the remainder after n terms of Taylor’s series
expansion. The various forms of remainders are
φ satisfies all the conditions of Rolle’s theorem so
that there exists c ∈ (a, a + h) such that (a) Schlomilch and Roche form of Rn :
hn (1 − θ)n−p (n)
φ (c) = φ (a + θh) = 0. f (a + θh)
(n − 1)!p
(b) Cauchy form of Rn : Maclaurin’s series

hn (1 − θ)n−1 (n)
f (a + θh) For a = 0, we get Maclaurin’s series
(n − 1)!
(c) Lagrange’s form of Rn : %
∞
xr
hn (n) f (x) = f (r) (0)
f (a + θh) r=0
r!
n!
Taking a + h = x ⇒ h = x − a in (3.14), we get This formula enables us to derive power series
Taylor’s series in the form expansions for most of the functions which are
differentiable.
%
n−1
(x − a)r (x − a)n (1 − θ)n−p Example 3.23
f (x) = f (r) (a) + Expand sin x in powers of x.
r=0
r! (n − 1)!p
(n)
f [(a + θ(x − a)] (0 < θ < 1) (3.15) Solution Here
Putting a = 0, we get Maclaurin’s theorem f (x) = sin x f (0) = 0
%
n−1 r
x π
f (x) = f (r) (0) + Rn (3.16) f (x) = cos x = sin x + f (0) = 1
r! 2
r=0 π
f (x) = − sin x = sin x + 2 · f (0) = 0
x (1 − θ)n−p (n)
n
2
where Rn = f (θx) (3.17) π
(n − 1)!p
f (x) = − cos x = sin x + 3 · f (0) = −1
2
is the Schlomilch remainder.
and generally,
Lagrange’s remainder put p = n in Eq. (3.17) and
π
we have f (n) (x) = sin x + n
2
xn (n) 0 (n even)
f (θx). (3.18) nπ
n! f (n) (0) = sin = n−1
2 (−1) 2 (n odd)
Cauchy remainder put p = 1 in Eq. (3.17) and we
have
x3 x5 (−1)n−1 x2n−1
xn (1 − θ)n−1 (n) sin x = x − + − ··· + + ··· .
f (θx). (3.19) 3! 5! (2n − 1)!
(n − 1)!
Example 3.24
Taylor’s series Expand ex about x = 1.
Since Rn → 0, as n → ∞ we obtain Taylor’s series Solution Taylor’s series about x = a is

in the form (x − a) (x − a)2
f (x) = f (a) + f (a) + f "(a) + · · ·
%
∞ 1! 2!
hr
f (a + h) = f (r) (a)
r! Here f (x) = ex f (1) = e
r=0
f (n) (x) = ex f (n) (1) = e
for all n
or, in powers of (x−a), we have (x − 1)2
ex = e + (x − 1)e + e + ···
%
∞
(x − a)r
2!

f (x) = f (r) (a). (x − 1)2
r=0
r! = e 1 + (x − 1) + + ··· .
2!
5 3 1
Note f (x) =
1
· · (1 − x)− 2 (−1)3
(x − 1)2 2 2 2
e = e · e = e 1 + (x − 1) +
x x−1
2! 15 1
f (θx)|x=1 = − (1 − θ)− 2

(x − 1)3 8
+ + ··· .
3! x = 1 ⇒ f (x) = f (1) = 0 [by Eq. (2)]
Example 3.25
x2 x3 Substituting x = 1 and the above values in Eq. (1)
Prove that log(1 + x) = x − + − ···
2 3
x n−1
xn 1 5 15 1 1 15 1
+ (−1)n−2 + (−1)n−1 . 0 = 1+ − + · − · (1 − θ)− 2
n−1 n(1 + θx)n 1! 2 4 2 6 8
5 15 5 1
Solution Here ⇒1− + − (1 − θ)− 2 = 0
2 8 16
5 1 8 − 20 + 15
f (x) = log(1 + x) f (0) = 0 ⇒ · =
16 (1 − θ) 12 8
(−1)n−1 (n − 1)!
f (n) (x) = 5
(1 + x)n ⇒ 3 (1 − θ) 2 =
1
f (n) (0) = (−1)n−1 (n − 1)! 2

5 1

1 2
(−1)n−1 (n − 1)! ⇒ (1 − θ) 2 =⇒ (1 − θ) 2
f (n) (θx) = 6
(1 + θx)n 2
5 25
= =
Maclaurin’s theorem is 6 36
% n−1 xr 25 11
xn ⇒θ =1− = ∈ (0, 1).
f (x) = lim f (r) (0) + f n (θx) 36 36
r=0 r! n!
% n−1 xr Hence Maclaurin’s theorem is verified.
log(1 + x) = lim (−1) (r − 1)!
r−1
r=1 r! Example 3.27
(−1)n−1 (n − 1)!xn Using Maclaurin’s series expand tan x up to the fifth
+ power of x and hence find series for log sec x.
n!(1 + θx)n
% n−1 xr (−1)n−1 xn
= lim(−1)r−1 + . Solution Here
r=1 r n(1 + θx)n
f (x) = tan x = t (say) f (0) = 0
Example 3.26
5 f (x) = sec x = 1 + t ,
2 2
Verify Maclaurin’s theorem for f (x) = (1 − x) 2 with dt
Lagrange’s form of remainder up to 3 terms with = sec2 x = 1 + t 2
dx
x = 1.
f (0) =1
Solution Maclaurin’s theorem for f (x) is f "(x) = 2t(1 + t 2 ) = 2t + 2t 3
x x2 x3 f (0) = 0
f (x) = f (0) + f (0)+ f (0)+ f (θx) · · · (1)
1! 2! 3! f (x) = (2 + 6t 2 )(1 + t 2 )
5
Here f (x) = (1 − x) 2 f (0) = 1 (2) = 2 + 8t 2 + 6t 4
5 3 5 f (0) = 2
f (x) = (1 − x) 2 (−1) f (0) = −
2 2 f (4) (x) = (16t + 24t 3 )(1 + t 2 )
5 3 1
= 16t + 40t 3 + 24t 5
f "(x) = · · (1 − x) 2 (−1)2
2 2 f (4) (0) = 0
15
f "(0) = f (5) (x) = (16 + 120t 2 + 120t 4 )(1 + t 2 )
4
= 16 + 36t 2 + 240t 4 + 120t 6 % ∞ (−1)r−1

loge x = lim (x − 1)r
f (5) (0) = 16 r=1 r
(x − 1)2 (x − 1)3
Maclaurin’s series for f (x) is = (x − 1) − +
2 3
x x2 x3 (x − 1)4
f (x) = f (0) + f (0) + f "(0) + f (0) − + ···
1! 2! 3! 4
x4 (4) x5 (5) x = 1.1 ⇒ log 1.1
+ f (0) + f (0) + · · ·
4! 5! 1 (0.1)2 1 1
1 3 2 5 = 0.1 − + (0.1)3 − (0.1)4 + · · ·
tan x = x + x + x + · · · 2 2 3 4
3 15 = 0.1 − 0.005 + 0.0003 − 0.00002
Integrating we get = 0.0953.
x2 1 x4 2 x6
log sec x = c + + · + · + ···
2 3 4 15 6 EXERCISE 3.3
When x = 0, log sec 0 = log 1 = 0 ⇒ c = 0 1. Using Maclaurin’s series show that
*
x2 x4 x6 1+x x3 x5
∴ log sec x = + + + ··· . (a) log =x+ + + ···
2 12 45 1−x 3 5
Example 3.28 x4 2x6
Expand loge x in powers of (x−1) and hence evaluate (b) cos2 x = 1 − x2 + − + ···
3 45
log 1.1 correct to 4 decimal places.
x2 x4
Solution (c) esin x = 1 + x + − + ···
2 8
f (x) = loge x f (1) = 0
1 sin−1 x 2 8
f (x) = f (1) = 1 (d) √ = x + x3 + x5 + · · ·
x 1−x 2 3 15
1 √
f "(x) = − 2 f "(1) = −1 and generally 2. Calculate the approximate value of 10 correct to
x 4 decimal places using Taylor’s theorem.
(n) (−1)n−1 (n − 1)! (n) Ans: 3.1623
f (x) = f (1)
xn
x2 x2 x3
= (−1)n−1 (n − 1)! 3. Prove that 1 − ≤ cos x ≤ 1− + (x > 0).
2 2 24
Taylor’s series is
1 x3 1 · 3 x5
4. Show that sin−1 x = x + · + + ··· .
%
∞
2 3 2·4 5
f (x) = (x − a)r−1 f (r) (a)
r=0 x x2 x4
%
∞ 5. Show that log(1+ex ) = log 2+ + − +· · · .
(−1)r−1 (x − 1)r−1 2 8 192
= (r − 1)! e x
1 x x 3
r! Hence deduce that x = + − + ··· .
r=1
e +1 2 4 48
Functions of Several Variables 4
4.1 Introduction Example 4.2
We have so far considered continuity and derivability The equation x2 +y2 = a2 represents the right circular
of real-valued functions of a real variable. Now cylinder of radius a with its axis oriented along the
we extend these concepts to functions of two or z-direction.
more variables. In this chapter, we study partial
differentiation, Euler’s theorem on homogeneous
4.3.1 Limit of a Function of Two
functions, Jacobians and Taylor’s theorem for a
Variables
function of two variables.
Let u = f (x, y) be a function defined for all (x, y) ∈ D
except possibly at (a, b) and suppose that l is a real
4.2 Functions of Several
number. Then u is said to tend to the limit l as (x, y)
Variables
tends to (a, b) if for every ∈> 0 there corresponds a
We are familiar with functions of a single variable. δ > 0 such that
But in nature there arise situations where we have to
deal with functions of more than one variable as, for |f (x, y)−l| < ∈ whenever max {|x − a|, |y − b|} < δ
instance, the area A of a rectangle A = xy which is
a function of the variable x (length) and the variable and we write
y (breadth). Also, the velocity v of a fluid particle
moving in 3-dimensional space is a function of three lim f x, y = l or lim f x, y = l.
(x,y)→(a,b) x→a
y→b
space variables x, y and z and the time variable t.
If u = f (x, y, z, t) then x, y, z and t are known as
independent variables and u is called the dependent
4.3.2 Continuity of a Function of
variable.
Two Variables
Similarly we can consider u = f (x1 , x2 , · · · xn ), a
function of n variables xi (i = 1, 2, · · · n). Let u = f (x, y) be a function defined for all (x, y) ∈ D
and (a, b) ∈ D. Then u = f (x, y) is said to be contin-
uous at (a, b) if for every ∈ > 0 there corresponds a
4.3 Function of Two Variables
δ > 0 such that
If for every x and y varying in a domain D there is a

unique value u = f (x, y), then u is called a function f x, y − f (a, b) < ∈
of the variables x and y. Geometrically, this equation
represents a surface. whenever m {|x − a| , |y − b|} <δ
This implies that lim f x, y = f (a, b). It
Example 4.1 (x,y)→(a,b)
The equation z = r 2 − x2 − y2 x2 + y2 ≤ r 2 is easy to extend these definitions to the case of a
represents the upper hemisphere. function of more than two variables.
Note Solution
The limit must exist and be equal in whatever manner xy xy
we approach the point (a, b). lim = lim lim
x→0
y→0 x2 + y2 x→0 y→0 x2 + y2
- / ,
. = lim 0 = 0
Also, lim lim f x, y = lim lim f x, y . x→0
y→b x→a x→a y→b xy xy
lim = lim lim
y→0
x→0
x2 + y2 y→0 x→0 x2 + y2
Example 4.3 = lim 0 = 0
2xy y→0
Evaluate lim 2 + y2
. xy mx
( )
x,y →(1,2) x lim = lim √ =0
y→mx
x→0 x2 + y2 x→0 1 + m2
Solution (independently of m)
/ ,
2xy 2xy The limit exists and is equal to zero.
lim = lim lim
x→1
y→2
x2 + y2 x→1 y→2 x 2 + y 2 Example 4.6
Examine for continuity at the origin of the function
4x 4
= lim =
x→1 x 2 + 4 5 x2
2xy f x, y = x, y = (0, 0)
Also, = lim 2 x2 + y2
y→2 x + y 2
x→1 =0 x, y = (0, 0) .
/ ,
2xy
= lim lim 2
y→2 x→1 x + y 2 Solution
2y 4 x2 x2
= lim = . lim = lim lim
y→2 1 + y 2 5 x→0
y→0 x2 + y2 x→0 y→0 x2 + y2
x2
2xy 4 = lim x = 0
= lim
x→0 x
∴ lim exists and is equal to . x→0
(x,y)→(1,2) x2+y 2 5 x2 x2
lim = lim lim
Example 4.4
y→0
x→0
x2 + y2 y→0 x→0 x2 + y2
x −y
2 2
= lim 0 = 0
Evaluate lim . y→0
(x,y)→(0,0) x2 + y2 2
x x
lim = lim √ = 0.
Solution y→mx
x→0 x2 + y2 x→0 1 + m2
x2 − y2 x2
lim lim = lim =1 Thus, the limit of the function exists along any
x→0 y→0 x 2 + y 2 x→0 x 2
path and is equal to zero, which is the value of the
x2 − y2 −y2 function.
Also, lim lim 2 = lim = −1.
y→0 x→0 x + y 2 y→0 y 2
Hence the function is continuous at the origin.
Here the limit does not exist.

4.4 Partial Differential
Coefficient
Example 4.5
xy Let u = f (x, y) be a function of two variables x and
Evaluate lim .
(x,y)→(0,0) x2 + y2 y. Then the partial differential coefficient of u with
Functions of Several Variables 4-3
∂u Example 4.8
respect to x, denoted by or ux is defined by
∂x Find the first order partial derivatives of
y
∂u f x + h, y − f x, y (i) u = ex sin y (ii) u = tan−1
= lim , x
∂x h→0 h (iii) u = log x2 + y2 .
if the limit exists. Solution
∂u ∂u
(i) = ex sin y; = ex cos y
Example 4.7 ∂x ∂y
∂u ∂u 1 y y
Find if u x, y = exy . (ii) =
∂x y2 · − 2 = − 2
∂x 1+ x x x + y2
Solution ∂u 1 1 x
= y2 · = 2
∂u e(x+h)y − exy ehy − 1 ∂y 1+ x x + y2
= lim = exy lim x
∂x h→0 h h→0 h ∂u 1 1
hy
− (iii) = · · 2x
= yexy lim
e 1
= yexy · 1 ∂x x2 + y2 2 x2 + y2
hy→0 hy x
ex − 1 =
∵ lim =1 x2 + y2
x→0 x ∂u 1 1
= · · 2y
∂u ∂y x2 + y2 2 x2 + y2
Thus, is obtained by differentiating u with y
∂x = .
respect to x, treating all but the variable x as constants. x2 + y2
∂u
Similarly can also be defined. Example 4.9
∂y
∂u ∂u If u = log(x3 + y3 + z 3 − 3xyz) then prove that
, are called the first order partial differential
∂x ∂y ∂u ∂u ∂u 3
coefficients of u with respect to x and y, respectively. (i) + + =
∂x ∂y ∂z x+y+z

4.4.1 Partial Differential ∂ ∂ ∂ 2 −9
(ii) + + = 2 .
Coef f icients of Higher Order ∂x ∂y ∂z x+y+z
∂u ∂u Solution
, are functions of x and y. Partial differential
∂x ∂y
coefficients of these functions may be defined as (i) u = log(x3 + y3 + z 3 − 3xyz)

∂ ∂u ∂2 u ∂ ∂u ∂2 u ∂u 1 2
= 2, = = 3 · 3x − 3yz
∂x ∂x ∂x ∂x ∂y ∂x∂y ∂x x + y3 + z 3 − 3xyz
∂u 1
∂ ∂u ∂u
2
∂ ∂u ∂2 u = 3 · 3y2 − 3xz
= , = 2 ∂y x + y + z − 3xyz
3 3
∂y ∂x ∂y∂x ∂y ∂y ∂y
∂u 1
These are called the second order partial differen- = 3 · 3z 2 − 3xy
∂z x + y + z − 3xyz
3 3
tial coefficients of u(x, y). If the partial derivatives
∂2 u ∂2 u ∂2 u ∂2 u
and are continuous, then = , Adding
∂x∂y ∂y∂x ∂x∂y ∂y∂x ∂u ∂u ∂u
that is, the order of differentiation is immaterial. + +
∂x ∂y ∂z
Partial differential coefficients of first order 2
and higher order can be defined in respect of 3 x + y2 + z 2 − yz − zx − xy
=
functions involving more than two variables. x + y + z x2 + y2 + z 2 − yz − zx − xy
∂u ∂u ∂u 3 Example 4.11
∴ + + = (1)
∂x ∂y ∂z x+y+z If u = eaθ cos(a log r), then show that
Hence proved. ∂2 u 1 ∂u 1 ∂2 u
(ii) From Eq. (1) we may write + + = 0.
∂r 2 r ∂r r 2 ∂θ 2

∂ ∂ ∂ 3 Solution
+ + u=
∂x ∂y ∂z x+y+z
∂u a
= eaθ − sin (a log r) . (1)
∂ ∂ ∂ 2 ∂r r
+ + u ∂2 u a2
∂x ∂y ∂z = e aθ
− cos (a log r)
∂r 2 r2
∂ ∂ ∂ ∂ ∂ ∂
= + + + + u a
∂x ∂y ∂z ∂x ∂y ∂z +eaθ − sin a log r − 2 (2)
r
∂ ∂ ∂ 3 1 ∂u a
= + + = eaθ − sin a log r · 2 (3)
∂x ∂y ∂z x + y + z r ∂r r
% ∂ 1 1 ∂2 u a2 aθ
=3 = 2 e cos a log r (4)
∂x x + y + z r 2 ∂θ 2 r
% 1 −9
= −3 2 = 2 . Adding Eqs. (2), (3) and (4)
x+y+z x+y+z
∂2 u 1 ∂u 1 ∂2 u
∴ + + 2 2 = 0.
Hence proved. ∂r 2 r ∂r r ∂θ
Example 4.10 EXERCISE 4.1

Show that V x, y, z = e5z sin 3x cos 4y satisfies the ∂u ∂u
Laplace’s equation 1. Find and if
∂x ∂y
∂2 V ∂2 V ∂2 V
+ + = 0. x−y
∂x2 ∂y2 ∂z 2 (a) u =
x+y
Solution 2y −2x
Ans: ux = 2 , uy = 2
∂V x+y x+y
= 3e cos 3x cos 4y,
5z
∂x (b) ex sin y
∂V Ans: ux = ex sin y, uy = ex cos y
= −4e5z sin 3x sin 4y,
∂y
∂V (c) u = log x + x2 + y2 .
= 5e5z sin 3x cos 4y
∂z Ans:
∂2 V 1 x
= −9e sin 3x cos 4y = −9V
5z ux = 1+
dx2 x+ x2 + y2 x2 + y2
∂2 V 1
= −16e5z sin 3x cos 4y = −16V =
∂y2 x2 + y2
∂2 V
= 25e5z sin 3x cos 4y = 25V 1 y
∂z 2 uy = 0+
x+ x2 + y2 x2 + y2
Adding we get
y 1
= ·
∂V
2
∂V ∂V 2 2
x2 + y2 x+ x2 + y2
+ 2 + 2 = (−9 − 16 + 25) V = 0.
∂x2 ∂y ∂z
2. If u = tan (y + ax) + (y − ax)3/2 , then show that 4.5 Total Differential

∂2 u ∂2 u Coefficient
= a2 2 .
∂x 2 ∂y If u = f (x, y) is a function of two variables x and y
3. Show thatthe Laplace’s equation then its total differential coefficient denoted by du or
∂2 ∂2 ∂2 df is defined by
∇ 2V = 0 ∇ 2 = 2 + 2 + is satisfied
∂x ∂y ∂z
du or df =
∂u ∂u
dx + dy, (4.1)
by V x, y, z = (a) x2 + y2 − 2z 2 ; ∂x ∂y
(b) eax+by cos a2 + b2 z ; (c)
2 −1/2
whether or not x and y are independent of each other.
x + y2 + z 2 .
4. If x2(a2 + u)−1 +y2 (b2 + u)−1 +z 2 (c2 + u)−1 = 1, 4.6 Chain Rules for Partial
then prove that Differentiation
2 2 2
∂u ∂u ∂u If x = x(t), y = y(t) in Eq. (4.1), we get
+ +
∂x ∂y ∂z
du ∂u dx ∂u dy
∂u ∂u ∂u = + (4.2)
= 2 x +y +z . dt ∂x dt ∂y dt
∂x ∂y ∂z
If, instead, y is a function of x then Eq. (4.1) yields
5. If u = f (x2 /y), then prove that
∂u ∂u du ∂u ∂u dy
(a) x + y =0 = + (4.3)
∂x ∂y dx ∂x ∂y dx
∂2 u ∂2 u ∂2 u
(b) x2 2 + 3xy + 2y2 2 = 0.
∂x ∂x∂y ∂y The above results can be extended to a function of
more than two variables.
6. If u = r then prove that
n
∇ 2 u = n(n + 1)r n−2 r= x2 + y2 + z 2 . Example 4.12

If u = f (x − y, y − z, z − x), then show that
7. If u = er cos θ cos (r sin θ) , ∂u ∂u ∂u
v = er cos θ sin (r sin θ) then prove that + + = 0.
∂x ∂y ∂z
1 )
ur = vθ Solution Let
r (Cauchy–Riemann Equations)
uθ = −rvr
p = x − y, q = y − z, r = z − x (1)
8. Show that u = r n (3 cos2 θ − 1) satisfies ∂p ∂q ∂r
= 1, = 0, = −1 (2)
∂ ∂u 1 ∂ ∂u ∂x ∂x ∂x
r2 + sin θ =0 ∂u ∂f ∂p ∂f ∂q ∂f ∂r
∂r ∂r sin θ ∂θ ∂θ Now = · + · + ·
∂x ∂p ∂x ∂q ∂x ∂r ∂x
if n = – 3 or 2. ∂f ∂f
= − by Eq. (2) (3)
9. If n =−3/2,then show that θ = t n e−r
2 /4t
satisfies ∂p ∂r
1 ∂ ∂θ ∂θ ∂u ∂f ∂f
r2 = . Similarly = − (4)
r 2 ∂r ∂r ∂t ∂y ∂q ∂p
y z x ∂u ∂f ∂f
10. If u = + + , then show that = − (5)
z x y ∂z ∂r ∂q
∂u ∂u ∂u ∂u ∂u ∂u
x +y +z = 0. Adding, we get + + = 0.
∂x ∂y ∂z ∂x ∂y ∂z

Example 4.13 ∂2 u ∂ ∂u ∂ sin θ ∂
= = cos θ −
If z = f (x, y) where x = eu + e−v , y = e−u − ev then ∂x 2 ∂x ∂x ∂r r ∂θ
∂z ∂z ∂z ∂z
prove that − =x −y . ∂u sin θ ∂u
∂u ∂v ∂x ∂y cos θ −
∂r r ∂θ

Solution ∂ ∂u sin θ ∂u
= cos θ cos θ −
∂z ∂z ∂x ∂z ∂y ∂z ∂z −u ∂r ∂r r ∂θ
= · + = eu + −e
∂u ∂x ∂u ∂y ∂u ∂x ∂y
∂z ∂z ∂x ∂z ∂y ∂z −v ∂z sin θ ∂ ∂u sin θ ∂u
= · + = −e + (−ev ) − cos θ −
∂v ∂x ∂v ∂y ∂v ∂x ∂y r ∂θ ∂r r ∂θ

∂2 u sin θ ∂u sin θ ∂2 u
Subtracting, = cos θ cos θ 2 + 2 −
∂r r ∂θ r ∂r∂θ
∂z ∂z ∂z u ∂z −u
− = e + e−v + −e + ev sin θ ∂u ∂2 u
∂u ∂v ∂x ∂y − − sin θ + cos θ
∂z ∂z r ∂r ∂θ∂r
= x −y .
∂x ∂y cos θ ∂u sin θ ∂2 u
− −
r ∂θ r ∂θ 2
Example 4.14
∂2 u ∂2 u ∂2 u 2 sin θ cos θ ∂u sin2 θ ∂u
Transform 2 + 2 into polar coordinates. = cos2 θ + +
∂x ∂y ∂r 2 r2 ∂θ r ∂r
2 sin θ cos θ ∂ u
2
sin θ ∂ u
2 2
Solution
− + 2 (1)
r2 ∂r∂θ r ∂θ 2
Also,
r
y ∂u ∂u ∂r ∂u ∂θ ∂u cos θ ∂u
= + · = sin θ +
θ
∂y ∂r ∂y ∂θ ∂y ∂r r ∂θ

x ∂2 u ∂ ∂u
=
∂y2 ∂y ∂y
Figure 4.1
∂ cos θ ∂ ∂u cos θ ∂u
= sin θ + sin θ +
∂r r ∂θ ∂r r ∂θ
x = r cos θ, y = r sin θ
∂ ∂u cos θ ∂u
⇔ r= x2 + y2 , θ = tan−1 y/x = sin θ sin θ +
∂r ∂r r ∂θ
∂r x
= = cos θ; cos θ ∂ ∂u cos θ ∂u
∂x x2 + y2 + sin θ +
r ∂θ ∂r r ∂θ
∂r y
= = sin θ ∂2 u cos θ ∂u cos θ ∂2 u
∂y x2 + y2 = sin θ sin θ 2 − 2 +
∂r r ∂θ r ∂r∂θ
∂θ 1 y y sin θ
= 2 − 2 = − 2 =− ; cos θ ∂u ∂2 u
∂x 1 + y/x x x +y 2 r + cos θ + sin θ
r ∂r ∂θ∂r
∂θ 1 1 x cos θ
= sin θ ∂u cos θ ∂2 u
2 · = 2 = − +
∂y 1 + y/x x x + y 2 r r ∂θ r ∂θ 2
∂u ∂u ∂r ∂u ∂θ ∂2 u 2 sin θ cos θ ∂u cos2 θ ∂u
Now = · + . = sin2 θ 2 − +
∂x ∂r ∂x ∂θ ∂x ∂r r2 ∂θ r ∂r
∂u sin θ ∂u 2 sin θ cos θ ∂2 u cos2 θ ∂2 u
= cos θ − + + (2)
∂r r ∂θ r ∂r∂θ r 2 ∂θ 2
Adding Eqs. (1) and (2), we get ∂z ∂z ∂x ∂z ∂y

Also, = +
∂2 u ∂2 u ∂2 u 1 ∂u 1 ∂2 u ∂v ∂x ∂v ∂y ∂v
+ = + + . ∂z ∂z
∂x2 ∂y2 ∂r 2 r ∂r r 2 ∂θ 2 = − sin α + cos α
∂x ∂y
Example 4.15
∂2 z ∂ ∂z
If u = sin x + y
2 2
where a2 x2 + b2 y2 = c2 =
du ∂v2 ∂v ∂v
find .
dx ∂ ∂
= − sin α + cos α
∂x ∂y
Solution
du ∂u ∂u dy ∂z ∂z
= + − sin α + cos α
dx ∂x ∂y dx ∂x ∂y
2 dy ∂z
2
= cos x + y2 · 2x + cos x2 + y2 · 2y · = sin2 α 2 − 2 sin α cos α

dx ∂x
dy dy a2
x ∂2 z ∂2 z
But 2a2 x + 2b2 y =0⇒ =− 2 + cos2 α 2 (4)
dx dx by ∂x∂y ∂y

du 2 a 2
x Adding Eqs. (3) and (4), we get
∴ = 2 cos x + y2 x − y · 2
dx b y ∂2 z ∂2 z ∂2 z ∂2 z
2 + 2 = 2 + 2.
2 b − a2 x 2 ∂u 2 ∂v ∂x ∂y
= cos x + y2 .
b2
∂2 z ∂2 z ∂2 z ∂2 z
Show that 2 + 2 transforms to 2 + 2 4 4
∂u ∂v ∂x ∂y 1. If u = x − y + y − z + (z − x)4 , then show
if x = u cos α − v sin α, y = u sin α + v cos α. ∂u ∂u ∂u
that + + = 0.
∂x ∂y ∂z
Solution

x = u cos α − v sin α, 2. If V = f x, y , x = u2 − v2 and y = 2uv, then
y = u sin α + v cos α (1) prove that
∂x ∂x ∂V ∂V ∂V
= cos α, = − sin α; (a) u −v = 2 u2 + v 2
∂u ∂v ∂x ∂x ∂x
2
∂y ∂y ∂2 V ∂2 V 1 ∂V ∂2 V
= sin α, = cos α (2) (b) + = + .
∂u ∂v ∂x2 ∂y2 4 u2 + v 2 ∂u2 ∂v2
∂z ∂z ∂x ∂z ∂y ∂z ∂z
Now = + = cos α + sin α 3. If z = f (x, y), x = u − v and y = uv, then prove
∂u ∂x ∂u ∂y ∂u ∂x ∂y
that
∂2 z ∂ ∂z ∂z ∂z ∂z
= (a) (u + v) =u −v
∂u2 ∂u ∂u ∂x ∂u ∂v

∂ ∂ ∂z ∂z ∂z
= cos α + sin α (b) (u + v) = + .
∂x ∂y ∂y ∂u ∂v

∂z ∂z
cos α + sin α 4. If z = f x, y , x = eu sin v and y = eu cos v, then
∂x ∂y prove that
∂ 2
z ∂2 z ∂z ∂z ∂z
= cos2 α 2 + 2 sin α cos α (a) =x +y
∂x ∂x∂y ∂u ∂x ∂y
∂z
2
∂z −u ∂z ∂z
+ sin2 α 2 (3) (b) =e sin v + cos v .
∂y ∂x ∂u ∂v
5. If u = f (x, y) = φ(r, θ), x = r cos θ and y = 15. If u = sin−1 √x + y√ , then prove that
r sin θ, then prove that x+ y
∂u ∂u 1
∂u ∂u ∂u ∂u ∂u ∂u x +y = tan u.
r =x +y and = −y + x . ∂x ∂y 2
∂r ∂x ∂y ∂θ ∂x ∂y
4.7 Homogeneous Functions
xy
6. If u = , then show that A polynomial function in two variables x and y of the
x+y form
∂u ∂u
x +y = u. %
n
∂x ∂y
f (x, y) = ar xn−r yr
x y r=0
7. If u = f , , then prove that
z z = a0 xn + a1 xn−1 y + a2 xn−2 y2 + · · ·
∂u ∂u ∂u
x +y +z = 0. +ar xn−r yr + · · · + an yn (4.4)
∂x ∂y ∂z
where the sum of the indices of x and y in each term
∂2 u ∂2 u is n is called a homogenous function of degree n; n
8. Verify that = if u =
∂x∂y ∂y∂x being a positive integer.
Note that f (x, y) can be expressed as
(a) log x2 + y2 /xy (b) xy

(c) sin−1 (y/x) (d) log[x tan−1 (x2 + y2 )]. y y 2
f (x, y) = x a0 + a1
n
+ a2 + ···
x x
9. If u = (y − z)(z − x)(x − y) then show that y r y n
∂u ∂u ∂u +ar + · · · + an
+ + = 0. x x
∂x ∂y ∂z = xn f 1, y/x
= xn · φ(y/x)
10. If V = log (x − a)2 + (y − b)2 where (x − a)
∂2 V ∂2 V where φ is a function of the variable (y/x). Further,
(y − b) = 0, then prove that 2 + 2 = 0.
∂x ∂y if f (x, y) is a homogenous function of degree n
then f (tx, ty) = t n f (x, y) where t is some param-
11. If r 2 = x2 + y2 , then
prove that
2 2 eter. These observations enable us to enlarge the
∂r2
∂r
2
1 ∂r ∂r definition of the so-called homogenous functions to
+ 2 = + .
∂x2 ∂y r ∂x ∂y include trigonometric, inverse trigonometric, loga-
rithmic and other transcendental functions whose
du arguments satisfy the above liberalised conditions
12. Find if u = x2 + y2 and y = (1 − x)/x. for some n, which may be a rational number.
dx
du 2(x4 + x − 1)
Ans: = Corollary If u = f (x, y, z) is a homogeneous
dx x3
function of degree n, then we can write u in one of
dy the following forms:
13. If x3 + y3 = 3axy find .
dx
dy x2 − ay y z x z x y
Ans: =− 2 x φ1 ,
n
or y φ2
n
, or z n φ3 , .
dx y − ax x x y y z z
3 Example 4.17
x + y3
14. If u = tan−1 , then prove that Prove that
x−y
∂u ∂u x+y x3 + y3
x +y = sin 2u. (i) sin , (ii) tan−1 ,
∂x ∂y x2 + y2 x−y

x2 y2 x4 + y4 Example 4.18
(iii) exp , (iv) log
x+y x2 + y2 Prove that √ √
x+ y
are homogenous functions of degrees –1, 2, 3, 2, (i) 2x − xy + 17y , (ii)
2 2
sin √ √ ,
x x+y y
respectively. 2
−1 x + y2
(iii) tan are homogenous functions of
Solution xy
x+y degrees 2, –1, 0, respectively.
(i) Let z = sin 2 ⇒ sin−1 z
x + y2
Solution
x+y −1 1 + (y/x) (i) Let u = 2x2 − xy + 17y2
= =x 2 2
x2 + y2 1 + y/x = x2 2 − y/x + 17 y/x
y
= x−1 φ = x2 φ
y
x x
y y 2
y 1 + y/x where φ = 2− + 17 y/x
−1 x x
where φ = 2 sin z is a homoge-
x 1 + y/x u is a homogeneous function of degree 2
neous function of degree −1 and hence z. √ √
x+ y
3 (ii) Let v = sin √ √ ⇒ sin−1 v
−1 x + y3 x x+y y
(ii) Let u = tan ⇒ tan u 1/2
x−y
−1
1 + y/x −1 y
1 + (y/x)3 y =x 3/2 = x φ
= x2 = x2 φ 1 + y/x x
1 − (y/x) x 1/2
y 1 + y/x
where φ = 3/2
y 1 + (y/x)3 x 1 + y/x
where φ = tan u is a homogeneous
x 1 − (y/x)
function of degree 2 and hence u. sin−1 v and hence v is a homogeneous function of
2 2 degree −1
xy 2
(iii) Let v = exp ⇒ log v x + y2
x+y (iii) Let w = tan −1
⇒ tan w
2 xy
y/x y 2
=x 3
= x3 φ 1 + y/x y
1 + y/x x = = x0 φ
y/x x
2 2
y y/x y 1 + y/x
where φ = log v is a homogeneous where φ =
x 1 + y/x x y/x
function of degree 3 and hence v.
tan w and hence w is a homogeneous function of
4
x + y4 degree 0.
(iv) Let w = log 2 ⇒ ew
x + y2
4 4.8 Euler’s Theorem on
1 + y/x y
=x 2
2 = x φ
2 Homogeneous Functions
1 + y/x x
4 Theorem 4.1 If f is a homogeneous function in

y 1 + y/x variables x and y of degree n, then
where φ = w
2 e is a homogeneous
x 1 + y/x ∂f ∂f
function of degree 2 and hence w. x +y = nf (4.5)
∂x ∂y
Proof: Since f is a homogenous function of degree Multiplying Eq. (4.12) by x and Eq. (4.13) by y
n, we can write and adding, we get
y
f (x, y) = xn φ (4.6) ∂2 f ∂2 f ∂2 f
x x2 + y2 2 + 2xy
∂x 2 ∂y ∂x∂y
Differentiating Eq. (4.6) partially with respect to
x and y in turn ∂f ∂f
= (n − 1) x + y (4.11)
∂x ∂y
∂f y
= nxn−1 φ + xn · φ · − 2 (4.7) = n(n − 1)f [by Eq. (4.5) and Eq. (4.11)]
∂x x
∂f 1
= x ·φ ·
n
(4.8) Corollary 2 If z is a homogeneous function of x
∂y x and y and z = f (u), then
( denotes differentiation with respect to the argument
y/x) ∂u ∂u nf
(i) x +y = (4.14)
Multiplying Eq. (4.7) by x and Eq. (4.8) by y and ∂x ∂y f
adding ∂2 u ∂2 u ∂2 u
(ii) x2 2 +2xy +y2 2 = g(g − 1) (4.15)
∂f ∂f ∂x ∂x∂y ∂y
x +y = nxn φ − xn−1 yφ + xn−1 yφ = nf (4.9)
∂x ∂y
where g = nf /f and denotes differentiation with
Thus, the effect of applying the differential respect to ‘u’.
∂ ∂
operator x + y on a homogeneous function
∂x ∂y Proof: (i) z = z(x, y) is a homogeneous function
of degree n is to multiply f by n. of x and y of degree n.
Note By Euler’s theorem

The above theorem can be extended to the case of a ∂z ∂z
x + y = nz (4.16)
homogeneous function of m variables. If f is a homo- ∂x ∂y
geneous function of m variables xi (i = 1, 2, · · · m) of
order n, then Since z = f (u) we have
%
m
∂f ∂z ∂u ∂z ∂u
xi = nf . (4.10) =f , =f (4.17)
∂xi ∂x ∂x ∂y ∂y
i=1
so that Eq. (4.16), becomes

∂u ∂u
4.9 Extension of Euler’s Theorem f x +y = nf
∂x ∂y
Corollary 1 If f is a homogeneous function of
degree n and fxy = fyx , then ∂u ∂u nf
or x +y = [See Eq. (4.14)]
∂x ∂y f
∂f
2
∂f
2
∂f 2
x2 + 2xy + y2 2 = n(n − 1)f (4.11)
∂x2 ∂x∂y ∂y (ii) Writing nf /f = g in Eq. (4.14) and differ-
Proof: Differentiating Eq. (4.5) with respect to x
entiating Eq. (4.14) partially with respect to
and y in turn, we get x,
∂f ∂2 f ∂2 f ∂f ∂2 u ∂u ∂2 u ∂u
+x 2 +y =n (4.12) x +1· +y = g
∂x ∂x ∂x∂y ∂x ∂x 2 ∂x ∂x∂y ∂x
∂f
2
∂f ∂f
2
∂f ∂u
2
∂u
2
∂u
x + +y 2 = n (4.13) ⇒ x 2 +y = (g − 1) (4.18)
∂y∂x ∂y ∂y ∂y ∂x ∂x∂y ∂x
Differentiating Eq. (4.14) partially with respect ∂u (x + y)y − xy · 1

Now,=
to y, we get ∂x (x + y)2
2
y
∂2 u ∂2 u ∂u = ;
y + x = g − 1 (4.19) (x + y)2
∂y 2 ∂y∂x ∂y ∂u x2
=
∂y (x + y)2
Multiplying Eq. (4.18) by x and Eq. (4.19) by y ∂u ∂u y2 x2
∴ x +y =x + y
and adding ∂x ∂y (x + y)2 (x + y)2
xy(y + x) xy
∂2 u ∂2 u = = =u
2∂ u
2
x2 + 2xy + y (4.20) (x + y)2 x+y
∂x2 ∂x∂y ∂y2
Hence, Euler’s theorem is verified.
∂u ∂u
= (g − 1) x + y
∂x ∂y x2 y2
(iii) u(x, y) = is a homogeneous function
= g(g − 1) [by Eq. (4.14)] x2+ y2
of degree 2.
∂2 u ∂2 u
∵ = . (4.21) 2
∂x∂y ∂y∂x ∂u x + y2 2xy2 − x2 y2 · 2x
Now = 2
∂x x2 + y2
Example 4.19
2xy4
Verify Euler’s theorem for u(x, y) = = 2 ;
x2 + y2
(i) x2 + y2 (ii) xy/(x + y) ∂u 2x4 y
−1 Similarly = 2 ,
(iii) x2 y2 /(x2 + y2 ) (iv) x2 + xy + y2 ∂y x2 + y2

(v) log (x4 + y4 )/(x + y) . ∂u ∂u 2xy4 2x4 y
∴ x +y = x 2 + y 2
∂x ∂y x2 + y2 x2 + y2
Solution
2x2 y2 y2 + x2
= 2
(i) u(x, y) = x2 + y2 is a homogeneous func- x2 + y2
tion of degree 1. x2 y2
=2 = 2u
x2 + y2
∂u 1
Now = · 2x
∂x 2 x + y2
2 Hence, Euler’s theorem is verified.
−1
x ∂u y (iv) u(x, y) = x2 + xy + y2 is a homogeneous
= ; = function of degree –2.
x + y ∂y
2 2 x2 + y2
∂u −2
∂u ∂u x2 + y2 Now = − x2 + xy + y2 2x + y ;
∴ x +y = ∂x
∂x ∂y x2 + y2 ∂u −2
= − x2 + xy + y2 x + 2y
= x2 + y2 = u ∂y
∂u ∂u x(2x + y) + y(x + 2y)
∴ x +y =− 2
Hence, Euler’s theorem is verified. ∂x ∂y x2 + xy + y2
xy 1
(ii) u(x, y) = is a homogeneous function of = −2 = −2u
x+y x2 + xy + y2
degree 1.
x4 + y4 Solution
(v) u(x, y) = log x2 y2
x+y u = sin −1
x+y
= log x4 + y4 − log(x + y) x2 y2
⇒ sin u = = z(x, y) (say) (1)
x +y
4 4
x+y
Let f (u) = eu =
x+y x2 y2
4 Now z(x, y) = (2)
x 1 + yx
4 x+y
y
= = x3 φ
x 1+ x y
x is a homogeneous function of degree 3. By Euler’s
theorem on homogeneous functions we have
So, f is a homogeneous function of degree 3.
∂z ∂z
By Corollary 2 (i) of Euler’s theorem, x +y = 3z (3)
∂x ∂y
∂u ∂u nf ∂z dz ∂u ∂u
x +y = =3 But = · = cos u ;
∂x ∂y f ∂x du ∂y ∂z
(∵ f = eu = f and n = 3) ∂z dz ∂u ∂u
= · = cos u
∂u 4x3 1 ∂y du ∂y ∂y
Now = − ;
∂x x4 + y4 x+y
∂u ∂u
∂u 4y3 1 Now Eq. (3) becomes cos u x + y = 3 sin u
= − ∂x ∂y
∂y x4 + y4 x+y
∂u ∂u 4x3 4y3 ∂u ∂u
∴ x +y = x 4 + y or x +y = 3 tan u. (4)
∂x ∂y x + y4 x4 + y4 ∂x ∂y
1 1
−x −y
x+y x+y Example 4.22
= 4−1=3 x3 + y3
If u = tan−1 show that
x−y
(i) xux + yuy = sin 2u
Example 4.20
Verify Euler’s theorem for u(x, y, z) = xy + yz + zx. (ii) x2 uxx + 2xyuxy + y2 uyy = sin 4u − sin 2u.
[JNTU 1998]
Solution u(x, y, z) = xy + yz + zx is a homoge-
nous function of degree 2. Solution
∂u ∂u ∂u x3 + y3
= y + z, = x + z, =y+x u = tan−1
∂x ∂y ∂z x−y
∂u ∂u ∂u x3 + y3
x +y +z = x(y+z)+y(x+z)+z(y + x) ⇒ f (u) = tan u =
∂x ∂y ∂z x−y

= 2(xy + yz + zx) = 2u = z x, y say (1)
3
Hence, Euler’s theorem is verified. x3 + y3 x3 1 + yx
Now z x, y = =
Example 4.21 x−y x 1 − yx

x2 y2 = x2 φ y/x
If u = sin−1 then show that
x+y
xux + yuy = 3 tan u. is a homogeneous function of degree 2.
By Euler’s Theorem 2. Verify Euler’s theorem for

x y z
xzx + yzy = 2z (2) f x, y, z = + + .
y+z z+x x+y
∂z dz ∂u 2
But = · x + y2
∂x du ∂x 3. If u = log , then prove that
∂u ∂u x+y
= f (u) = sec2 u xux + yuy = 1.
∂x ∂x
∂z dz ∂u x y z
= · 4. If u = sin−1 + cos−1 + tan−1 , then prove
∂y du ∂y y z x
∂u ∂u ∂u ∂u ∂u
= f (u) = sec2 u · that x + y + z = 0. [JNTU 1993]
∂y ∂y ∂x ∂y ∂z
Substituting in Eq. (2), we get 5. Verify Euler’s theorem for V = f (u, v, w) where
y z x
u = , v = and w = .
∂u ∂u z x y
sec2 u x + y = 2 tan u
∂x ∂y
6. Verify Euler’s theorem for the function
∂u ∂u tan u f (x, y, z) = 3x2 yz+5xy2 z+4z 4 .
⇒ x +y =2 2 [JNTU 1997]
∂x ∂y sec u
3
= 2 sin u cos u −1 x + y3 + z 3
7. If u x, y, z = cos , then prove
= sin 2u (3) ax + by + cz
that xux + yuy + zuz = −2 cot u.
n
f (u) tan u x + yn
Since g (u) = n =2 2 8. If u = sin−1 √ √ , then prove that
f (u)
x +y
sec u
= sin 2u 1
xux + yuy = n − tan u.
and g (u) − 1 = 2 cos 2u − 1 2

by Corollary 2 (ii) of Euler’s theorem 9. If u x, y = xm yn tan−1 y/x , then prove that
xux +yuy = (m + n) u and x2 uxx +2xyuxy +y2 uyy =
x2 uxx + 2xyuxy + y2 uyy = g (u) g (u) − 1 (4) (m + n) (m + n − 1) u.
= sin 2u(2 cos 2u − 1) 1/2
x + y1/2
= sin 4u − sin 2u 10. If u = cosec −1
, then prove that
x1/3 + y1/3
x2 uxx + 2xyuxy + y2 uyy
EXERCISE 4.3 1
= tan u 13/12 + tan2 u/12 . [JNTU 1999S]
12
1. Verify Euler’s theorem if u(x, y) =
11. If u = x2 tan−1 (y/x) − y2 tan−1 (x/y), then prove
(a) x3 + y3 that x2 uxx + 2xyuxy + y2 uyy = 2u.
x+y
(b) tan−1
x−y 4.10 Jacobian
5/2
(c) ax + by In the case of transformation of variables we need
−1 −1
to use determinants called the Jacobians whose ele-
(d) cos (x/y)+cot y/x
n ments are functions. Let u(x, y) and v(x, y) be two
x + yn
(e) log (f) xy/ x2 + y2 · given functions of the two independent variables x
x+y and y.
The Jacobian
of u, v with respect to x, y denoted We now have
u, v ∂ (u, v) ∂u ∂u
by J or is a second order functional ∂x ∂x
x, y ∂ x, y ∂x ∂y ∂u ∂v
determinant defined by J · J = ∂v ∂v ∂y ∂y
∂u ∂u ∂x ∂y ∂u ∂v

u, v ∂ (u, v) ∂x ∂y ∂u ∂x ∂u ∂y ∂u ∂x ∂u ∂y
J or = + +
x, u ∂ x, y ∂v ∂v ∂x ∂u ∂y ∂u ∂x ∂v ∂y ∂v
=
∂x ∂y ∂v ∂x ∂v ∂y ∂v ∂x ∂v ∂y
+ +
and the Jacobian of three functions u, v, w of three ∂x ∂u ∂y ∂u ∂x ∂v ∂y ∂v
independent variables x, y, z is defined by 1 0
= = 1.
0 1
∂u ∂u ∂u
∂x ∂y ∂z
Chain rule for the Jacobians
u, v, w ∂ (u, v, w) ∂v ∂v ∂v
J or = . Theorem 4.3 If u, v are functions of r, s and r, s
x, y, z ∂ x, y, z ∂x ∂y ∂z
are in turn functions of x, y, then
∂w ∂w ∂w ∂(u, v) ∂(u, v) ∂(r, s)
= .
∂x ∂y ∂z ∂(x, y) ∂(r, s) ∂(x, y)
4.10.1 Properties Proof: Differentiating u, v partially with respect to
Inverse x, y, we get
∂u ∂u ∂r ∂u ∂s
Theorem 4.2 = + (4.22)
∂x ∂r ∂x ∂s ∂x
∂(u, v)∂(x, y)
If J = J = , , then ∂u ∂u ∂r ∂u ∂s
∂(x, y)∂(u, v) = + (4.23)
∂y ∂r ∂y ∂s ∂y
∂ (u, v) ∂ x, y
J .J = · = 1. ∂v ∂v ∂r ∂v ∂s
∂ x, y ∂ (u, v) = + (4.24)
∂x ∂r ∂x ∂s ∂x
Proof:
∂v ∂v ∂r ∂v ∂s
= + (4.25)
Let u = f (x, y) and v = g(x, y) (4.20) ∂y ∂r ∂y ∂s ∂y
be two given functions of x and y. Solving these we By definition of the Jacobian
get x and y as functions of u and v. ∂u ∂u ∂r ∂r
x = φ (u, v) and y = ψ (u, v) (say) (4.21) ∂ (u, v) ∂ (r, s) ∂r ∂s ∂x ∂y
· =
∂ (r, s) ∂ x, y ∂v ∂v ∂s ∂s
Differentiating Eq. (4.20) partially with respect to u
and v in turn, we have ∂r ∂s ∂x ∂y
∂u ∂r ∂u ∂s ∂u ∂r ∂u ∂s
∂u ∂u ∂x ∂u ∂y + +
1= = + ∂r ∂x ∂s ∂x ∂r ∂y ∂s ∂y
∂u ∂x ∂u ∂y ∂u =
∂v ∂r ∂v ∂s ∂v ∂r ∂v ∂s
∂u ∂u ∂x ∂u ∂y + +
0= = + ∂r ∂x ∂s ∂x ∂r ∂y ∂s ∂y
∂v ∂x ∂v ∂y ∂v
∂v ∂v ∂x ∂v ∂y ∂u ∂u
0= = +
∂u ∂x ∂u ∂y ∂u ∂x ∂y ∂ (u, v)
= = [by Eqs. (4.22)–(4.25)]
∂v ∂v ∂x ∂v ∂y ∂v ∂v ∂ x, y
1= = + ·
∂v ∂x ∂v ∂y ∂v ∂x ∂y
0
Thus the Jacobians in a certain way behave like ∂(u, v, w) ∂ f1 , f2 , f3 ∂(x, y, z)
= (−1)3 0
derivatives. The above results can be extended to ∂(x, y, z) ∂ f1 , f2 , f3 ∂(u, v, w)
more than two variables. 0
∂(x, y, z) ∂ f1 , f2 , f3 ∂(u, v, w)
and = (−1)−3 0 .
∂(u, v, w) ∂ f1 , f2 , f3 ∂(x, y, z)
Functional dependence
Theorem 4.4 If u, v, w are ‘functionally related’ 4.11 Standard Jacobians

∂(u, v, w)
functions of x, y, z, then = 0. We will now find the Jacobians for changing the
∂(x, y, z) variables from cartesian coordinates (x, y, z) to
Proof: Let u, v, w be not independent. Suppose
(i) Polar coordinates: (r, θ)
f (u, v, w) = 0 (4.26) x = r cos θ, y = r sin θ (in plane z is absent)
Differentiating Eq. (4.26) with respect to x, y, z in (ii) Cylindrical polar coordinates: (r, θ, z)
turn, we have x = r cos θ, y = r sin θ, z = z
∂f ∂u ∂f ∂v ∂f ∂w
· + · + · =0 (4.27) (iii) Spherical polar coordinates: (r, θ, φ)
∂u ∂x ∂v ∂x ∂w ∂x
x = r sin θ cos φ, y = r sin θ sin φ, z = r cos θ
∂f ∂u ∂f ∂v ∂f ∂w
· + · + · =0 (4.28)
∂u ∂y ∂v ∂y ∂w ∂y Solution (i) x = r cos θ, y = r sin θ
∂f ∂u ∂f ∂v ∂f ∂w
· + · + · =0 (4.29) ∂x ∂x
∂u ∂z ∂v ∂z ∂w ∂z = cos θ, = −r sin θ;
∂r ∂θ
∂f ∂f ∂f ∂y ∂y
Eliminating f , i.e., , , from Eqs. (4.27) – = sin θ, = r cos θ
∂u ∂v ∂w ∂r ∂θ
(4.29)
∂x ∂x

∂u ∂v ∂w ∂ x, y ∂r ∂θ cos θ −r sin θ
∂x ∂x ∂x ∴ = =
∂ (r, θ) ∂y ∂y sin θ r cos θ
∂u ∂v ∂w
∂y ∂y ∂y
=0 (4.30) ∂r ∂θ
= r cos2 θ + sin2 θ = r
∂u ∂v ∂w
∂z ∂z ∂z Solving for r and θ :
∂u ∂u ∂u r = x2 + y2 , θ = tan−1 (y/x)
∂x ∂y ∂z ∂r x ∂r y
∂v ∂v ∂v = , = ;
=0 (4.31) ∂x r ∂y r
∂x ∂y ∂z
∂θ −y ∂θ x
∂w ∂w ∂w = 2, = 2
∂x r ∂y r
∂x ∂y ∂z
∂r ∂r x y
interchanging the rows and columns.
∂ (r, θ) ∂x ∂y r r
∴ = =
∂ x, y ∂θ ∂θ −y x
Jacobians of implicit functions
∂x ∂y r2 r2
If u, v, w are implicit functions of x, y, z, i.e., x2 y2 1
fi (u, v, w, x, y, z) = 0 for i = 1, 2, 3 then = 3
+ 3
= ∵ x2 + y2 = r 2 .
r r r
(ii) x = r cos θ, y = r sin θ, z = z ex 0

= = ex · ey = uv
0 ey
∂x ∂x ∂x
= cos θ, = −r sin θ, =0 Solving for x, y we get x = log u, y = log v
∂r ∂θ ∂z
∂x ∂x
∂y ∂y ∂y
= sin θ, = r cos θ, =0 ∂ x, y ∂u ∂v 1
0 1
∂r ∂θ ∂z J = = = u 1 =
∂ (u, v) ∂y ∂y 0 v
uv
∂z ∂z ∂z ∂u ∂v
= 0, = 0, =1
∂r ∂θ ∂z
Note that JJ = 1.
cos θ −r sin θ 0
∂ x, y, z (ii) u = ex+y , v = e−x+y
∴ J = = sin θ r cos θ 0
∂ (r, θ, z) 0 0 1 ∂u ∂u
2 ∂ (u, v) ∂x ∂y
= r cos θ + sin θ = r
2 J = = ∂v ∂v
∂ x, y
∂ (r, θ, z) 1 1 ∂x ∂y
J = = = .
∂ x, y, z J r ex+y ex+y
= = e2y + e2y = 2e2y
(iii) x = r sin θ cos φ, y = r sin θ sin φ, −e−x+y e−x+y
z = r cos θ u 1
Solving = e2x ⇒ x = log(u/v);
∂ x, y, z v 2
uv = e2y ⇒ y = 12 log(uv)
∂ (r, θ, φ)
∂x ∂x
sin θ cos φ r cos θ cos φ −r sin θ sin φ
∂ x, y ∂u ∂v
= sin θ sin φ r cos θ sin φ r sin θ cos φ J = =
∂ (u, v) ∂y ∂y
cos θ −r sin θ 0 ∂u ∂v
v 1 v −u
= r 2 sin θ cos2 θ + r 2 sin θ sin2 θ · ·
2u v 2u v2
= r 2 sin θ =
1 1
∂ (r, θ, φ) 1 1 ·v ·u
∴ = = 2 . 2uv 2uv
∂ x, y, z ∂ x, y, z r sin θ 1 1 1 1
= + = = 2y
∂ (r, θ, φ) 4uv 4uv 2uv 2e
∴ JJ = 1.
Example 4.23
∂ (u, v) ∂ x, y Example 4.24
Find and and verify that x+y ∂ (u, v)
∂ x, y ∂ (u, v) If u = , v = tan−1 x + tan−y find .
∂ (u, v) ∂ x, y 1 − xy ∂ x, y
. =1 Show that u and v are functionally dependent and
∂ x, y ∂ (u, v)
(i) u = ex , v = ey find the relation between them.
(ii) u = ex+y ; v = e−x+y .
Solution

Solution ∂u 1 − xy · 1 − x + y −y
∂u ∂u = 2
∂x ∂x 1 − xy
∂ (u, v) vy
(i) J = = 1 + y2
∂ x, y ∂v ∂v = 2
∂x ∂y 1 − xy

∂u 1 − xy 1 − x + y (−x) 1 + x2 = (1 − v) u2 v (1 − w) + u2 vw
= 2 = 2
∂y 1 − xy 1 − xy +u uv2 (1 − w) + uv2 w
∂v 1 ∂v 1 = (1 − v) u2 v + u2 v2
= =
∂x 1+x 2 ∂y 1 + y2 = u2 v (1 − v + v) = u2 v.
∂u ∂u
∂x ∂y Example 4.26
∂ (u, v)
∴ = If x = u (1 − v) , y = uv, then prove that
∂ x, y ∂v ∂v ∂ x, y ∂ (u, v)
∂x ∂y · = 1.
∂ (u, v) ∂ x, y
1 + y2 1 + x2
2 2 Solution
1 − xy 1 − xy
= ∂x ∂x
1 1
∂ x, y ∂u ∂v 1−v −u
1 + x2 1 + y2 = =
∂ (u, v) ∂y ∂y v u
1 1 ∂u ∂v
= 2 − 2 = 0 = u (1 − v) + uv = u
1 − xy 1 − xy
u − uv = x ⇒ u = x + y
So, u and v are functionally dependent. y y
v= =
Also, v = tan−1 x + tan−1 y u x+y
x+y
= tan−1 = tan−1 u ∂u ∂u
1 − xy
or u = tan v. ∂ (u, v) ∂x ∂y
∴ =
∂ x, y ∂v ∂v
Example 4.25
If u = x + y + ∂x ∂y
z, uv = y + z, uvw = z, then prove
∂ x, y, z 1 1
that = u2 v.
∂ (u, v, w)
= y x+y ·1−y·1
− 2 2
Solution x+y x+y
z = uvw; y = uv − uvw = uv (1 − w) ; x y
= 2 + 2
x = u − uv = u (1 − v) x+y x+y
1 1
∂x ∂x ∂x = =
x+y u
∂u ∂v ∂w
We see that,
∂ x, y, z ∂y ∂y ∂y
=
∂ (u, v, w) ∂u ∂v ∂w ∂ x · y ∂ (u, v) 1
· = u · = 1.
∂z ∂z ∂z ∂ (u, v) ∂ x, y u
∂u ∂v ∂w
1−v −u 0 Example 4.27
x2 x3 x3 x1 x1 x2
If y1 = , y2 = , y3 = , then show
= v (1 − w) u (1 − w) −uv x1 x2 x3
∂ y1 , y2 , y3
that = 4.
vw uw uv ∂ (x1 , x2 , x3 )
Solution 2lx + 2my 2mx + 2ny

∂y1 ∂y1 ∂y1 ⇒ =0
∂x1 ∂x2 ∂x3 2l x + 2m y 2m x + 2n y

∂ y1 , y2 , y3 ∂y2 ∂y2 ∂y2 ⇒ lx+my m x + n y = l x+m y mx + ny
=
∂ (x1 , x2 , x3 ) ∂x1 ∂x2 ∂x3
⇒ lm x2 + ln + mm xy + mn y2
∂y3 ∂y3 ∂y3

∂x1 ∂x2 ∂x3 = l mx2 + l n + mm xy + m ny2
−x2 x3 x3 x2
Since x and y are independent, the coefficients of
x12 x1 x1
like terms must be equal.
x3 x3 x1 x1
= − 2
x2 x2 x2
x2 x1 x1 x2 ∴ lm = l m ln = l n and mn = m n
− 2 l m n
x3 x3 x3 ⇒ = = .
l m n
−x2 x3
x3 x2
x1 Example 4.29
1 x3 x1
= x3 − x1 If x = a cosh
ucos v and y = a sinh u sin v, then
x1 x2 x3 x2 ∂ x, y a2
x1 x 2 show that = (cosh 2u − cos 2v) .
x2 x1 − ∂ (u, v) 2
x3
1 1 1 Solution
taking out , , as common factors from
x1 x2 x3
R1 , R2 , R3 , respectively ∂x ∂x

1 x2 x3 2 ∂ x, y ∂u ∂v
= − (x − x12 ) − x3 (−x1 x2 − x1 x2 ) =
x1 x2 x3 x1 1 ∂ (u, v) ∂y ∂y

∂u ∂v
+x2 (x3 x1 + x3 x1 ) a sinh u cos v −a cosh sin v
=
1
= · 4x1 x2 x3 = 4. a cosh u sin v a sinh u cos v
x1 x2 x3
= a2 sinh2 u cos2 v + cosh2 u sin2 v

Example 4.28 = a2 cosh
2 u − 1 cos2 v

If u = lx2 + 2mxy + ny2 and v = l x2 + 2m xy + n y2 . + cosh2 u 1 −cos
2 v
If u and v are functionally dependent, then prove that
l m n = a2 cosh2 u − cos2 v
= = .
l m n a2
= (cosh 2u − cos 2v) .
2
Solution u and v are functionally dependent
∂u ∂u
Example 4.30
∂ (u, v) ∂x ∂y
⇒ = =0 If F = xu + v − y, G = u2 + vy + w, H = zu − v + vw
∂ x, y ∂v ∂v ∂ (F, G, H )
compute .
∂x ∂y ∂ (u, v, w)
Solution 8. If u = x + y + z, v = x2 + y2 + z 2 ,
∂ (u, v, w)
∂F ∂F ∂F w = x3 + y3 + z 3 − 3xyz find .
∂ x, y, z
∂u ∂v ∂w Are u, v, w functionally dependent? If so, find the
∂ (F, G, H ) ∂G ∂G ∂G relation.
=
∂ (u, v, w) ∂u ∂v ∂w Ans: Yes. 2w = u(3v − u2 )
∂H ∂H ∂H −1/2 −1/2
∂u ∂v ∂w 9. If u = x 1 − r 2 , v = y 1 − r2 ,

2 −1/2
x 1 0 w =z 1−r where r = x + y + z 2 show
2 2 2
∂ (u, v, w) −5/2
= 2u y 1 that = 1 − r2 .
∂ x, y, z
z −1 + w v y2 x2 + y2
10. If u = , v = , then show that
2x 2x
u, v −y
= x yv + 1 − w − 1 (2uv − z) J = .
x, y 2x
= x yv + 1 − w + z − 2uv.
EXERCISE 4.4 4.12 Maxima and Minima

4.12.1 Taylor’s Theorem for a
∂ (u, v) Function of Two Variables
1. If u = ax + by and v = cx + dy find .
∂ x, y
In calculus, we have studied Taylor series expan-
Ans: (ad − bc)
sion of a function of a single variable. If we know
2. If u = x + ωy, v = ω2 x + y where ω is a cube root the values of a function f and all its derivatives
of unity, then show that u and v are functionally f , f , · · · at a point P(x), then we can find its value at
dependent and find the relation between them. a neighbouring point Q(x+h) through the expansion
Ans: u = ωv
h h2
f (x + h) = f (x) + f (x) + f (x) + · · ·
3. If u = (cos α)x + (sin β)y, v = (sin α)x + (cos β)y 1! 2!
and u and v are functionally dependent, then show h(r) (r)
+ f (x) + · · · (4.32)
that (α + β) is equal to an odd multiple of π/2. r!
4. If x = u cosv, y = u sin v, then prove that d

Let D ≡ be the differential operator such that
∂ (u, v) ∂ x, y dx
· = 1. d n
∂ x, y ∂ (u, v) Dn f = f (n = 1, 2, 3, · · ·) and Do f ≡ 1 · f = f .
dxn
Assuming that D obeys the laws of algebra, we can
5. If x = eu cos v, y = eu sin v, then show that
∂ x, y ∂ (u, v) express Eq. (4.32) in a compact form:
· = 1.
∂ (u, v) ∂ x, y hD (hD)2 (hD)r
f(x + h) = 1+ + +· · ·+ +· · · f (x)
1! 2! r!
6. If x = u, y = u tan v, z = w, then show that
∂ x, y, z = e f (x)
hD
(4.33)
= u sec2 v.
∂ (u, v, w)
where ehD is a differential operator producing the
7. If x = v + w, y = w + u, z = u + v, then show same effect as the differential expression in the
∂ x, y, z brackets, when applied on a function possessing
that = 2. derivatives of all orders.
∂ (u, v, w)
We now extend this to a function of more than one 1

= f x, y + hfx x, y + kfy x, y
variable. Consider a function f of two variables x 1!
and y. Let Dx , Dy denote partial differential operators 1 2
+ h fxx x, y + 2hkfxy x, y
∂ ∂ 2!
, , respectively. If we know the values of f and 1
∂x ∂y
all its derivatives at P(x, y), we can express the value +k 2 fyy x, y + h3 fxxx (x, y)+3h2 kfxxy (x, y)
3!
of the function at a neighbouring point Q(x+h, y+k) +3hk 2 fxyy x, y + k 3 fyyy (x, y) + · · · (4.35)
in terms of these.
Incrementing the variable x keeping y fixed
Note 1

f (x + h, y) = ehDx f x, y Here
% ∞
(hDx )r ⎫
= f x, y ∂2 f ⎪
⎪
r! = fyx = fxy , ⎪
⎪
r=0 ∂x∂y ⎪
⎪
⎬
hDx (hDx )2 (hDx )r ∂3 f
= 1+ + +· · ·+ +· · · f (x, y) = fyxx = fxxy , (4.36)
1! 2! r! ∂x2 ∂y ⎪
⎪
⎪
⎪
∂3 f ⎪
⎪
h ∂ h2 ∂2 = fyyx = fxyy ⎭
= f x, y + f x, y + f x, y + · · · ∂x∂y2
1! ∂x 2! ∂x2
hr ∂r
+ f x, y + · · · (4.34) because of the continuity of these derivatives.
r! ∂xr
Note 2
Then incrementing the variable y in Eq. (4.34) f (a + h, b + k)

f x + h, y + k = ehDx f x, y + k
1
% ∞
(hDx )r = f (a, b) + hfx (a, b) + kfy (a, b)
= f x, y + k 1!
r=0
r! 1 2
kDy + h fxx (a, b) + 2hkfxy (a, b) + k 2 fyy (a, b)
=e hDx
e f x, y = ehDx +kDy f x, y 2!

1 ∂ ∂ 1 3
= f x, y + h +k f x, y + h fxxx (a, b) + 3h2 kfxxy (a, b)
1! ∂x ∂y 3!
2
1 ∂ ∂ + 3hk 2 fxyy (a, b)+k 3 fyyy (a, b) +· · · (4.37)
+ h +k f x, y
2! ∂x ∂y

1 ∂ ∂ 3 Note 3
+ h +k f x, y + · · ·
3! ∂x ∂y Put a = b = 0 and h = x, k = y, we get Maclaurin’s

1 ∂ ∂ series
= f x, y + h f x, y + k f x, y
1! ∂x ∂y
1
1 ∂ 2 ∂ 2 f x, y = f (0, 0) + xfx (0, 0) + yfy (0, 0)
+ h2 2 f x, y + 2hk f x, y 1!
2! ∂x ∂x∂y 1 2
+ x fxx (0, 0) + 2xyfxy (0, 0)
∂ 2 1 ∂3 2!
+k 2 2 f x, y + h3 3 f x, y
∂y 3! ∂x 1 3
∂ 3 ∂3 +y2 fyy (0, 0) + x fxxx (0, 0)
+3h2 k 2 f x, y + 3hk 2 f x, y 3!
∂x ∂y ∂x∂y 2
+3x2 yfxyy (0, 0) + 3xy2 fxyy (0, 0)
∂ 3
+k 3 3 f x, y + · · · +y3 fyyy (0, 0) + · · · (4.38)
∂y
e
Note 4
fxx x, y = ex cos y fxx (1, π/4) = √
f x, y = f a + x − a, b + y − b 2
e
fxy x, y = −ex sin y fxy (1, π/4) = − √
1 2
= f (a, b) + [(x − a) fx (a, b) e
1! fyy x, y = −ex cos y fyy (1, π/4) = − √
1 2
+(y − b)fy (a, b) + (x − a)2 fxx (a, b)
2! Substituting in Eq. (4.39), we have
+2 (x − a) y − b fxy (a, b)
2 e -
ex cos y = √ 1 + [(x − 1) + (y − π/4)]
+ y − b fyy (a, b) 2
1 1
+ (x − a)3 fxxx (a, b) (x − 1)2 − 2(x − 1)(y − π/4)
2!
2 .
3!

+3(x − a)2 (y − b)fxxy (a, b) − y − π/4 + · · · .
2
+3 (x − a) y − b fxyy (a, b)
3 Example 4.33
+ y − b fyyy (a, b) + · · · . (4.39) Find Taylor’s series expansion for tan−1 (y/x) about
(1, 1).
Example 4.31
Expand ex+y in a neighbourhood of (1,1). Solution

Solution Let Let f x, y = tan−1 (y/x) f (1, 1) = π/4
1 y y
f x, y = ex+y f (1, 1) = e2 fx x, y =
2 − 2 = − 2
1 + y/x x x + y2
fx x, y = ex+y fx (1, 1) = e2
1
fy x, y = ex+y fy (1, 1) = e2 fx (1, 1) = −
2
fxx x, y = ex+y fxx (1, 1) = e2 1 1 x
fy x, y = 2 · = 2
fxy x, y = ex+y fxy (1, 1) = e2 1 + y/x x x + y2

fyy x, y = ex+y fyy (1, 1) = e2 1
fy (1, 1) =
Substituting in Eq. (4.39), we have 2
2xy 1
fxx x, y = 2 fxx (1, 1) =
f x, y = ex+y x2 + y2 2
/ 2
1 x + y2 1 − x (2x) y2 − x2
= e2 1+[(x − 1)+(y − 1)]+ (x − 1)2 fxy x, y = =
2! 2 2
x2 + y2 x2 + y2
+2(x − 1)(y − 1)+(y − 1) + · · · .
2
fxy (1, 1) = 0
Example 4.32 −2xy 1
Obtain Taylor’s series expansion for ex cos y about fyy x, y = 2 fyy (1, 1) = −
x + y2 2
the point (1, π/4).
Substituting in Eq. (4.39), we have
Solution
e y
Let f x, y = ex cos y f (1, π/4) = √ f x, y = tan−1
2 x
e π 1 1
fx x, y = ex cos y fx (1, π/4) = √ = − (x − 1) + y−1
2 4 2 2

fy x, y = −ex sin y fy (1, π/4) = − √
e 1 2
+ (x − 1)2 − y − 1 + · · ·
2 4

π 1 1 (x + h) y + k
= − (x − y)+ (x − y)(x + y − 2)+· · · 3. Expand in powers of h, k up to the
4 2 4 x+h+y+k
second degree term.
on simplification. (x + h) y + k
Hint: Let f x + h, y + k =
Example 4.34 x+h+y+k
xy
Expand ex log(1 + y) in Taylor’s series about (0, 0). put h = k = 0, f x, y = . Calculate fx , fy ,
x+y
[JNTU 1996]
fxx , fxy , fyy and substitute in Eq. (4.35).
Solution xy y2 x2 y2
Ans: + 2 h + 2 k − 3 h
2
x+y x+y x+y x+y
Let f x, y = ex log 1 + y f (0, 0) = 0
2xy x2 k 2
fx x, y = e log 1 + y
x
fx (0, 0) = 0 + hk − 3 + · · ·
(x + y)3 x+y
e x
fy x, y = fy (0, 0) = 1
1+y
(1 + x + y ) at (1,2 0). 2
2 1/2
4. Expand
fxx x, y = ex log 1 + y = fxxx x, y √ x − 1 (x − 1) y
Ans: 2 1 + − + + ···
fxx (0, 0) = 1 = fxxx (0, 0) 4 32 4
ex
fxy x, y = = fxxy x, y 5. Expand cos(x + h)(y + k) in increasing powers of
1+y h and k.
fxy (0, 0) = 1 = fxxy (0, 0) 1 2 2
Ans: cos xy− hy sin xy + kx sin xy − h y cos xy
−ex 2!
fyy x, y = 2 = fxyy x, y +2hk sin xy + xy cos xy +k x cos xy + · · ·
2 2
1+y
fyy (0, 0) = −1 − fxyy (0, 0) 6. Expand e2x sin 3y in powers of x and y up to third
2ex degree.
fyyy x, y = 3 1 2
1+y Ans: 3y + 6xy + 36x y − 27y3 + · · ·
3!
fyyy (0, 0) =2
Substituting in Eq. (4.38), we have 4.12.2 Maxima and Minima of a

Function of Two Variables
f x, y = ex log 1 + y Let u = f (x, y) be a function of two variables x and y.
1 1 1 1
= y+xy− y2 + x2 y− xy2 + y3 +· · · . Relative maximum The function u = f (x, y) is
2 2 2 3 said to have a relative maximum at a point P(a, b) if
EXERCISE 4.5 f (a, b) > f (a + h, b + k) for all points Q(a + h, b + k)
in some neighbourhood of P(a, b).
1. Expand xy
e about (1, 1) by Taylor’s theorem. Relative minimum The function u = f (x, y) is
(x − 1)2 said to have a relative minimum at a point P(a, b) if
Ans: e 1 + (x − 1) + (y − 1) +
2! f (a, b) < f (a + h, b + k) for all points Q(a + h, b + k)
(y − 1)2 in some neighbourhood of P(a, b).
+(x − 1)(y − 1) + + ···
2! Let f = f (a + h, b + k) − f (a, b), then f has
a minimum at P(a, b) if f > 0 and a maximum at
2. Expand ex sin y as Taylor’s series up to the third
P(a, b) if f < 0.
degree term.
1 1 The maximum or minimum of a function is called
Ans: y + xy + x2 y − y3 + · · · its extreme value or an extremum.
2 6
We know that u = f (x, y) represents a surface. Thus the sign of f depends upon rt − s2 , which
The maximum of f corresponds to the top point and we denote by
the minimum of f to the bottom point of the sur-
face. Saddle point or minimax is a point where f has f > 0 if > 0 and r >0
neither minimum nor maximum. An example for a f < 0 if > 0 and r <0
saddle point is the leather seat on the back of a horse.
Hence u = f (x, y) has an extreme value at P(a, b)
The origin is a saddle point for the surface z = xy
if p = fx = 0, q = fy = 0 at P(a, b). At P(a, b), f
(hyperbolic paraboloid).
has a
4.12.3 Necessary and Sufficient (i) maximum if = rt − s2 > 0 and r < 0;

Conditions for the Extrema
of a Function of Two (ii) minimum if = rt − s2 > 0 and r > 0;
Variables
(iii) saddle point (minimax) if = rt − s2 < 0
By Taylor’s theorem ( f > 0 or < 0 depending on h, k); and
f = f (a + h, b + k) − f (a, b) there occurs;
1 2
= hfx (a, b) + kfy (a, b) + h fxx (a, b) (iv) case of failure: if rt − s2 = 0, and the test fails.
2! A further investigation is required to decide the
+2hkfxy (a, b) + k 2 fyy (a, b) + · · · case.

= hfx + kfy (a,b) (4.40) Example 4.35
1 Find the extrema of f (x, y) if
+ h2 fxx +2hkfxy + k 2 fyy (a,b) +· · ·
2! (i) f x, y = a2 − x2 − y2
f will have the same positive or negative sign if
(ii) f x, y = x3 y2 − xy.
hfx + kfy (a,b) = 0 (4.41)
A necessary condition for a function u = f (x, y) to Solution (i) Differentiating f with respect to x
have an extremum is that fx |(a,b) = 0 and fy (a,b) = 0. and y, we get
Then the point P(a, b) is called a stationary point of fx = −2x, fy = −2y
u = f (x, y) and the value of the function a stationary
value. Extremum occurs only at stationary points. fx = 0, fy = 0
At the point P(a, b) let the partial differential ⇒ (0, 0)is a critical point or stationary point of f
coefficients of u = f (x, y) be denoted as follows:
r = fxx = −2, t = fyy = −2, s = fxy = 0
∂f ∂f ∂2 f ∂2 f ∂2 f
= p, = q, 2 = r, = s, 2 = t (4.42) ∴ = rt − s2 = (−2) (−2) − 0
∂x ∂y ∂x ∂x∂y ∂y
= 4>0 and r = −2 < 0
Then from Eq. (4.41) we have, on neglecting higher
powers of h, k ∴ (0, 0) is a point of maximum value of
1 2
f = h r + 2hks + k 2 t f and fmax = a2
2!
So the sign of f is the same as that of .
(ii) Differentiating f with respect to x and y, we
h2 r 2 + 2hkrs + k 2 rt get
h r + 2hks + k t =
2 2
r fx = 3x2 y2 − y, fy = 2x3 y − x
(hr + ks)2 + k 2 rt − s2
= fx = fy = 0 when x, y = (0, 0)
r
∴ (0, 0) is a stationary point of f At Q = (0, 0), = rt − s2 = 0 and so fur-

ther investigation is required. Along the line y =
Also r = fxx = 6xy2 , s = fxy = 6x2 y − 1, x, f (x, y) = x5 (1 − 2x) > 0 for x = 0.1 and < 0
t = fyy = 2x3 for x = −0.1, i.e., in the neighbourhood of (0,0)
∴ = rt − s2 = −1 < 0 at (0, 0) there are points where f (x, y) > f (0, 0) and there are
points where f (x, y) < f (0, 0). Hence f (0, 0) is not
(0, 0) is a saddle point of f . an extreme value.

Discuss the minima and maxima of Find the maximum or minimum of
f (x, y) = x3 y2 (1 − x − y). f (x, y) = x3 + y3 − 3x − 12y + 20.
Solution We have Solution We have fx = 3x2 − 3, fy = 3y2 − 12

fx = 0 ⇒ x = ±1 and fy = 0 ⇒ y = ±2
fx = 3x2 y2 − 4x3 y2 − 3x2 y3 , ∴ Stationary points are A = (1, 2), B = (1, −2),
fy = 2x3 y − 2x4 y − 3x3 y2 C = (−1, 2), D = (−1, −2)
Also, r = fxx = 6x, s = fxy = 0; t = fyy = 6y
r = fxx = 6xy2 − 12x2 y2 − 6xy3 , ∴ = rt − s2 = 36xy
s = fxy = 6x2 y − 8x3 y − 9x2 y2 |A = 36 × 1 × 2 = 72 > 0 and
t = fyy = 2x3 − 2x4 − 6x3 y. r = 6x = 6 × 1 = 6 > 0
fx = 0, fy = 0
∴ f x, y has a minimum at A and

⇒ x2 y2 3 − 4x − 3y = 0, fmin = 1 + 8 − 3 − 24 + 20 = 2

x3 y 2 − 2x − 3y = 0 |D = 36 (−1) (−2) = 72 > 0 and
Solving the last equations we get the stationary r = 6 (−1) = −6 < 0
points as (0, 0) , (1/2, 1/3)
∴ f x, y has a maximum at D and

= rt − s2 = x4 y2 12 1 − 2x − y
2 fmax = −1 − 8 + 3 + 24 + 20 = 38
1 − x − 3y − 6 − 8y − 9y |B = 36 (1) (−2) = −72 < 0

1 1 |C = 36 (−1) .2 = −72 < 0
At P = , , = rt − s2
2 3
∴ f has neither maximum nor minimum.
1 1 1 1
= · 12 1 − 1 − 1− −1
16 9 3 2
Example 4.38
− (6 − 4 − 3) 2
A rectangular box open at the top is to have a volume
1 of 32 cft. Find the dimensions of the box requiring
= >0
14 least material for its construction.

1 1 2 1 1 1 1
Also, r = 6 . − . − . = − <0 Solution Let x, y, z, be the length, breadth and
2 9 4 9 2 27 9
height of the box, respectively, V its volume and S

1 1 its surface area.
Hence f has a maximum at P = , and
2 3
1 1 1 1 1 S = xy + 2yz + 2zx (1)
fmax = · 1− − = . V = xyz = 32 (2)
8 9 2 3 432
From Eqs. (1) and (2), we have Solving Eq. (2) we get x = 0, y = 0,
64
∴ z = ±1 [from Eq. (1)].
S = xy + y+x
xy
The stationary points are
1 1 P1 = (0, 0, 1), P2 = (0, 0, −1).
S = xy + 64 +
x y At these points
∂S 64 ∂S 64
Sx = = y − 2 , Sy = =x− 2 = rt − s2 = 2.2 − 12 = 3 > 0 and r = 2 > 0
∂x x ∂y y
∴ f is minimum at P1 and P2 .
∂S ∂S
Solving the equations = 0 and = 0, Hence the points on the surface Eq. (1) nearest to
∂x ∂y the origin are P1 = (0, 0, 1) and P2 = (0, 0, −1).
we get x = y = 4
∴ P = (4, 4) is a stationary point. Example 4.40
128 128 Discuss the maxima and minima of
Now, r = Sxx = , s = Sxy = 1, t = Syy = 3 f (x, y) = sin x sin y sin(x + y).
x3 y
128 128 Solution
|P = rt − s2 = 3 · 3 − 1 = 3>0
P 4 4 f (x, y) = sin x sin y sin (x + y) (1)
128
r|P = 3 = 2 > 0 fx = cos xsin ysin(x + y)
4
+ sin xsin ycos(x + y)
Hence S is minimum at P = (4, 4). From Eq. (2),
= sin y sin 2x + y (2)
32 32
we have z = = = 2. fy = sin xcos ysin(x + y)
xy 4×4
∴ The dimensions of the box requiring least mate- + sin xsin ycos(x + y)

rial are 4, 4, 2. = sin x sin 2y + x (3)

Example 4.39 fx = 0 ⇒ sin y sin 2x + y = 0 (4)

Find the points on the surface z = xy + 1 nearest to
2 fy = 0 ⇒ sin x sin 2y + x = 0 (5)
the origin. Solving Eqs. (4) and (5), we get
Solution Let P = (x, y, z) be any point on the x = y = 0 and x = y = π/3
surface
∴ The stationary points are
z 2 = xy + 1 (1) P1 = (0, 0), P2 = (π/3, π/3).

Then OP = x2 + y2 + z 2 Also r = fxx = 2 sin y cos 2x + y

s = fxy = cos xsin 2y + x +sin xcos 2y + x
= x2 + y2 + xy + 1 [by (1)]
= sin 2x + 2y

Let f x, y = OP 2 = x2 + y2 + xy + 1 t = fyy = 2 sin x cos 2y + x
At P1 = (0, 0), we get f = 0
fx = 2x + y; fy = 2y + x;
Since fxx = fxy = fyy = 0, we cannot arrive at any
fx = 0, fy = 0 conclusion. At P2 = (π/3, π/3), we get
,
2x + y = 0 √ √
⇒ 3 √ 3
x + 2y = 0 r = 2· (−1) = − 3 s=−
√2 2
3 √
Also, r = fxx = 2, s = fxy = 1, t = fyy = 2 (2) t = 2· (−1) = − 3
2
3 9 √
|P2 = rt − s2 = 3 − = > 0 and r = − 3 < 0 Wecan easilycheck that at
4 4 a b
Hence f is a maximum at P2 = (π/3, π/3) . P= √ ,√ , = rt − s2 > 0 and r < 0
3 3
π π 2π
and fmax = sin + sin + sin a b
3 3 3 ∴ f is maximum at P = √ , √
√ √ √ √ 3 3
3 3 3 3 3 Also fmax = 64c2
= + + = . 2 2
2 2 2 2 a b 1 a4 b2 1 b4 a2
Example 4.41 · − · · − · ·
3 3 a2 9 3 b2 9 3
Find the volume of the greatest rectangular par-
allelopiped that can be inscribed in the ellipsoid 64a2 b2 c2
=
x2 /a2 + y2 /b2 + z 2 /c2 = 1. 27
8abc
so that Vmax = √ cubic units.
Solution Let the length, breadth and height of 3 3
the parallelopiped be 2x, 2y and 2z, respectively, so
EXERCISE 4.6
that its volume, V = 8xyz.
V 2 = 64x2 y2 z 2 (1)
1. Investigate the maximum and minimum values of
Equation of the ellipsoid the functions:
x2 y2 z2
+ + =1 (2) (i) x3 − 3xy2 − 15x2 − 15y2 + 72x
a2 b2 c2
Ans: Max. at (4, 0) , min. at (0, 0)
Eliminating z from Eqs. (1) and (2), weget
(ii) x2 + y4 − 2x2 + 4xy − 2y2
x2 y2 √ √
f x, y = 64x2 y2 c2 1 − 2 − 2 Ans: No conclusion at (0, 0) ; min. at 2, − 2
a b

x 4 2
y x2 y4 (iii)x3 + 3xy2 − 3x2 − 3y2 + 4
or f x, y = 64c x y − 2 − 2
2 2 2
Ans: Max. at (0, 0) ; and pts. (0, 2) , (1, ±1) are sad-
a b
3 2
dle pts.
4x y 2xy4
fx = 64c 2xy − 2 − 2
2 2
(iv) ax3 y2 − x4 y2 − x3 y3
a b
Ans: Max. at (a/2, a/3)
4 2 3
2x y 4x y
fy = 64c2 2x2 y − 2 − 2 2. Find the maxima and minima of f = x3 +y3 −3axy.
a b
Ans: f is min. at (a, a) if a > 0, f is max. at (a, a) if
Setting fx = 0 and fy = 0, we get a < 0, (0, 0) is a saddle pt.
2x2 y2
+ =1
maxima and minima of f = x + y +
(3) 3. Discuss the 2 2
a2 b2
1 1
x2 2y2 2 + .
and + =1 (4) x y
a2 b2 Ans: f is min. at (1, 1)
Solving
the equations,
we get the stationary point
a b 4. Discuss the maxima and minima of f = x2 +y2 +z 2
P= √ ,√ where x, y, z are connected by xyz = a2 .
3 3
Ans: f is min. at (a, a), fmin. = 3a2
12x2 y2 2y4
Also r = fxx = 64c2 2y2 − − 5. In a plane ABC, find the maximum value of
a2 b2
cos A cos B cos C.
8x y 8xy3
3
s = fxy = 64c2 4xy − 2 − 2 [Hint:f = cos A cos B cos C
a b
4 2 2
= cos A cos B cos [π − (A + B)]
2x 12x y
t = fyy = 64c2 2x2 − 2 − = − cos A cos B cos (A + B).]
a b2
Ans: Max. at A = B = π/3, C = 2π/3, Multiplying Eq. (4.47) by λ and adding Eq. (4.46),
1 we get
max. value =
8
∂u ∂φ ∂u ∂φ
6. Find themaximum and minimum values of f = +λ dx + +λ dy
∂x ∂x ∂y ∂y
xy + a 3
1 1
+
. ∂u ∂φ
x y + +λ dz = 0 (4.48)
Ans: Min. at (a, a), fmin = 3a2 ∂z ∂z
7. The sum of three numbers is constant. Prove that Eq. (4.48) will be satisfied if
their product is a maximum when they are equal.
∂u ∂φ ∂u ∂φ ∂u ∂φ
+λ = 0, +λ = 0, +λ = 0(4.49)
8. Prove that if the perimeter of a triangle is con- ∂x ∂x ∂y ∂y ∂z ∂z
stant then its area is maximum when the triangle is
The stationary values for u will be obtained from
equilateral.
Eqs. (4.44) and (4.49). Note that the above condi-
9. Find the minimum value of f = x2 + y2 + z 2 given tions are necessary but not sufficient for extreme
that ax + by + cz = p. values of the function. So, we have to ultimately
Ans: p2 /(a2 + b2 + c2 ) settle the problem based on geometrical or physical
considerations.
4.13 Constrained Maxima and Example 4.42
Minima Find the point on the plane lx + my + nz = p which
Sometimes we may be required to find extreme values is nearest to the origin.
of a function of several variables which are not all
independent but satisfy some prescribed condition. Solution Let P(x, y, z) be a point on the plane,
Such a problem could be solved by the method of
Lagrange’s undetermined multipliers. φ(x, y, z) = lx + my + nz − p = 0 (1)
and let f (x, y, z) = OP 2 = x2 + y2 + z 2 (2)
4.13.1 Lagrange’s Method of The problem is to minimize Eq. (2) subject to the
Undetermined Multipliers condition Eq. (1)
Suppose we have to find the extreme values of
Let F x, y, z = f x, y, z + λφ x, y, z
u = f (x, y, z) (4.43) (where λ is a constant to be determined)
F(x, y, z) = x2 +y2 +z 2 +λ(lx+my+nz−p) (3)
subject to the condition
For a stationary value of F
φ(x, y, z) = 0 (4.44)
⎫
Fx = 0 ⇒ 2x + λl = 0, ⎬
For u to have stationary values, it is necessary that Fy = 0 ⇒ 2y + λm = 0, (4)
⎭
Fz = 0 ⇒ 2z + λn = 0
ux = 0, uy = 0, uz = 0 (4.45)
Multiplying the above equations by x, y, z, respec-
Now from Eqs. (4.43) and (4.44), we have
tively and adding, we get
∂u ∂u ∂u
du = dx + dy + dz = 0 (4.46) x (2x + λl) + y 2y + λm + z (2z + λn) = 0
∂x ∂y ∂z
∂φ ∂φ ∂φ or 2 x2 + y2 + z 2 + λ lx + my + nz = 0
and dφ = dx + dy + dz = 0 (4.47) or 2f + λp = 0 by Eq. (1) or λ = −2f /p (5)
∂x ∂y ∂z
Substituting the value of λ in Eq. (4), we get (ii) Here (l, m, n) = (3, −4, 12) and p = 26
p2 262
λ lf λ mf f = =
x=− l= , y=− m= , (6) l 2 + m2 + n2 32 + (−4)2 + 122
2 p 2 p
λ nf 262
z =− n= = 2 = 22 = 4
2 p 13
f2 The required point is
∴ f = x 2 + y 2 + z 2 = 2 l 2 + m2 + n2 f 4
p x, y, z = (l, m, n) = (3, −4, 12) ·
p2 p 26
⇒ fmin = 2 . (7)
l + m2 + n2 6 −8 24
= , , .
13 13 13
Note
We observe that the square of the distance from O to Example 4.44
any point P(x, y, z) on the plane is Find the maximum and minimum distances of the
point (3, 4, 12) from the sphere x2 + y2 + z 2 = 1.
f = OP 2 = x2 + y2 + z 2
Solution Let P = (x, y, z) be any point on the
and the perpendicular distance from O to the plane sphere and A = (3, 4, 12) be the given point.
is We have to find maximum and minimum of the
p
OM = √ function
l + m2 + n 2
2
f (x, y, z) = AP 2
Geometrically, f is least when P coincides with
M , the foot of the perpendicular from O to the plane. = (x − 3)2 + (y − 4)2 + (z − 12)2 (1)
p2 subject to the condition

∴ fmin = .
l 2 + m2 + n2 φ(x, y, z) = x2 + y2 + z 2 − 1 = 0 (2)
Let F(x, y, z) = f (x, y, z) + λφ(x, y, z)
Example 4.43
or F(x, y, z) = (x−3)2 +(y−4)2 +(z−12)2 (3)
Corollary Find the point on the plane (i) 2x + +λ(x2 + y2 + z 2 − 1)
3y − z = 5, (ii) 3x − 4y + 5z = 26 which is nearest
to the origin. where λ is a constant to be determined.
For a stationary value of F, we have
Solution ⎫
Fx = 0, ⇒ 2 (x
− 3)
+ λ · 2x = 0, ⎬
(i) Here (l, m, n) = (2, 3, −1) and p = 5 Fy = 0, ⇒ 2 y − 4 + λ · 2y = 0, (4)
⎭
Fz = 0 ⇒ 2 (z − 12) + λ · 2z = 0
p2 52 25
f = = = Multiplying these equations by x, y, z, respectively,
l +m +n
2 2 2
2 + 3 + (−1)
2 2 2
14
and adding, we get
The required point is
2λ x2 + y2 + z 2 + 2 x2 + y2 + z 2

f 5 −2 3x + 4y + 12z = 0
x, y, z = (l, m, n) = (2, 3, −1) ·
p 14 or λ + 1 = 3x + 4y + 12z

10 15 5
= , ,− 3 4 12
14 14 14 From Eq. (4), x = ,y= ,z=
1+λ 1+λ 1+λ
Substituting in Eq. (5) Lagrange’s equations Fx = 0, Fy = 0, Fy = 0

give
3 4 12
(1 + λ) = 3 · +4· + 12 · ⎫
1+λ 1+λ 1+λ 2x + 2aλ1 x + lλ2 = 0, ⎪
⎬
or (1 + λ)2 = 32 + 42 + 122 = 132 2y + 2bλ1 y + mλ2 = 0, (5)
⎪
⎭
or 1 + λ = ±13 2z + 2cλ1 z + nλ2 = 0
λ has two values: Multiplying these equations by x, y, z, respectively,
λ1 = 13 − 1 = 12, λ2 = −13 − 1 = −14 and adding, we get
Corresponding to these values, we get two points
2(x2 + y2 + z 2 ) + 2λ1 (ax2 + by2 + cz 2 )
P1 and P2 giving extreme values of f :
+λ2 lx + my + nz = 0

3 4 12 3 4 12
P1 = , , and P2 = − , − , −
13 13 13 13 13 13
⇒ 2f + 2λ1 .1 + λ2 (0) = 0
⇒ λ1 = −f [ by Eqs. (1), (2), and (3)]

3 2 4 2 12 2
Also, AP12= 3− + 4− + 12−
13 13 13 ∴ 2x − 2axf + lλ2 = 0,
2
12
= 32 +42 +122 . = 122 2y − 2byf + mλ2 = 0,
13
2z − 2czf + nλ2 = 0
3 2 4 2 12 2
AP22 = 3 + + 4+ + 12+
13 13 13
2 lλ2
14 or x = ,
= 32 +42 +122 = 142 2 af − 1
13
mλ2
∴ P1 is at minimum distance = 12 from A y= ,
2 bf − 1
and P2 is at maximum distance = 14 from A. nλ2
z=
Example 4.45 2 cf − 1
Find the maxima and minima of x2 + y2 + z 2 subject
Substituting in Eq. (3), we have
to the conditions that
l2 m2 n2
ax2 + by2 + cz 2 = 1 and lx + my + nz = 0. + + =0 (6)
af − 1 bf − 1 cf − 1
Solution which gives the maximum and minimum values of f .

Let f x, y, z = x +y +z
2 2 2
(1)

φ1 x, y, z = ax + by + cz − 1 = 0
2 2 2
(2) EXERCISE 4.7

φ2 x, y, z = lx + my + nz = 0 (3)
1. If xyz = 8 find x, y and z so that
Let F x, y, z = f x, y, z + λ1 φ1 x, y, z

+λ2 φ2 x, y, z 5xyz
u=
= x +y +z
2 2 2 x + 2y + 4z

+λ1 ax2 + by+cz
2 2
−1 is a maximum.

+λ2 lx+my+nz (4) Ans: (x, y, z) = (4, 2, 1)
2. Find the maximum value of x2 + y2 + z 2 subject 5. Prove that the maximum value of xyz subject to
to the condition, the condition x + y + z = a is a3 /27.
(a) xyz = a3 , 6. Find the stationary values of x2 + y2 + z 2 subject
Ans: 3a2 x2 y2 z2
to + + = 2 and 3x + 2y + z = 0.
(b) ax + by + cz = p. 1 2 3
Ans: p2 / a2 + b2 + c2 Ans: 28/5, 3
3. If u = a3 x2 + b3 y2 + c3 z 2 where 7. Prove that the stationary values of u = x2 /a4 +

1 1 1 y2 /b4 + z 2 /c4 where lx + my + nz = 0 and
+ + = 1 then show that the stationary x2 /a2 + y2 /b2 + z 2 /c2 = 1 are the roots of the
x y z
value$of uis given by equation
l 2 a4/ 1 − a2 u + m2 b4 / 1 − b2 u +
$ $
x= a /a, y = a /b, z = a /c. n c / 1 − c2 u = 0.
2 4
4. Find the maximum and minimum distance from 8. Prove that the maximum value of mx yn z p
the origin to the curve 5x2 + 6xy + 5y2 − 8 = 0. under the condition x + y + z = a is
Ans: 4, 1 mm nn pp am+n+p / (m + n + p)m+n+p .
i
Radius of Curvature 5
5.1 Introduction is the average bending or average curvature of the
Let C be a smooth plane curve at each point on which arc.
a tangent can be drawn as in Fig. 5.1. Suppose that Curvature of a curve C at a point P, denoted by κ
between P and Q on C, the curve bends continuously is given by
in one direction.
κ = lim |κav |
Q→P
Let AP= s and PQ= δs. Suppose PL and QM are
tangents drawn to the curve at P and Q intersecting δψ dψ
= lim =
at R and cutting the X -axis at L and M , respectively. δs→0 δs ds
5.1.1 Angle of Contingence 5.2 Curvature of a Circle

The angle QRT is called the angle of contingence of In the case of a circle (Fig. 5.2), its curvature is the
the arc PQ. Clearly it is the angle through which the same at every point. It is measured by the reciprocal
tangent line turns as it touches each point between P of the radius. If r is the radius and O the centre then
and Q.
PQ
RTQ = POQ =
r
5.1.2 Measure of Curvature the angle being measured in circular measure
(radians)
The fraction
Angle of contingence 1
∴ =
κav = Angle of contingence
Length of arc
Length of arc r
Figure 5.1 Curvature of a smooth plane curve Figure 5.2 Curvature of a circle
5.2.3 Intrinsic Equation

The equation of a curve may be written in various
forms. A relation between s and ψ is called the intrin-
sic equation of the curve: s = f (ψ). The formula for
ρ in this case is
ds
ρ= [From Eq. (5.1)]
dψ
For example, the intrinsic equation of a circle of
radius ‘a’ is s = aψ.
Figure 5.3 Circle of curvature through P, Q, R on
ds 1 1
a curve C ∴ ρ= = a and the curvature κ = = .
dψ ρ a
5.2.1 Circle of Curvature 5.2.4 Transformations

dx dy
A unique circle can be drawn through three From Fig. 5.4, we have cos ψ = , sin ψ =
ds ds
neighbouring points P, Q, R on any curve C (Fig.
5.3). As the points move closer together, PQ and QR Differentiating w.r.t ‘s’
ultimately become tangents both to the curve and to dψ d 2x dψ d 2y
the circle. So, the curvature of a curve at a given point − sin ψ = 2, cos ψ = 2
ds ds ds ds
is defined with respect to the circle described at the
point. We call the circle, the circle of curvature at d 2x d 2y
1 −
that point. Its radius and centre are called the radius ds2 = ds2
= (5.2)
and centre of curvature, respectively. A chord of this ρ dy dx
circle drawn at the point of contact in any direction is ds ds
referred to as the chord of curvature in that direction.
Squaring and adding,
2 2 2
5.2.2 Formula for Radius of 1 d x d y
Curvature = + (5.3)
ρ 2 ds 2 ds2
Let PQ and QR be equal chords and the correspond-

since δs2 = (Chord PQ)2 = (δx)2 + (δy)2
ing arcs PQ = QR = δs and the angle of contingence 2 2
dx dy
RQZ = δψ. or 1= +
ds ds
Now radius of circumcircle of
PR
PQR =
1
2 sin P QR
PR 2δs
∴ ρ = lim = lim
1
2 sin P QR 2 sin δψ
δs δψ ds
= lim · lim = (5.1)
δψ sin δψ dψ
Also, the perpendicular bisectors of chords PQ
and QR intersect at the circumcentre of PQR. Thus,
in the limit, the centre of curvature at any point on a
curve C is the point of intersection of the normal at
that point with the normal at a neighbouring point. Figure 5.4 Elemental right triangle
Radius of Curvature 5-3
Formulas (5.2) and (5.3) are only suitable for the Solution Equation of the parabola is
case where x and y are known functions of s.
s = a log(tan ψ + sec ψ) + a tan ψ sec ψ (1)
Example 5.1
Find the radius of curvature at any point of the Differentiating Eq. (1) w.r.t ψ, we have
catenary (chainette) s = c tan ψ.
ds
Solution Equation of the curve is ρ=
s = c tan ψ (1) dψ
a(sec2 ψ + sec ψ tan ψ)
Differentiating Eq. (1) w.r.t ψ, we have =
tan ψ + sec ψ
ds +a(sec3 ψ + tan2 ψ sec ψ)
ρ= = c sec2 ψ
dψ = a sec ψ + a sec ψ(sec2 ψ + sec2 ψ − 1)

s2 c 2 + s2 = 2a sec3 ψ.
= c 1+ 2 = .
c c
EXERCISE 5.1
Example 5.2
Find√
the radius of curvature at any point of the cycloid 1. Find the radius of curvature at any point of the
s = 8ay having the intrinsic equation s = 4a sin ψ. tractrix s = c log sec ψ.
Ans: ρ = c tan ψ
Solution Equation of the curve is
s = 4a sin ψ (1) 2. Prove that for the curve s = m(sec3 ψ − 1),
the radius of curvature is ρ = 3m tan ψ sec3 ψ.
Differentiating Eq. (1) w.r.t ψ, we have Hence show that 3my1 y2 = 1 and this differential
ds
ρ= = 4a cos ψ equation is satisfied by the semicubical parabola
dψ 27my2 = 8x3 .
*
s2
= 4a 1 − 3. Prove that for the curve
16a2
*
y π ψ sin ψ
= 16a2 − s2 = 4a 1 − . s = a log cot − +a 2 ,
2a 4 2 cos ψ
Example 5.3
the radius of curvature is ρ = 2a sec3 ψ.
Find the radius of curvature at any point of the
ψ
cardioid s = 4a sin . 4. Determine the radius of curvature of the curve
3 s s
x = a cos , y = a sin .
Solution Equation of the curve is a a
ψ [Hint: Use Equations (5.2) and (5.3)]
s = 4a sin (1) Ans: ρ = a
3
Differentiating Eq. (1) w.r.t ψ, we have
5.3 Radius of Curvature—
ds 4a ψ
ρ= = cos . Cartesian Formula
dψ 3 3
[Equation of the curve is of the form y = f (x)]
Example 5.4 We know that the slope of the tangent at any point P
Find the radius of curvature at any point of the on a curve C is
parabola
dy
s = a log(tan ψ + sec ψ) + a tan ψ sec ψ. = tan ψ
dx
3/2
Differentiating w.r.t ‘x’ 1 + y12 [1 + (−1)2 ]3/2
∴ ρ= = √
y2 2
d 2y dψ dψ ds √
2
= sec2 ψ = sec2 ψ . 2 2
dx dx ds dx = √ = 2 = radius of the circle.
dψ dx 2
= sec3 ψ ∵ = cos ψ
ds ds Example 5.6
3a 3a
ds sec ψ 3
Find the radius of curvature at P = , on the
or ρ= = (5.4) 2 2
dψ d2y
2 dx curve x3 + y3 = 3axy.
(1 + tan ψ) 2 3/2
(1 +y12 )3/2
= = (5.5) Solution The equation of the curve is
d2y y2
dx2
x3 + y3 = 3axy (1)
dy d 2y Differentiating Eq. (1) w.r.t ‘x’, we have
where y1 = , y2 = .
dx dx2 dy
3x2 + 3y2 y1 = 3a(y + xy1 ) where y1 =
Note
dx
If the equation is of the form x = φ(y) then ay − x2
⇒ y1 = (2)
y2 − ax

2 3/2
1+ dx Differentiating Eq. (2) w.r.t ‘x’, we have
dy
ρ= . (5.6) 2
d2x y −ax ay1 −2x − ay−x2 2yy1 −a
dy2 y2 = 2
y2 −ax
Example 5.5 √ √ (3)
Find the radius of curvature at P = ( 2, 2) on the
curve x2 + y2 = 4. 3a 9a2 6a2 9a2
a· − −
Solution The equation of the given curve, which y1 |p = 2 4 = 4 4 = −1
is a circle, is 9a2 3a 9a2 − 6a2
−a· 4 4
4 2

x2 + y2 = 4 (1) 3a2 2 2
−a − 6a2 − 6a4 − 9a4 (−3a − a)
4
y2 |p = 4 9a
Differentiating Eq. (1) w.r.t ‘x’, we have 16
dy 32
2x + 2y =0 (2) =
dx 3a
√
x 2 The radius of curvature at P
⇒ y1 |p = − = − √ = −1 3/2 3/2
y p 2 1 + y12 1 + (−1)2 3a
ρ= = = √ .
Differentiating Eq. (2) again w.r.t ‘x’, we get |y2 | 32
3a 8 2
2 Example 5.7
dy d 2y Find the radius of curvature at (3, 3) on the curve
2+2 + 2y 2 = 0 (3)
dx dx x3 + xy2 − 6y2 = 0.
1 + y12
⇒ y2 |p = − Solution The equation of the curve, in implicit
y p
form, is
1 + (−1)2 √
=− √ =− 2
2 φ x, y = x3 + xy2 − 6y2 = 0 (1)
Differentiating Eq. (1) w.r.t ‘x’, we have 5.4 Radius of Curvature—

Parametric Form
3x + y + 2xyy1 − 12yy1 = 0
2 2
(2)
At P = (3, 3) [Equations of the curve are of the form: x = x(t),
y = y(t)]
3 · 32 + 32 + 2 · 3 · 3y1 p − 12 · 3 · y1 p =0
Let x = x(t), y = y(t) be the parametric equations
⇒ y1 =2 (3)
p
of the curve y = f (x).
Differentiating Eq. (2) w.r.t ‘x’, we have
dy dy/dt y
6x + 2yy1 + 2yy1 + 2xy12 + 2xyy2 y1 = = =
dx dx/dt x
−12y12 − 12yy2 = 0
Putting (x, y) = (3, 3) and y1 p =2 where denotes differentiation w.r.t parameter ‘t’.
we get 18 + 12 + 12 + 24 + 18y2 p
d 2y d
−48 − 36y2 p = 0 ⇒ y2 p = 1 y2 = 2
= y /x
dx dx
The radius of curvature at P = (3, 3) on Eq. (1) x y − x y 1 x y − x y
= =
3/2 3/2 (x ) 2
x (x )3
1 + y12 1 + 22 √ 3
ρ= = = 53/2 = 5 5. 1 + y12 2 2 32 (x )3
|y2 | 1 ρ= = 1 + y /x ·
y2 x y − x y
Example 5.8 2 2 3
2
Show that the radius of curvature at any point of the x1 + y1 (x )3
= ·
catenary y = c cosh(x/c) is equal to the length of the (x )3 x y − x y
portion of the normal intercepted between the curve 2 2 3
and the x-axis. (x )2 + y1

=
x y − x y
Solution Equation of the catenary
x
y = c cosh (1)
c 2 32
Differentiating Eq. (1) w.r.t ‘x’, we have (x )2 + y
∴ Radius of curvature, ρ = (5.6)
dy x 1 x [x y − x y ]
y1 = = c sinh · ⇒ y1 p = sinh (2)
dx c c c 1 [x y − x y ]
Curvature, =
d 2y x 1 1 x ρ 2 32
y2 = 2 = cosh · ⇒ y2 p = cosh (3) (x )2 + y
dx c c c c
∴ Radius of curvature
3/2 3/2
1 + y12 1 + sinh2 xc Example 5.9
ρ= = 1 Show that the radius of curvature at any point of
y2 cosh xc
c the cycloid x = a (t + sin t), y = a (1 − cos t) is
cosh3 x
x y2 4a cos(t/2).
= 1
e
x
= c cosh2 = (4)
c
cosh c
c c
Solution The parametric equations of the cycloid
Length of the normal at P(x, y) is are
+
x y2
PN = y 1 + y12 = y cosh = [by Eq. (2)]
c c x = a (t + sin t) , y = a (1 − cos t) (1)
∴ Radius of curvature at any point P on the
catenary = Length of the normal. Differentiating w.r.t ‘t’ we get
dx y2 = 4ax (1)
x = = a (1 + cos t) ,
dt
dy Differentiating Eq. (1) w.r.t ‘x’, we get
y = = a (0 + sin t) (2)
dt
2
d 2x 2yy1 = 4a ⇒ y1 = y (2)
x = 2 = a (0 − sin t) , a
dt
2
Differentiating again
d y
y = 2 = a cos t (3) 2y12 + 2yy2 = 0
dt
2 3/2 y12 4a2
(x )2 + y or y2 = − =− 3 (3)
y y
ρ=
x y − x y
3/2 The radius of curvature is
3/2
a3 (1 + cos t)2 + (sin t)2 1 + y12
= (4) ρ=±
a2 (1 + cos t) cos t + sin2 t y2
a [2 (1 + cos t)]3/2 4a2
3/2
2 3/2
= 1+ y + 4a2 1
(1 + cos t) =±
y2
= ·
√ 2 4a2 y3
= a · 2 2 (1 + cos t)1/2 − 4ay3 y3
1/2 3/2 √
√ 2 t t 4ax + 4a2 4a · 2 a (x + a)3/2
= a·2 2· 2 cos = 4a cos = =
2 2 4a2 4a2
2
Example 5.10 = √ (x + a)3/2 .
Find the radius of curvature at any point ‘t’ of the a
curve
Example 5.12
x = a (cos t + t sin t) , y = a (sin t − t cos t). (1)
Find the radius of curvature at any point of the curve
Solution Differentiating Eq. (1) w.r.t ‘t’, we have y = c log sec (x/c).
dx Solution The equation of the curve is

x = = a (− sin t + sin t + t cos t) = at cos t
dt y = c log sec (x/c) (1)
dy
y = = a (cos t − cos t + t sin t) = at sin t
dt Differentiating Eq. (1) w.r.t ‘x’, we get
dy y at sin t x x x 1 x
= = = tan t y1 = c cos · sec · tan · = tan (2)
dx x at cos t c c c c c
d 2y dt sec2 t 1 2 x 1
= sec2 t = = sec3 t y2 = sec · (3)
dx2 dx at cos t at c c
3/2 3/2
1 + y12 1 + (tan t)2 The radius of curvature is
ρ= = 1 3/2 3/2
y2 at
sec3 t 1 + y12 1 + tan2 xc
sec3 t ρ=± = 1
= = at y2 c
sec2 xc
1
sec3 t x
at sec3 x
= 1
c
2 x
= c sec . (4)
Example 5.11 c
sec c
c
Find the radius of curvature at any point of the
parabola y2 = 4ax. Example 5.13
Find the radius of curvature at (–2, 0) on the curve
Solution The equation of the curve is y2 = x3 + 8.
Solution Here we take y as the independent 3/2 3/2

1 + y12 1 + t12
variable and use the formula ρ= ± =±
y2 − 2at1 3
2 2 3/2
1 + dy dx
1 + t2 2at 3 3/2
ρ= = 3
· = 2a 1 + t 2 (5)
2 d x t 1 2
dy2 S = (a, 0) , P = at , 2at
2
Differentiating the given equation w.r.t to ‘y’, we (SP)2 = at − a2 + (2at − 0)2
get 2
= a2 t 2 − 1 + 4t 2 = a2 t 2 + 1
dx
3x2 = 2y SP = a t2 + 1 (6)
dy

dx 2y From Eqs. (4) and (5)
= =0 3/2
dy p 3x2 (−2,0)
2 3/2
2 a 1 + t2
ρ = 2a 1 + t = √ (7)
Differentiating again w.r.t to ‘y’, we get a
2
2
2 = √ (SP)3 .
2d x dx a
3x + 6x =2
dy2 dy
2
2 − 6x dx
d 2x dy 2 1 5.5 Radius of Curvature—Polar
= = = Form
dy2 p 3x2 3 (−2)2
6
p [Equation of the curve is of the form r = f (θ)]

2 3/2
Let O be the pole and OX be the initial line as shown
1+ dx
dy (1 + 0)3/2 in Fig. 5.5. Let p be the length of the perpendicular
ρ= = 1
= 6. ON on the tangent PT at P on C.
d2x
dy2 6
Let r = f (θ) be the equation of the curve C. Sup-
Example 5.14 pose that the tangent PT at P(r, θ) on the curve C
Find the radius of curvature at any point P(at 2 , 2at) make an angle ψ with the initial line OX and the
3/2
on the parabola y2 = 4ax. Show that it is 2 (SP) √
a
angle between the radius vector OP and the tangent
PT be φ.
where S is the focus of the parabola. Then
ψ =θ+φ
Solution The equations of the parabola are
x = at 2 , y = 2at (1)
Differentiating Eq. (1) w.r.t to ‘t’, we get

dx dy
x = = 2at, y = = 2a
dt dt
dy dy/dt 2a 1
y1 = = = = (2)
dt dx/dt 2at t

d 2y d 1 dt
y2 = 2 = · (3)
dx dt t dx
1 t 1
=− 2· =− 3 (4)
t 2at 2at Figure 5.5 Tangent at P(r, θ) on a polar curve
3/2
1 dψ dθ dφ dθ dφ θ 2 θ θ 2
= = + = 1+ (5.7) =a 3 2
+ 2 sin cos
ρ ds ds ds ds dθ 2 sin
2 2 2
Since 2· 32 3/2
rdθ r r θ 2 θ 2 θ
tan φ = = dr = (5.8) = a · 2 sin
3
× sin + cos
dr r1 2 2 2
dθ
θ
= 8a3 sin3 (3)
dr 2
where r1 =
dθ 2
Differentiating Eq. (5.8) w.r.t θ, we get dr d 2r
r 2 + 2r1 − rr2 = r 2 + 2 −r
dr 2 d2r dθ dθ 2
dφ −r dθ 2 r12 − rr2
sec φ
2
=
dθ
dr 2 = = a2 (1 − cos θ)2 + 2a2 sin2 θ
dθ r12
dθ −a2 cos θ (1 − cos θ)

dφ r 2 − rr2 1 r12 − rr2 = a 1 + cos2 θ − 2 cos θ
2
= 1 2 · = (5.9)
dθ r1 2
r 2 + r12 +2 sin2 θ − cos θ + cos2 θ
1+ r
r1
θ
dr d 2r = 3a2 (1 − cos θ) = 6a2 sin2
where r1 = , r2 = 2 2
dθ
+
dθ 2 3/2
ds r + r12 8a3 sin3 2θ
Also, = r 2 + r12 (5.10) ρ= 2 =
dθ r + 2r1 − rr2 6a2 sin2 2θ
Substituting in Eq. (5.7) we get 4 θ 2√
= a sin = 2ar·
1 1 r12 − rr2 3 2 3
= 1/2 1 +
ρ r 2 + r12 r 2 + r12 Example 5.16
Show that the radius of curvature of the curve
r 2 + 2r12 − rr2
= 3/2 r n = an cos nθ is an r −n+1 / (n + 1) for n = −1.
r 2 + r12
2 3/2 Solution The equation of the given curve is
r + r12
ρ= 2 (5.11) r n = an cos nθ (1)
r + 2r12 − rr2
By logarithmic differentiation w.r.t ‘θ’ we get
Example 5.15
Find the radius of curvature at any point (r, θ) of the n dr n sin nθ dr
=− ⇒ r1 = = −r tan nθ (2)
cardioid r = a(1 − cos θ). r dθ cos nθ dθ
Solution Equation of the curve is Differentiating again w.r.t θ

r = a(1 − cos θ) (1) d 2r dr
r2 = 2
= − tan nθ − nr sec2 nθ
Differentiating w.r.t ‘θ’ dθ dθ
dr d 2r = r tan2 nθ − nr sec2 nθ (3)
r1 = = a sin θ, r2 = 2 = a cos θ (2) 2
2 3/2
dθ dθ r + r1
∴ ρ=
r + 2r1 − rr2
2
2 3/2
2 3/2 r + r 2 tan2 nθ
dr =
(r 2 + r12 )3/2 = r 2 + r 2 + 2r 2 tan2 nθ − r 2 tan2 nθ + nr 2 sec2 nθ
dθ
3/2 r 3 sec3 nθ r
= a2 (1 − cos θ)2 + a2 sin2 θ = =
(n + 1) r 2 sec2 nθ (n + 1) cos nθ
ran an r −n+1 Example 5.17

= = . (4)
(n + 1) r n (n + 1) Show that the p–r equation of the cardioid r = a(1 −
Particular cases cos θ) is p2 = r 3 /2a.
n = −1, a = r cos θ (Straight line) Solution Polar equation of the carioid

n = −2, r 2 cos 2θ = a2 (Rectangular hyperbola) r = a(1 − cos θ) (1)
n = 1, r = a cos θ (Circle with diameter a)
Differentiating Eq. (1) w.r.t θ, we get
n = 2, r 2 = a2 cos 2θ (Lemniscate of Bernoulli)
n = − 12 , r(1 + cos θ) = 2a (Parabola) dr
= a sin θ (2)
n = 12 , r = a2 (1 + cos θ) (Cardioid) dθ

n = 13 , r 1/3 = a1/3 cos 3θ (Cardioid) 1 1 1 dr 2 1 1
∴ 2
= 2
+ 4
= 2 + 4 a2 sin2 θ
p r r dθ r r
5.6 Pedal Equation or p—r r 2 + a2 sin2 θ
Equation of a Curve =
r4
Let OX be the initial line, O being the pole. Let P(r, θ) a (1 − cos θ)2 + a2 sin2 θ
2
be any point on the curve C as shown in Fig. 5.6. Let =

r4
p be the length of the perpendicular ON drawn from a 2
O on the tangent PN touching the curve at P. Then = 1 − 2 cos θ + cos2 θ + sin2 θ
r4
ON = p = r sin O1
PN = r sin φ 2a2 2ar 2a
= 4
(1 − cos θ) = 4 = 3 (3)
r r r
r3
∴ p2 = . (4)
2a
Example 5.18
Given that the polar equation of the parabola with
2a
respect to the focus as the pole is = 1 − cos θ,
r
show that its p–r equation is p2 = ar.
Solution Polar equation of the parabola is

2a
= 1 − cos θ (1)
r
Figure 5.6 Pedal equation curve Differentiating w.r.t ‘θ’
2a dr
⇒
1
=
1 1
= 2 cosec2 φ − = sin θ
r 2 dθ
p2 r 2 sin2 φ r
1 1 1 dr 2 1 sin θ 2
1 1 dr 2 Now, = + = +
= 2 1 + cot φ = 2 1 +
2 p2 r2 r 4 dθ r2 −2a
r r rdθ
(1 − cos θ)2 sin2 θ
2 = +
1 1 1 dr 4a 2 4a2
⇒ 2 = 2+ 4
p r r dθ 2 − 2 cos θ 2(1 − cos θ)
= =
1 dr 1 du 4a2 4a2
Put r = , =− 2 1 2a 1
u dθ u dθ = 2· =
2 2a r ar
1 1 du 2 du
2
= u 2
+ u 4
− 2
= u 2
+ . ⇒ p2 = ar.
p u dθ dθ
5.6.1 Radius of Curvature of a dr

2r = 2a2 cos 2θ (2)
Pedal Curve dθ
2
From Fig. 5.6 we see that [Equation of the curve is 1 1 1 dr
Now, = 2+ 4
of the form p = f (r)] p2 r r dθ
ψ = θ+φ (5.12) 1 1 a4 cos2 2θ
= 2+ 4
r r r2
Differentiating w.r.t ‘s’
a sin 2θ + a4 cos2 2θ
4 2
a4
1 dψ dθ dφ = =
= = + (5.13) r6 r6
ρ ds ds ds
r3
Now p = r sin φ. ⇒p= 2 (3)
a
Differentiating w.r.t ‘r’
Differentiating Eq. (3) w.r.t ‘r’, we get
dp dφ dθ dr dφ
= sin φ + r cos φ =r +r · dp 3r 2 dr a2 a2
dr dr ds ds dr = 2 ⇒ ρ=r =r· 2 = .
dr a dp 3r 3r
dθ dr
∵ sin φ = r , cos φ =
ds ds EXERCISE 5.2

dp dθ dφ r 1. Find the p − r equation of the following curves
=r + = [by Eq.(5.13)]
dr ds ds p whose polar equations are:
dr
∴ ρ=r . (5.14) (a) r sin θ + a = 0.
dp
Ans: p + a = 0
Example 5.19 a
Find the radius of curvature of the cardioid (b) r = (1 − cos θ).
2
r = a(1 − cos θ). Ans: ap2 = r 3
Solution The p–r equation of the cardioid whose (c) r = a sin θ.

polar equation is r = a(1 − cos θ) is Ans: ap = r 2
2p2 a = r 3 (1) (d) r 2 sin 2θ + a2 = 0.

Ans: pr + a2 = 0
Differentiating Eq. (1) w.r.t p, we get
(e) r m = am sin mθ.
dr Ans: pam = r m+1
4pa = 3r 2
dp
1/2 (f) r = aeθ cot α .
dr 4pa 4a r3 2
ρ=r = = = (2r)1/2 . Ans: p = r sin α
dp 3r 3r 2a 3
2. Find the radius of curvature at any point on
Example 5.20 the curve
Find the radius of curvature of the curve
r 2 = a2 sin 2θ. (a) x2/3 + y2/3 = a2/3
Ans: (axy)1/3 or 3a sin θ cos θ
Solution The p–r equation of the curve whose
polar equation is (b) ay2√= x3
Ans: x(4a + 9x)3/2 /6a
r 2 = a2 sin 2θ (1)
(c) x = a(t − sin t); y = a(1 − cos t)
Differentiating Eq. (1) w.r.t ‘θ’, we get Ans: 4a sin t/2
(d) r 2 cos 2θ = a2 dy
⇒ y1 (0) = =0 (5.16)
Ans: r 3 /a2 dx (0,0)
3. Find the radius of curvature of each of the x2 2x 1 1

Now lim = lim = lim = (5.17)
following curves at the specified point on it. x→0
y→0
2y 2y1 y2 y2 (0)
√ √ √ (Applying L’Hospital’s rule).
(a) x + y = a at (a/4, a/4)
√
Ans: a/ 2 L’ Hospital’s1 Rule is a rule to evaluate indeterminate
0 ∞
(b) xy2 = 16(x + 4) at (-4, 0) forms of the type and .
Ans: 54/7 0 ∞
Let f and g be differential functions with g (x) =
0 is an open interval containing c (except possi-
3a 3a
(c) x3 + y3 = 3axy at , bly at c itself). Suppose limx→c g(x) f (x)
produces an
√ 2 2
Ans: 3a/8 2
0 ∞
indeterminate form or and that
0 ∞
(d) x = a cos θ(1 + sin θ); y = a sin θ(1 + cos θ)
f (x)
at θ = −π/4 lim =L
Ans: a x→c g (x)
dθ where L is a finite number, ∞ or −∞. Then
4. Prove that in any curve p = r 2 and f (x)
ds lim = L.
dr x→c g(x)
r 2 − p2 = r . ∴ Radius of curvature at the origin
ds
5. Find the radius of curvature at (r, θ) on r n = 3/2
1 + y12 (0) (1 + 0)3/2
an sin nθ. = ρ|(0,0) = =
y2 (0) y2 (0)
6. Show that in the cardioid r = a(1 + cos θ), ρ2 /r 1 x2
is constant. = = lim [by Eq.(5.17)]
y2 (0) x→0 2y
y→0
7. Show that the curvatures of r = aθ and rθ = a
at their point of intersection are in the ratio 3:1. Similarly, if the y-axis is a tangent to the curve C
at (0, 0) we have
8. Show that the radius of curvature of r = a sin nθ 2
y
at the pole is na/2. ρ|(0,0) = lim . (5.18)
x→0
y→0
2x
5.7 Radius of Curvature at the (ii) Expansion method

Origin
If neither the x-axis nor the y-axis is a tangent to the
(i) Newton’s method
curve C at the origin, then by Maclaurin’s theorem,
Let the given curve C pass through the origin. Then x x2
y(x) = y(0) + y1 (0) + y2 (0) + · · ·
the radius of curvature at the origin is given by 1! 2!
2 y(0) = 0 [∵ C passes through (0, 0)]
x
ρ|(0,0) = lim (5.15)
x→0
y→0
2y 1 L’ Hospital (alias L’ Hospital) Guillaume Francois Antoine
Marquis de (1661–1704), French analyst and geometer, is known
if the x-axis is a tangent to C. for the rule that bears his name. The rule was actually discovered
by Johann Bernoulli who was his teacher. But L’ Hospital did cite
his sources in his book, which was the first book on differential cal-
Proof: The x-axis is a tangent to the curve C at culus, for he paid for his results, and for keeping the arrangements
(0, 0) confidential.
Radius of curvature at the origin Substituting in Eq. (1)

x2
3/2 3/2 y = px + q + ··· (2)
1 + y12 (0) 1 + p2 2
ρ|(0,0) = = (5.19)
y2 (0) q where p = y1 (0), q = y2 (0) we have
2
Note x2
1. The tangents at the origin are obtained by equating a px + q + · · · − x = x3 2
(3)
2
to zero the lowest degree terms in the equation of the
curve. Equating the coefficients of x2 , x3 on both sides
2. Radius of curvature at the pole (origin) ap − a = 0 ⇒ p = ±1, apq = 1
2
1
x2 ⇒ q=± (4)
ρ|pole = lim a
x→0 2y 3/2
y→0
1 + p2 (1 + 1)3/2 √
r 2 cos2 θ r 1 ∴ ρ= = = ±2 2a.
= lim = lim · q ± 1a
r→0
θ→0
2r sin θ 2 θ
(∵ cos θ = 1, sin θ = θ, approximately) EXERCISE 5.3
1. Find the radius of curvature at the origin to the
1 dr
= lim . curves:
2 dθ
(a) x3 − 2x2 y + 3xy − 4y3
Example 5.21
+ 5x2 − 6xy + 7y2 − 8y = 0.
Find the radius of curvature at the origin for the curve
Ans: ρ = 4/5, y = 0, the x-axis is
2x4 + 3y4 + 4x2 y + xy − y2 + 2x = 0 (1) a tangent at (0, 0)
(b) x4 − y4 + x3 − y3 + x2 − y2 + y = 0.
Solution Equating to zero the lowest degree term Ans: ρ = 1/2, y = 0, the x-axis is a tangent at
we get the equation of the tangent at (0, 0) as (0, 0)
x=0 (the y-axis) (2) (c) 2(x4 + y4 ) + 4x2 y + xy − y2 + 2x = 0.
Ans: ρ = 1, x = 0, the y-axis is a tangent at
Dividing Eq. (1) by x, we get (0, 0)
(d) y − x = x2 + 2xy + y2 .
y2 y2 Ans: y = x is a tangent at (0, 0); expansion
2x + 3y ·
3
+ 4xy + y −
2
+2=0 (3)
x x method, ρ = 2√1 2
y2 a+x
At the origin lim = 2ρ, x = 0, y = 0 so that (e) x3 = x2 · .
x a−x
Eq. (3) gives Ans: y = x are tangents
2 2
√ at (0, 0); expansion
method, ρ = ± 2a
−2ρ + 2 = 0 ⇒ ρ = 1. (f) x3 + y3 = 3axy.
Ans: Both the axes are tangents at (0, 0)
Example 5.22
ρ = 3a2
Find the radius of curvature at the origin to the curve

x3 = a y2 − x2 (1) 2. Find the radius
of curvature
at (2a, 0) to the curve
2a − x
y = 4a
2 2
·
Solution Equating the lowest degree terms to x
zero we find that y = ±x are the tangents to the [Hint: Shift the origin to (2a, 0).]
curve (1) at (0, 0). Ans: ρ = a
3. Find the curvature κ at the origin for the curve Hence

5x3 + 7y3 + 4x2 y + xy2 + 2x2 + 3xy + y2 + 4x = 0. ⎫
y1 1 + y12 ⎪ ⎪
1 x̄ = x − ⎪
Ans: x = 0 is a tangent at (0, 0); κ = − y2 ⎪
⎬
2
⎪ (5.24)
1 + y12 ⎪
⎪
5.8 Centre of Curvature ȳ = y + ⎪
⎭
y2
We now find the cartesian coordinates of the centre
of curvature. Equation of the circle of curvature
Consider Fig. 5.7. Let Q(x̄, ȳ) be the centre of cur-
(x − x̄)2 + (y − ȳ)2 = ρ2 (5.25)
vature corresponding to the point P(x, y) on the curve
C. Let the tangent PT at P make an angle ψ with OX , Evolute The locus of the centre of curvature of all
and points on a given curve C is called the evolute of the
PM , QL be perpendiculars on OX curve C.
The given curve C itself is called the involute of
PR be perpendicular on QL.
the evolute.
Example 5.23
Also PQR = PTX = ψ.
Find the centre of curvature of the parabola y2 = 4ax
Then and hence find its evolute.
x̄ = OL = OM − LM = OM − PR Solution The parametric equations of the

= OM − PQ sin ψ = x − ρ sin ψ (5.20) parabola
ȳ = QL = LR + QR = PM + QR y2 = 4ax (1)
= PM + PQ cos ψ = y + ρ cos ψ (5.21) are x = at 2 , y = 2at (2)
dy
But 2a 1
y1 = dt
dx
= = ,
y1 2at t
tan ψ = y1 , sin ψ = ; dt
1 + y12 1 dt 1
y2 = − 2
=− 3 (3)
1 t dx 2at
cos ψ = (5.22)
1 + y12 The coordinates of the centre of curvature are
⎫
y1 1 + y12
3
(1 + y12 ) 2 ⎪
⎪
Also ρ = (5.23) x̄ = x − ⎪
⎪
y2 y2 ⎪
⎪
⎪
2⎪
= at − t 1 + t 2 −2at = 2a + 3at ⎬
2 1 1 3
(4)
1 + y12 ⎪
⎪
ȳ = y + ⎪
⎪
y2 ⎪
⎪
⎪
⎪
= 2at + 1 + t 2 −2at = −2at
1 3 3 ⎭
Eliminating t between Eqs. (4) and dropping bars,

we get the equation of the evolute as
27ay2 = 4(x − 2a)3 (5)
Example 5.24
x2 y2
Figure 5.7 Centre of curvature of a curve C Find the centre of curvature of the ellipse+ =1
a2 b2
and hence find the equation of the evolute.
Solution The parametric equations of the ellipse Solution
x2 y2 dy/dt a sin t
+ =1 (1) y1 = =
a2 b2 dx/dt a(1 − cos t)
are x = a cos θ, y = b sin θ (2) 2a sin 2t cos 2t t
dy = = cot (2)
b cos θ b 2a sin 2t 2
y1 = dθ
= = − cot θ
dx
−a sin θ a 1 t dt
dθ
y2 = − cosec2 ·
b dθ 2 2 dx
y2 = cosec2 θ · 1 1 1
a dx = − ·
b 1 1 −b 2 2 sin 2 2a sin2 2t
2 t
= · · (− )=
a sin θ
2
a sin θ a sin3 θ
2 1
= − (3)
The coordinates of the centre of curvature are 4a sin4 2t
y1 (1 + y12 )
y1 (1 + y12 ) b cos θ x̄ = x −
x̄ = x − = a cos θ − + y2
y2 a sin θ
t 2 t t
a sin θ
2 3
a sin θ + b2 cos2 θ
2 2 = a(t − sin t) + cot 1 + cot 4a sin4
× − 2 2 2
b a2 sin2 θ
cos 12 1 4 t
= a(t − sin t) + · 4a sin
(a2 − b2 ) sin 2t sin2 2t 2
= cos3 θ
a = a(t − sin t) + 2a sin t
ax̄ = a2 cos θ − cos θ = a(t + sin t) (4)

× a2 − a2 cos2 θ + b2 cos2 θ 1 + y12
2 ȳ = y +
= a − b2 cos3 θ (3) y2

1 + y1 2 t 1
ȳ = y+ = a(1 − cos t) + (−4a) sin4
y2 2 sin2 2t

a2 sin3 θ a2 sin2 θ + b2 cos2 θ = a(1 − cos t) − 2a(1 − cos t)
= b sin θ − = −a(1 − cos t) (5)
b a2 sin2 θ
−(a2 − b2 ) sin3 θ The locus of (x̄, ȳ) is the cycloid represented by
=
b
bȳ = b2 sin θ − sin θ(a2 sin2 θ + b2 − b2 sin2 θ) x = a(t + sin t), y = a(1 − cos t). (6)
= −(a2 − b2 ) sin3 θ (4)
EXERCISE 5.4
From Eqs. (3) and (4), we have the evolute as
2 2 2 1. Find the centre of curvature at any point on the
(ax) 3 + (by) 3 = (a2 − b2 ) 3 (5) 2 2 2
curve x 3 + y 3 = a 3 .
1 1 1 1
where we have dropped the bars. Ans: x + 3x 3 y 3 , y + 3x 3 y 3
Example 5.25 2. Find the

equation of the circle of curvature at
Prove that the evolute of the cycloid a b 1 1 1
, on the curve x 2 + y 2 = a 2 .
x = a(t − sin t), y = a(1 − cos t) (1) 4 4

3a 2 3a 2 a2
is another cycloid. Ans: x − + y− =
y y 2
3. Find the centre of curvature and evolute of the

x2 y2
hyperbola 2 − 2 = 1.

a b
a2 + b2 −a2 − b2
Ans: Centre: sec θ,
2
tan θ
2
a b
Figure 5.8 Envelope of single parameter family
4. Determine the circle of curvature at the origin of of circles
2
the parabola y = mx + xa .
f (x, y, α, β) = 0 (5.27)
Ans: x2 + y2 = a(1 − m2 )(y − mx)
where α, β are parameters.
the evolute of the
5. Show that tractrix
t
x = a cos t + log tan , y = a sin t is the Example 5.26
2
The equation
catenary y = a cosh c .
x
6. Find the centre of curvature of the curve f (x, y, α) ≡ (x − α)2 + y2 − r 2 = 0 (1)

xy = c2 and hence find
its evolute. represents a single parameter family of circles with
3x2 + y2 3y2 + x2
Ans: Centre: , fixed radius r and with their centres lying on the x-
2x 2y axis.
23 2 2

Evolute: x + y − x − y 3 = (4c) 3 Example 5.27
The equation
7. Show that the evolute of the curve
x = a (a cos t + t sin t) , y = a (sin t − t cos t) g(x, y, α) ≡ x cos α + y sin α − p = 0 (2)
is x2 + y2 = a2 .
represents a family of straight lines inclined at (90°+
8. Find the equation of the circle of curvature at (1,1) α) with the x-axis and with a fixed perpendicular
of the curve x3 + y3 = 2xy. distance p from the origin.
Ans: (8x − 7)2 + (8y − 7)2 = 2
Definition A curve which touches each member
9. Show that the circle of curvature of the origin of a family of curves and at each point of it, it is
for x + y = ax2 + by2 + cx3 is touched by some member of the family of curves is
(a + b) x2 + y2 = 2x + 2y. called an envelope of the family of curves.
Example 5.28
Each of the two straight lines x = ±a is an envelope
5.9 Envelopes of the family of circles
[One-parameter family of curves] 2
x2 + y − α = a2
The equation
f (x, y, α) = 0 (5.26) with a fixed radius ‘a’ and with centres (0, α) (varying
α) lying on the y-axis.
where x and y are variables and α is a constant that
represents a plane curve in the x-y plane. If we vary Example 5.29
‘α’ we get different curves in the x-y plane, having The parabola y2 = 4ax is an envelope of the family of
a
one common property. We call the totality of curves, a straight lines y = mx + where ‘m’ is a parameter.
m
one-parameter family of curves, α being the param-
eter. Similarly a two-parameter family of curves is Note
represented as shown in Fig. 5.8. A family of curves may have one, two or more
envelopes or may have no envelope at all as in the (b) The given equation may be written as (y−mx)2 =
2
case of circles x2 + y − α = α2 . 1 + m2 which may be put in the form
Method for obtaining an envelope φ(x, y, m) = (x2 − 1)m2 − 2xym + (y2 − 1) = 0

(1)
(1) An envelope of a given family of curves
which is a quadratic equation in ‘m’. Equating the
f (x, y, α) = 0 (if it exists) discriminant to zero we obtain
may be obtained by eliminating ‘α’ between (−2xy)2 − 4(x2 − 1)(y2 − 1) = 0 ⇒ x2 + y2 = 1

(2)
f (x, y, α) = 0
∂f which is the unit circle.
and (x, y, α) = 0. Alternative method: Differentiating Eq. (1) partially
∂α
w.r.t ‘m’, we obtain
(2) If the equation of the given family of curves ∂φ xy
= (x2 − 1)2m − 2xy = 0 ⇒ m = 2 (3)
∂m x −1
f (x, y, α) = 0
Substituting for m in Eq. (1), we obtain
is a quadratic in α then the equation of the
(xy)2 xy
envelope is obtained by equating to zero, the (x2 − 1) · · 2 + (y2 − 1) = 0
discriminant of the quadratic. (x − 1) (x − 1)
2 2
Simplifying we get the envelope of Eq. (1) as

Example 5.30
Find the envelope of each of the families√
of straight x2 + y2 = 1 (4)
lines (a) y = mx + ma (b) y = mx + 1 + m2
√ (c) The given equation may be written as
(c) y = mx + a2 m2 + b2 [JNTU 2008S (Set 3)]
(y − mx)2 = ( a2 m2 + b2 )2
Solution (a) The given equation may be written
in the form Let
φ(x, y, m) = m2 x − my + a = 0 (1) φ(x, m) = (x2 − a2 )m2 − 2xym + (y2 − b2 ) = 0

(1)
This is a quadratic equation in ‘m’. The required
envelope is obtained by equating the discriminant to which is a quadratic in ‘m’. Equating the discrimi-
zero. That is, nants to zero, we obtain,
(−2xy)2 = 4(x2 − a2 )(y2 − b2 )
(−y)2 − 4x0 = 0 ⇒ y2 = 4ax (2)
simplifying we get the envelope as
which is the standard parabola.
Alternative method: Differentiating Eq. (1) partially x2 y2
+ 2 = 1 (Ellipse)
w.r.t ‘m’, we obtain a2 b
∂φ y Alternative method: Differentiating Eq. (1) partially
= 2mx − y = 0 ⇒ m = (3) w.r.t ‘m’, we get
∂m 2x
Substituting Eq. (3) into Eq. (1), we get ∂φ
= (x2 − a2 ) · 2m − 2xy = 0
∂m
y2 y 2xy
x− · y + a = 0 or y2 = 4ax (4) ⇒m= 2 (2)
4x2 2x x − a2
Substituting Eq. (3) into Eq. (1) Substituting Eq. (2) in Eq. (1), we obtain
(xy)2
xy
· (x2 − a2 ) − 2xy · 2 x 1/3 x x 1/3 3x
(x2 −a )2 2 (x − a2 ) − − +x =y ⇒ − =y
4 4 4 4
+ (y − b ) = 0
2 2
cubing both sides and simplifying we obtain the
Simplifying we get the equation of the envelope as equation of the envelope as
x2 y2
+ =1 (3) 256y + 27x4 = 0 (3)
a2 b2
(d) The given family of curves as Note
x y Here the method of discriminants is not applicable.
f (x, y, a, b) = + − 1 = 0 (1)
a b
Example 5.32
where the parameters a and b are connected by Find the envelope of the family of straight lines
a +b =c
n n n
(2) x y
+ =1
differentiating Eqs. (1) and (2) partially w.r.t ‘a’ a b
treating b as a function of ’a’, we get if the parameter a and b are connected by
x y db (a) a + b = c (b) a2 + b2 = c2 (c) ab = c2
− 2− 2 =0 (3)
a b da (d) an + bn = cn (e) am bn = cm+n .
db
and nan−1 + nbn−1 =0 (4)
da Solution
db an−1 (a) The given family of straight lines is
From Eq. (4), we have = − n−1 . Substituting
da b
this in Eq. (3), we get x y
+ =1 (1)
x y an−1 x y a b
− 2 − 2 ( n−1 ) = 0 ⇒ n+1 = n+1
a b b a b where a and b are connected by
( ax ) ( by ) x
+ y
1
⇒ n = n = an b n = n (5) a+b=c
a b a +b c (2)
1 1
⇒ a = (c x) n n+1 , b = (c y) n n+1 (6) So, we may write the equation as
Substituting these in Eq. (2), we get x y
φ(x, y, a) = + −1=0 (3)
n n a c−a
a = (c x) n n+1 + (c y)
n n+1 =c n
where c is constant, and ‘a’ is the parameter. Differ-

n n n2 n entiating Eq. (3) partially w.r.t ‘a’, we get
⇒x n+1 +y n+1 =c n− n+1
=c n+1
*
Example 5.31 ∂φ x y c−a y
=− 2 + =0⇒ = (4)
Find the envelope of the family of curves y = mx+m4 ∂a a (c − a) 2 a x
* √ √
for different values of ‘m’ c y x+ y
∴ =1+ = √
Solution a x x
√ √ √
1 x + y 1 1 x
Let φ(x, y, m) = m4 + mx − y = 0 (1) = √ ; = ·√
a c x c−a a y
∂φ x 1/3
√ √ √ √ √
= 4m3 + x = 0 ⇒ m = − x+ y x x+ y
∂m 4 = √ √ = √ (5)
(2) c x y c y
Substituting Eq. (5) into Eq. (3), we get where a and b are connected by
√ √ √ √ ab = c2 or b = c2 /a (2)
x+ y x+ y
√ ·x+ √ ·y =1 x ay
c x c y Let φ(x, y, a) = + 2 − 1 = 0 (3)
a c
After simplification the equation for the envelope is Differentiating Eq. (3) partially w.r.t ‘a’, we get
attained as
∂φ x y x y
√ √ √ =− 2 − 2 =0⇒ 2 = 2
x+ y= c (6) ∂a a c a c
(introducing a as a numberator )
(b) The given family of straight lines is
ax ay ( ax ) y x
+ by 1
x y ⇒ = ⇒ + b
= a
=
+ =1 (1) a2 c2 1 1 1+1 2
a b x y
∵ + = 1 ⇒ a = 2x,
where a and b are connected by a b
b = 2y ⇒ c2 = ab = 2x · 2y or 4xy = c2
a2 + b2 = c2 ⇒ b2 = c2 − a2 (2)
x y which is the envelope of the family of curves (1)
Let φ(x, y, a) = + √ −1=0 (3) subject to (2).
a c − a2
2
Differentiating Eq. (3) partially w.r.t ‘a’, we get (e) Let

x y
f (x, y, a) =+ −1=0 (1)
∂φ x 1 ay x y a b
=− 2 − (−2a) = 0 3 = 3
∂a a 2 (c − a )
2 2 3/2 a b where am bn = cm+n (2)
(4)
From Eq. (2) we get
x2/3 y2/3 x2/3 + y2/3
⇒ 2 = 2 = , 1 am/n
a b c2 = (m+n)/n (3)
Sum of Nrs b c
= , by Eq. (2)
Sum of Drs Using condition (3) Eq. (1) can be written as
+
1 x2/3 + y2/3 1 x2/3 + y2/3 x am/n
⇒ = , = (5) f (x, y, a) = + (m+n)/n · y − 1 = 0 (4)
a cx 1/3 b cy1/3 a c
Differentiating (4) partially w.r.t ‘a’
Substituting these values in Eq. (1)
∂f (x, y, a) x m am/n − 1
1 =− 2 + y=0 (5)
1 = (x2/3 x2/3 + y2/3 + y2/3 x2/3 + y2/3 ) ∂a a n c(m+n)/n
c
Multiplying Eq. (5) by a2 and transposing we get
1 2/3 3/2
= x + y2/3 m+n n
c m a(m+n)/n nx
⇒ x2/3 + y2/3 = c2/3 (6) x = · (m+n)/n y ⇒ a = c (6)
n c my
which is the equation of the envelope. On multiplying Eq. (4) by ‘a’ we have
a (m+n)/n
nx
(c) The given family of straight lines is a=x+ y=x+ ·y
c my

x y m+n
+ =1 (1) ⇒a= x by Eq. (6) (7)
a b m
From Eqs. (6) and (7) we get Example 5.34

n/(n+n) Find the envelope of the family of straight line
nx m+n x y
c = x (8) + = c where α is the parameter.
my m sin α cos α
Solution The given family of curves is
Taking (m + n)th powers on both sides of (8)
x y
nx n m + n (m+n) m+n φ(x, y, α) = + −c =0 (1)
c m+n
= x sin α cos α
my m
differentiating Eq. (1) partially w.r.t ‘α’, we get
⇒ cm+n mm · nn = (m + n)m+n xm yn (9)
∂φ −s cos α sin α
=x +y =0
Example 5.33 ∂α sin α cos2 α
Find the envelope of the family of curves x y
⇒ =
gx2 sin3 α cos3 α
y = x tan α − 2 α being the parameter.
2u cos2 α x1/3 y2/3 x2/3 + y2/3
⇒ = =
Solution The given family of curves is sin2 α cos α
2 1
1 x2/3 + y2/3 1/2
⇒ = ,
gx2 sin α x2/3
y = x tan α − 2 (1) 1 x2/3 + y2/3 1/2
2u cos2 α = (2)
cos α y2/3
The equation may be written in the form
Substituting these values in Eq. (1), we get
φ(x, y, α) = gx2 (tan2 α) − 2u2
x2/3 + y2/3 1/2 x2/3 + y2/3 1/2
x(tan α) + (gx2 + 2u2 y) = 0 (2) x + y =c
x2/3 y2/3
= gx · m − 2u x · m + (gx + 2u y) = 0
2 2 2 2 2
⇒x + y = c
2/3 2/3 2/3
(3)
(m = tan α) (3)
on simplification.
The above equation is a quadratic equation in m =
tan α. Equating the discriminant to zero we have Example 5.35
x2 y2
(−2u x) − 4gx (gx + 2u y) = 0
2 2 2 2 2
(4) Find the envelope of 2 + 2 = 1 where a is a
a c − a2
parameter.
⇒ u = g x + 2gyu
4 2 2 2
(5)
2u 2
u 2 Solution We can write the given equation as
⇒ x2 = − (y − ) (6)
g 2g φ(x, y, a) = a4 − (x2 − y2 + c2 )a2 + c2 x2 = 0 (1)
which is the required envelope of the given curves. which is a quadratic in a2 . To find the envelope we
The envelope is a parabola with the negative y-axis have to equate to zero the discriminant of Eq. (1).
u2 This gives
as its axis. Semi latus-rectum = and vertex at
2g
u2 (x2 − y2 + c2 )2 = 4c2 x2 (2)
0,
2g
Factorizing Eq. (2), we get the following two pairs of
Note parallel lines
The given curves are paths of projectiles with the (x − c − y)(x − c + y) = 0 and
angle of projection α inclined to the horizontal. (x + c − y)(x + c + y) = 0 (3)
forming a square which is the envelope of the given 1

= (2)
curves. x2 + y2
y x
⇒ cos θ = , sin θ = (3)
Example 5.36 x2 + y2 x2 + y2
Find the envelope of the curves
a2 b2 c2 Eliminating θ between Eqs. (1) and (3), we obtain
cos θ − sin θ = for different values of θ.
x y a the equation of the envelope as
Solution Let
y3 x3
a2 b2 c2 x· + y · =a
φ(x, y, θ) = cos θ − sin θ − = 0 (1) (x2 + y2 )3/2 (x2 + y2 )3/2
x y a 1 1 1
Differentiating Eq. (1) partially w.r.t ‘θ’, we get ⇒ 2+ 2 = 2 (4)
x y a
∂φ a2 b2
= − sin θ − cos θ = 0 on simplification.
∂θ x y
+ Example 5.38
a4 4
+ by2
2
( −by )
2
( ax ) x2 Show that the envelope of a circle whose centre on
⇒ = = (2) xy = c2 and passes through the origin is the lemnis-
cos θ sin θ 1
(a /x)
2 cale r 2 = 8c2 sin 2θ.
⇒ cos θ = + ,
a4 b4
x2
+ y2
Solution Let (ct, c/t) be any point on the
hyperbola
(−b2 /y) c2
sin θ = + = xy = c2 (1)
a4 4
+ by2 a
x2
by the method of proportions Equation of the circle with (ct, c/t) as centre and
passing through the origin is given by
Substituting these values in Eq. (1)
c2
a2 (a2 /x) b2 (−b2 /y) c2 (x − ct)2 + (y − c/t)2 = c2 t 2 + (2)
+ − + = t2
x a4 + b4 y a4 + b4 a c
x2 y2 x2 y2 ⇒ x2 + y2 − 2xct − 2 y=0
t
a4 b4 c2 a4 b4 c4 ⇒ 2xt 2 − (x2 + y2 )t + 2cy = 0 (3)
+ = ⇒ + = (3)
x2 y2 a2 x2 y2 a2
which is a quadratic in ‘t’. The required envelope is
which is the required envelope. obtained by equating the discriminant of Eq. (3) to
Example 5.37 zero.
Find the envelope of x cos3 θ + y sin3 θ = a for dif- (x2 + y2 )2 = 4(2cx)(2cy) = 16c2 xy (4)
ferent values of θ.
Putting x = r cos θ, y = r sin θ we get Eq. (4) in
Solution Let polar form
φ(x, y, θ) = x cos3 θ + y sin3 θ − a = 0 (1)
r 2 = 8c2 sin 2θ (5)
Differentiating Eq. (1) w.r.t ‘θ’, we get
∂φ EXERCISE 5.5
= x · 3 cos3 θ(− sin θ) + y · 3 sin2 θ cos θ = 0 1. Find the envelope of the straight lines
∂θ
(a) y = mx + am √
3
cos θ sin θ cos2 θ + sin2 θ
⇒ = = (b) y = mx + a b2 + m2
y x y2 + x2 where ’m’ is the parameter.
2. Find the envelope of the family of circles 5. Find the envelope of the family of curves
(a) x2 + y2 − 2ax cos α − 2ay sin α = c2 , α being (a) x sec2 α + ycosec2 α = a (α : parameter)
the parameter. (b) x cosec α − y cot α = c (α : parameter)
(b) (x − α)2 + (y − α)2 = 2α
6. Find the envelope of parabolas (x/a)1/2 +
3. Find the equation of the envelope of the straight (y/b)1/2 = 1 where a and b are connected by
ax by (a) a + b = c (b) ab = c2
lines − = a2 − b2 , θ being the
cos θ sin θ x y
parameter. 7. Find the envelope of + = 1 with the condition
a b
an + bn = cn
4. Find the envelope of the system of concentric and
coaxial ellipses of constant area. 8. Find the envelopes of
x2 y2 (a) y = mx + am2
[Hint: Take 2 + 2 = 1;
a b (b) y = mx + am4
a, b parameters such that ab = c2 ] (c) y = mx + amp
Curve-Tracing
6
6.1 Introduction
An elementary method of drawing the graph of a
plane curve is to tabulate the x and y values that
satisfy the equation of the curve, plot the points and
join them by means of a smooth curve. A systematic
and more elegant method is the analytical method
of studying the characteristics of the curve such as
its symmetry, tangents, asymptotes and regions and
drawing its graph. Knowledge of curve-tracing helps
in finding its length (rectification), area bounded by
it (quadrature) and the volume of the solid of its
revolution.
In this chapter, we discuss the tracing of plane
curves represented by (1) cartesian (2) polar and
(3) parametric equations.
6.2 Curve-Tracing—Cartesian
Equations
Let f (x, y) be an algebraic function of two variables
x and y. f (x, y) = 0 represents a plane curve called
an algebraic curve.
A general procedure for tracing such a curve
consists of a study of the following characteristics.
6.2.1 Symmetry
(1) Symmetry about the x-axis A curve is sym-
metric about the x-axis if it contains only even powers
of y, i.e., if its equation is unaltered when we replace
y by −y or if f (x, −y) = f (x, y).
(a) y2 = 4ax (Parabola with the x-axis as its axis)
(b) y2 (a + x) = x2 (3a − x) Figure 6.1 Curves symmetric about the x-axis
(c) ay2 = x3 (Semi-cubical parabola) (Vertical
line PP is bisected at M by the x-axis.)
y
(2) Symmetry about the y-axis A curve is sym-
metric about the y-axis if its equation contains only
even powers of x, i.e., if its equation is unaltered when
M
P'(–x, y) P (x, y) we replace x by −x or f (−x, y) = f (x, y).
(a) x2 = 4ay (Parabola with the y-axis as its axis)
x (b) a2 x2 = y3 (2a − y)
O
(Any horizontal line PP cutting the curve is
bisected at M by the y-axis.) (Refer Fig. 6.2)
(a) x2 = 4ay
(Parabola with y-axis as its axis)
(3) Symmetry about the straight line y = x A
curve is symmetric about the straight line y = x if its
y equation is unaltered when x and y are interchanged
(0, 2a) or f (x, y) = f (y, x). (Refer Fig. 6.3)
(a) x3 + y3 = 3axy (Folium of Des Cartes)
P'(–x, y) P (x, y)
M (b) xy = c2 (Rectangular hyperbola)
(4) Symmetry about the straight line y = −x A

x curve is symmetric about the straight line y = −x if
O
its equation is unaltered when x and y are replaced
(b) a2x2 = y3(2a–y)
simultaneously by −y and −x, respectively.
(a) x3 − y3 = 3axy (Folium of Des Cartes)
Figure 6.2 Curves symmetric about the y-axis (b) xy = c2 (Rectangular hyperbola)
y
y)
x, y
P( P(
x,
y)
x) y = x − a (α > 0)
(y, M
P' x P(y, x) x
O O
x
y=
y = −x
y+x+a=0
(a) x3 + y3 = 3axy (a) x3 + y3 = 3axy
(Folium of Des Cartes) (Folium of Des Cartes)
y
Y y
x Y x
x x
O O
x' x'
Y' Y'
(b) xy = c2 (b) xy = c2
(Rectangular hyperbola) (Rectangular hyperbola)
Figure 6.3 Curves symmetric about the Figure 6.4 Curves symmetric about
line y = x y = −x
Curve-Tracing 6-3
(5) Symmetry about the origin (0, 0) A curve is

symmetric about the origin if its equation is unaltered
when both x and y are replaced simultaneously by −x
and −y, respectively.
(a) xy = c2 (Rectangular hyperbola)
1
(b) y = x +
x
(c) a2 y = x3
Every line segment joining a pair of points P and

P on either side of the origin is bisected at 0. The
curve lies in opposite quadrants.
(6) Symmetry about both the axes A curve is
symmetric about the x and y axis if both the con-
ditions (1) and (2) of Section 6.2.1 are satisfied when
applied separately or simultaneously.
(a) x2 + y2 = a2 (Circle)
x2 y2
(b) 2
+ 2 = 1 (Ellipse)
a b
x2 y2
(c) 2
− 2 = 1 (Hyperbola)
a b
A curve which is symmetric about both the x and y

axes is also symmetric about the origin. But a curve
which is symmetric about the origin need not be
symmetric about both the axes.
1
(a) xy = c2 and (b) y = x + (Rectangular
x
hyperbola) are symmetric about the origin but are
not symmetric about both the axes.
6.2.2 Curve Passing Through the

Origin
Find whether the curve passes through the origin
by checking whether its equation is satisfied by
(0, 0). Find the equations of the tangent at the ori-
gin by equating to zero, the lowest degree term in the
equation.
If there are two or more tangents at the origin, then
Figure 6.5 Curves symmetric about the origin the origin is called a multiple point (double point,
(0, 0). triple point, etc.). The point is a node if two or more
distinct tangents exist at the origin, and it is a cusp
if the tangents are coincident.
√
If the imaginary unit i = −1 appears in the For example, x5 + y5 = 5a2 x2 y has no asymp-
equations of the tangents at the origin, then it is an totes parallel to the axes since the coefficients of
isolated or conjugate point. x5 , y5 are 1, 1, respectively which are constants.
(a) y2 = 4ax (Equating to zero the lowest degree 4. Oblique asymptotes (asymptotes not parallel to the
term 4ax we get x = 0 (y-axis) as the tangent axes)
at the origin.) Method 1
(b) x3 + y3 = 3axy (Equating to zero the lowest
degree term xy = 0 we get x = 0, y = 0 as Let y = mx + c (6.1)
tangents at the origin.)
be an asymptote of the curve
(c) x4 = a2 (x2 − y2 ) (Equating to zero, the low-
est degree term y2 = x2 , i.e., y = ±x as f (x, y) = 0 (6.2)
tangents at the origin.)
Eliminate y between the two equations, equate the
6.2.3 Intersection of the Curve coefficient of the highest power of x to zero and solve
with the Coordinate Axes for m and c. When we substitute these values in Eq.
Find the points where the curve meets the axes by (6.1) we get the asymptotes.
putting x = 0 and y = 0 in turn. Also, find the
Example 6.1
equations of the tangents at these points, if they exist.
x2 + 2x − 1
y= (1)
6.2.4 Asymptote of a Curve x
(See Section 6.5) Solution Let
A straight line cutting a curve in two points at an y = mx + c (2)
infinite distance from (0, 0) but is not itself wholly at
infinity is called an asymptote of the curve. be an asymptote then
To find the asymptotes of the curve: x2 + 2x − 1
mx + c =
1. Asymptotes parallel to the x-axis are obtained by x
equating to zero the coefficients of the highest or mx2 + cx = x2 + 2x − 1
power of x in the equation. ⇒ m = 1, c = 2 (3)
2. Asymptotes parallel to the y-axis are obtained by ∴ The asymptote is y = x + 2. (4)
equating to zero the coefficients of the highest
power of y in its equation. Method 2
For example, the equation x2 y − y − x = 0 In the highest degree terms (i.e., nth degree terms)
is of degree 3. It has a maximum number of 3 of f (x, y) put x = 1 and y = m and obtain a func-
asymptotes. tion φn (m) of m. Solve φn (m) = 0 for m. Suppose
The coefficients of the highest power of m1 , m2 , · · · , mn are the n roots of φn (m) = 0.
x (i.e., x2 ) is y so that the asymptote parallel to In the next highest degree terms (i.e., (n − 1)th
the x-axis is y = 0 (the x-axis itself). degree terms) of f (x, y) put x = 1 and y = m and
The coefficient of the highest power of y (i.e., y) obtain φn−1 (m) of m. Substituting m1 , m2 , · · · , mn in
is (x2 − 1) so that the asymptotes parallel to the the equation
y-axis are x = ±1. φn−1 (m)
c=−
φn (m)
3. Asymptotes parallel to the axes do not exist if the
coefficients of the highest powers of x and y are ( denotes differentiation with respect to m) we get
real constants or imaginary numbers. c1 , c2 , · · · cn of c.
Curve-Tracing 6-5
Example 6.2 6.2.6 Rising and Falling Curves

Find the asymptotes of x3 + y3 = 3axy. dy
Calculate from f (x, y) = 0. Then if
Solution Let f (x, y) = x3 + y3 − 3axy = 0 dx
In the highest degree terms x3 + y3 put dy
x = 1, y = m, we get (1) >0 ∀ x ∈ [a, b] then the curve is rising in
dx
[a, b];
φ3 (m) = 1 + m3 = 0 (1)
(whose real solution is m = −1) dy
(2) < 0 ∀ x ∈ [a, b] then the curve is falling in
φ3 (m) = 3m2 (2) dx
[a, b]; and
In the next highest degree terms −3axy put dy
x = 1, y = m, we get (3) if = 0 at any point P(x0 , y0 ), then it is nei-
dx
ther rising nor falling: P(x0 , y0 ) is a stationary
φ2 (m) = −3am
point at the top-most or bottom-most point of
φ2 (m) −3am a a
∴ c=− =− = = the curve.
φ3 (m) 3m 2 m −1
Hence the equation of the asymptote is
6.2.7 Concavity or Convexity
y = −x − a or x + y + a = 0. (3) d 2y
Calculate . Then if
Example 6.3 dx2
Find the asymptotes of y3 − x2 y + 2y2 + 4y + x. d 2y
(1) > 0 then the curve is concave upwards (cup
Solution Let dx2
holding coffee)
f (x, y) = y3 − x2 y + 2y2 + 4y + x = 0. (1) d 2y
φ3 (m) = m3 − m (2) (2) < 0 then the curve is concave downwards
dx2
φ2 (m) = 2m2 (3) (cap on the head) and
φ3 (m) = 0 ⇒ m = 0, 1, −1 (4) d 2y
(3) = 0 at a point P(x0 , y0 ) then P is called an
φ2 (m) 2m2 dx2
c=− =− 2 = 0, −1, −1 (5) inflexion point where the curve changes direction
φ3 (m) 3m − 1
of concavity/convexity from downward to upward
(corresponding to m = 0, 1, −1)
or vice versa.
∴ The asymptotes are ⎫
y=0 ⎬
Example 6.4
y =x−1 . (6)
⎭
y = −x − 1
Trace the curve x2 y2 = a2 (y2 − x2 ). (1)
6.2.5 Region or Extent [JNTU 1995S)]

Solution
We can determine the region in which the curve lies
(1) Symmetry Since Eq. (1) contains even powers
by solving the equation f (x, y) = 0 for y and knowing
of x and y the curve is symmetric about both the
the domain of variation of x and similarly solving the
coordinate axes.
equation for x and knowing the domain of variation
of y. (2) Passing through the origin Eq. (1) is satisfied
If y (or x) is imaginary for any range of values of x by x = 0 and y = 0. So, the curve passes through the
(or y) it means that the curve does not lie in that range. origin.
(3) Tangents at the origin Equating to zero the (8) Concavity or Convexity
lowest degree terms of Eq. (1), we get
ax dy
a2 (y2 − x2 ) = 0 ⇒ y = ±x y = ±√ ;
a −x
2 2 dx
The tangents are real and distinct. So the origin is a3 d 2y
a node. = ± 3/2 ;
a2 − x2 dy2
(4) Extent or region Eq. (1) can be written as
3a3 x
ax = 5/2 (3)
y = ±√ (2) a2 − x2
a2 − x 2
If |x| > a, then y is imaginary. So no part of the dy
Since > 0 for all x the curve is always rising.
curve lies outside the straight lines x = ±a. dx
y = 0 if x = 0. As x varies from 0 to a, y increases The origin is a point of inflexion since
from 0 to ∞.
d 2y
=0
(5) Asymptotes Asymptotes parallel to the y-axis dx2 (0,0)
are obtained by equating to zero the coefficient of the
highest power of y: thus x2 − a2 = 0 or x = ±a are Example 6.5
asymptotes parallel to the y-axis. There is no other Trace the curve
asymptote.
y2 (2a − x) = x3 (Cissoid of Diocles). (1)
(6) Extension to infinity As x increases from 0 to
a, y increases from 0 to ∞ in the first quadrant. The Solution
curve is symmetrical with respect to both the axes (1) Symmetry Eq. (1) contains even powers of y.
and the curve lies in all the quadrants. So, the curve is symmetric about the x-axis
(7) Intersection with the coordinate axes The
(2) Passing through the origin Eq. (1) is satis-
curve intersects the axes only at the origin and at
fied by x = 0; y = 0. So, the curve passes through
no other points.
the origin.
The graph for the curve is given in Fig. 6.6.
(3) Tangents at the origin Equating to zero the
lowest degree terms,
2ay2 = 0 we get y = 0, y = 0 (double root)
The two tangents to the curve at the origin are real

and coincident so that the origin is a cusp.
(4) Extent or region y2 is negative for negative
values of x.
Thus, y is imaginary for x < 0; so, no part of the
curve lies to the left of the y-axis.
Again, y is imaginary for x > 2a; so, no part of the
curve lies to the right of the straight line x = 2a.
(5) Asymptotes Asymptotes parallel to the y-axis
highest power of y. Thus x = 2a is an asymptote, and
Figure 6.6 there is no other asymptote to the curve.
Curve-Tracing 6-7
Figure 6.8
As x → ∞, y → ∞. The curve extends from the

origin to infinity in the I quadrant and because it is
Figure 6.7 symmetric about the x-axis the reflection of the curve
in the x-axis lies in the IV quadrant.
(6) Extension to infinity Writing Eq. (1) in the
x3/2 (6) Asymptotes Since the coefficients of the high-
form y = √ we see that as x varies from 0 to est and the next highest degree terms (3rd and 2nd
2a − x
2a, y ranges from 0 to ∞. degree terms) are 1 and 0, respectively, there are no
The graph for the curve is given in Fig. 6.7. asymptotes to the curve.
The curve is called a semi-cubical parabola whose
Example 6.6 graph is shown in Fig. 6.8.
Trace the curve
Example 6.7
ay2 = x3 (Semi-cubical parabola). (1) Trace the curve
a+x
Solution y2 = x2 · (Strophoid). (1)
a−x
(1) Symmetry The equation contains only even Solution
powers of y, so the curve is symmetric about the
x-axis. (1) Symmetry Eq. (1) contains only even powers
of y. So, the curve is symmetrical about the x-axis.
(2) Origin The curve passes through the origin
since (0, 0) satisfies the equation. (2) Origin The curve passes through the origin
since x = 0, y = 0 satisfy Eq. (1).
(3) Tangents at (0, 0) Equating the lowest degree
terms we get y2 = 0 ⇒ y = 0, y = 0. Thus the (3) Tangents at the origin Equating to zero the
x-axis is a tangent to the two parts of the curve at the lowest degree terms, we get y = ±x as the tangents
origin. at the origin. These are real and different. So, the
∴ The origin is a cusp. origin is a node.
(4) Intersection points with the coordinate axes (4) Extent or regionEq. (1) can be written as
x = 0 ⇔ y = 0. So, the curve does not cut the *
coordinate axes at any point other than the origin. a+x
y=±x
a−x
(5) Region in which the curve lies If x < 0, then y
is imaginary. So, no part of the curve lies to the left When x < −a or x > a, y is imaginary. So no part
of the y-axis. of the curve lies outside the lines x = −a, x = a.
(2) Origin The curve passes through the origin.

(3) Tangent at the origin The lowest degree term
in Eq. (1) is 3xy. Equating it to zero, we have xy = 0,
i.e., the coordinate axes are both tangent to the curve
at the origin.
(4) Intersection with the axes The curve does not
intersect the coordinate axes except at the origin.
(5) Region in which thecurve lies Put y = x in
3a
Eq. (1) 2x = 3ax ⇒ x x −
3 2 2
= 0.
2
In the first quadrant,
thecurve lies in between the
3a 3a
points (0, 0) and , .
2 2
(6) Asymptotes Since the coefficients of the high-
est degree terms are constants there are no asymp-
totes parallel to the coordinate axes.
Figure 6.9 Let y = mx + c (m = 0) (2)

a
As x increases from −a to − , y decreases from be an oblique asymptote. Eliminating y between Eqs.
2
a a (1) and (2)
0 to − and as x increases from − to 0, y increases
2 2
a
from − to 0. y
2 y=x
(5) Asymptotes Asymptotes parallel to the y-axis
highest power of y; thus x = a is an asymptote.
The coefficient of the highest power of x is 1. So, Asymptote
there is no asymptote parallel to the x-axis. x
O
(6) Extension to infinity As x increases, y also y+x+a=0
increases steadily to ∞. The curve extends to infin-
ity in the first quadrant. By its symmetry about the
x-axis, it extends to infinity in the fourth quadrant.
(7) Intersection with the coordinate axes y = 0 Figure 6.10
when x = 0 or − a. The curve cuts the x-axis at
x = −a and x = 0.
The graph for the curve is given in Fig. 6.9. x3 + (mx + c)3 − 3ax (mx + c) = 0

1 + m3 x3 + 3x2 cm2 − am
Example 6.8
Trace the curve + 3 mc2 − ac x + c3 = 0
x3 + y3 = 3axy (Folium of Des Cartes). (1) Equating to zero the coefficient of x3 , we get
Solution
m3 = −1 ⇒ m = −1
(1) Symmetry There is no symmetry about the
x-axis or y-axis. But there is symmetry about the Equating to zero the coefficient of x2 , we get
straight line y = x since the equation remains a
unaltered where x and y are interchanged. m2 c − am = 0 ⇒ c = = −a(m = −1)
m
Curve-Tracing 6-9
∴ y = −x − a is the oblique asymptote to the Since there are no real roots for this equation there
curve. are no asymptotes parallel to the x-axis.
Refer Fig. 6.10 for the graph associated with the Equating to zero the coefficient of the highest
curve. power y2 of y, x = 0 (the y-axis).
Example 6.9 The y-axis is an asymptote. (Refer Fig. 6.11)
Trace the curve
Example 6.10
xy2 = 4a2 (2a − x) (witch of Agnesi or Versiera). (1) Trace the curve
Solution xy2 = a2 (x − a) (a > 0) . (1)
(1) Symmetry Eq. (1) contains even powers of y.
So, the curve is symmetric about the x-axis. It is not Solution
symmetric about the y-axis. It is not symmetric about (1) Symmetry Eq. (1) contains only even powers
the origin. of y, so the curve it represents is symmetrical about
(2) Origin The curve does not pass through the the x-axis.
origin since (0, 0) do not satisfy the equation. (2) Special points If x = a, then y = 0. So, the
(3) Intersection with the axes Put y = 0 in Eq. (1) curve cuts the y-axis at A(0, 0).
*
we get x = 2a. The curve cuts the x-axis at (2a, 0). x−a
It does not cut the y-axis. If x > a, then y = ± a . There are two
x
(4) Region in which it lies Solving Eq. (1) for y branches for y in the range −a < y < a.
√ Also lim y = ± a. A part of the above has
2a 2a − x
y=± √ x→∞
two branches, one above and one below the x-axis
x
extending up to ∞, y = ± a being the asymptotes.
y is imaginary when x < 0 and x > 2a. The curve lies
Again, the equation can be written as
only in the strip 0 ≤ x ≤ 2a, y ∈ R.
(5) Asymptotes Parallel to the x-axis x−a y2 x−a
= 2 ⇒ 1−
Equating to zero the coefficient of the highest power x a x
of x: y2 a2 − y2
x(y2 + 4a2 ) = 8a3 = 1− 2 =
a a2
we get y2 + 4a2 = 0 a 3
or x = 2
a − y2
Now if y > a, then x is negative. Thus a branch of
Asymptote
the curve lies in the second quadrant above the line
y = a. By symmetry, there is another branch below
the line y = −a both extending up to −∞.
a
(3) Asymptotes Equating to zero the coefficient of
O (2a, 0) the highest power of y we get x = 0, i.e., the y-axis
is an asymptotes to the curve.
Again, equating to zero the coefficient of the high-
Asymptote est power of x, we get y = ±a as the asymptotes to
the curve.
x y2=4 a 2 (2a - x) (4) Tangent at A(a, 0)
(witch of Agnesi or Versiera)
a3
Figure 6.11 x=
a2 − y2
Tangent
Asymptote
Asymptote
y=a
A(a, 0)
x
O
y =−a
Asymptote
Asymptote
x y2=a2 (x − a)
Figure 6.12 Figure 6.13
Differentiating this with respect to y, we get The x-axis is tangent to Eq. (1) at the points
(± a, 0). The y-axis is a tangent at the points (0, ± b)

dx 2a3 y dx (when x = 0, y = b). As x increases from 0 to a, y
= 2 ⇒ = 0. decreases from (b, 0). If b = a, the curve becomes
dy a −y
2 2 dy (a,o)
an astroid (Refer Fig. 6.13).
The tangent at A(a, 0) to the curve is perpendicular x 2/3 y 2/3
to the x-axis at A(a, 0). + = 1.
a b
The graph for the curve is given in Fig. 6.12. Example 6.12
Trace the curve
Example 6.11
y2 (a − x) = (a + x) . (1)
Trace the curve Solution
x 2/3 y 2/3 (1) Symmetry The equation contains even powers
+ =1 of y only. So, the curve is symmetric about the x-axis.
a b
or x = a cos3 t, y = b sin3 t (Hypocycloid). (2) Origin The curve does not pass through the
origin since (0, 0) does not satisfy Eq. (1).
Solution (3) Region in which the curve lies
(1) Symmetry Eq. (1) contains even powers of x
and even powers of y and so the curve it represents (i) Solving Eq. (1) for y
*
is symmetrical about the axes. a+x
y=±
(2) Special points Eq. (1) gives no real values of x a−x
if |y| > b and similarly it gives no real values of y if y is real only if −a < x < a. So the curve does
|x| < a. The curve cuts the x-axis at the points A(a, 0). not lie outside the lines x = −a and x = a.
A = (−a, 0) and the y-axis at the points B(0, b) and a+x
B = (0, −b). (ii) Also, = y2
a−x
(3) Tangents Differentiating Eq. (1), we get By componendo-dividendo
3/2 a y2 + 1
dy b y 1/3 = 2
=− x y −1
dx a x
y2 − 1
dy dx or x = a· 2
=0 =0 y +1
dx (±a,o) dy (o,±b)
Curve-Tracing 6-11
Solution
(1) Symmetry The equation contains only even
powers of x. So, the curve it represents is symmetrical
about the y-axis.
(2) Special points
(a) y = −1 when x = 0. For y = 0, x has no real
values. This shows that the curve does not intersect
the x-axis. It crosses the y-axis at the point (0, −1).
(b) Eq. (1) can be written as
y+1
x2 = .
y−1
Figure 6.14
When n lies between −1 and +1, x2 is negative
and hence x has no real values. Thus the curve does
When x = 0 we get y = ±1 so that the curve cuts not lie between the horizontal lines y = ±1.
the y-axis at P(0, 1) and Q(0, −1).
y
Similarly when y = 0, x = −a, the curve cuts the
x-axis at A(−a, 0).
(4) Asymptotes parallel to the axes Equating to
zero the coefficient of the highest power y2 of y, we
get the asymptote parallel to the y-axis as x = a.
There is no asymptote parallel to the x-axis since
y=1 y=1 y=1
by equating to zero the coefficient of x, we get y2 +
1 = 0 which has no real roots. x' O (0, –1) x
(5) Tangents to the curve Differentiating
a+x x2 + 1
y=
y2 = with respect to ‘x’, we get x2 −1
a−x
dy (a − x) 1 − (a + x) (−1)
2y =
dx (a − x)2 Figure 6.15
dy a
⇒ = (3) Asymptotes Equating to zero the coefficient of
dx y (a − x)2
the highest power of x, we get y = 1 as the asymptote
dy 1 parallel to the x-axis.
Slope of tangent at P(0, 1) is =
dx (0,1) a Equating to zero the coefficient of the highest
dy 1 power of y, we get x = ±1.
and at Q (0, −1) is =− ·
dx (0,−1) a (4) Tangents Differentiating Eq. (1), we get
Fig. 6.14 shows the graph associated with the dy 4x
graph. = − 2 ,
dx x −1
2

Example 6.13 d 2y 4 x2 + 1
= 3
Trace the curve dx2 x2 − 1
x2 + 1 dy
y= . (1) = 0 when x = 0.
x2 − 1 dx
The curve attains its maximum value when x = 0 Substituting values of θ, we get values for r. It is
and maximum value is −1. important to plot the points where r is maximum or
The graph for the curve is given in Fig. 6.15. minimum and to find the values of θ, when the curve
passes through the origin, if at all it does.
6.3 Tracing of Curves—Polar
Coordinates
6.3.1 Introduction
The position of a point P on a plane can be indicated
by stating
(1) its distance r from a fixed point O (called the
pole) and
(2) the inclination θ of OP to a fixed straight line
OA through O.
OA is called the initial line and OP is called the
radius vector. Then r and θ are called the polar coor-
dinates of the point, r is called the radial distance
and θ is the vectorial angle. θ is considered positive
when measured in the counter-clockwise direction. r
is considered positive when measured away from O
along the line bounding the vectorial angle.
Let P = (r, θ). If OP is extended backwards
to P such that |OP| = |OP |, then P = (−r, θ) or
(r, θ + π). (Refer Fig. 6.16)
For converting cartesian coordinates to polar
coordinates or polar to cartesian, it is customary to
take the pole ‘O’ as the origin and the initial line OA
as the x-axis.
Equation of a curve in polar coordinates is
expressed implicitly as F(r, θ) = 0 or explicitly as
r = f (θ), which is obtained from the former by
solving for r.
Figure 6.16 Figure 6.17 Symmetries of the curves

Curve-Tracing 6-13
These are readily verified by considering the

symmetrically located points as shown in Fig. 6.17.
6.3.2 Curve-tracing when the

Equation is in Polar
Coordinates
If the equation can be transformed to cartesian form
by putting x = r cos θ, y = r sin θ, we can trace the
curve by the method discussed earlier. Otherwise, we
have to observe the following points to trace a curve
whose equation is given in polar coordinates.
(1) Symmetry Find out whether the curve is sym-

metrical about any line or pole, by verifying the
following points.
(a) A curve is symmetrical about the initial line (the

x-axis) if its equation remains unaltered when θ
is replaced by −θ
Example: (a) r 2 = a2 cos 2θ,
(b) r = a(1 + cos θ),
(c) r = a + b cos θ
Figure 6.19 Curves symmetrical about the

π
vertical line θ =
2
(c) A curve is symmetrical about the pole if its

(a) Lemniscate of equation remains unaltered when r is replaced
Bernoulli (b) Cardioid by −r, that is, when it contains only even powers
of r or when θ is replaced by (π + θ).
Example: (a) r 2 = a2 cos 2θ (Lemniscate of
Bernoulli) (b) r = a sin 3θ (Three-leafed rose)
(d) Put r = 0 and solve the equation for θ. If θ is

real then the curve passes through the pole and
(c) Pascal’s Limacon θ gives the tangent at the pole.
Figure 6.18 Curves symmetrical about the initial Example: r = aθ (Spiral of Archimedes)
line
(b) A curve is symmetrical about the vertical line

through the pole (the y-axis), that is, θ = π if its
equation remains unaltered when θ is replaced
by (π − θ).
Example: (a) r = a(1 + sin θ), (b) r = a sin 3θ
Figure 6.20 Spiral of Archimedes
Solution
(1) Symmetry Replace θ by −θ. r changes to −r.
Hence the curve has symmetry with respect to the
π
line θ = (y-axis).
2
(2) Region or extent As θ increases from 0 to ∞, r
also increases from 0 to ∞. Also as θ decreases from
0 to −∞, r increases.
θ=0 π/2 π 3π/r 2π

r=θ πa/2 πa 3πa/2 2πa
= 1.6a = 3.1a 4.7a 6.21a
(3) Asymptotes No asymptotes exist to the curve,

Figure 6.21 4-leafed rose r = aθ since as θ → ∞, r → ∞.
The shape of the curve is shown in Fig. 6.22.
(e) A curve is symmetrical about the line θ = π/4,
π
(i.e., y = x) if f r, − θ = f (r, θ)
2
Example: r = a sin 2θ (Four-leafed rose) Q
(f) A curve is symmetrical

aboutthe line θ = 3π/4,
3π
(i.e., y = x) if f r, − θ = f (r, θ)
2
x' x
Example: r = a sin 2θ (Four-leafed rose) O
(2) Region or extent Find the limits for r and θ.

P
(a) Find the greatest numerical value of r, so that
the curve lies within the circle of radius r. Figure 6.22 Hypocycloid
(b) Find the values of θ for which r is imaginary. Example 6.15
This gives us an idea of the region into which Trance the curve
the curve does not extend.
rθ = a (Hyperbolic or Reciprocal Spiral).
(3) Asymptotes Find the asymptotes of the curve,
if any. Solution
(1) Symmetry On changing r to −r and θ to −θ
(4) Special points Give different values to θ. Find simultaneously, we see that the equation remains
the corresponding values of r and determine the unaltered so that the curve is symmetrical about the
points where the tangent coincides with the radius vertical line θ = π/2 through 0.
vector or perpendicular to the radius vector, i.e.,
dθ (2) Pole r is never zero for any finite real value of
points where tan φ = r = 0 or ∞.
dr θ. So, the curve does not pass through 0.
Example 6.14 (3) Asymptote As θ → 0, r → ∞. So, the curve
Trace the curve r = aθ (Spiral of Archimedes). has an asymptote.
Curve-Tracing 6-15
Figure 6.23 Hyperbolic spiral or reciprocal spiral
Writing the equation in the form

1 θ
= = f (θ)
r a Figure 6.24 Cardioid with negative x-axis as its
axis
We have f (θ) = 0 ⇒ θ = 0
1 1 cos θ = 1 or θ = 0. Hence θ = 0 is tangent to the
Also, f (θ) = ⇒ f (0) = . curve.
a a
So, the equation of the asymptote is (3) Table of values As θ increases from 0 to π, r
1 also increases from 0 to 2a as is evident from the
r sin(θ − 0) = = a ⇒ r sin θ = a corresponding values of r and θ as given.
f (0)
which is a straight line parallel to the initial line at a
distance ‘a’ from it. θ=0 π/3 π/2 2π/3 π
r=0 a/2 a 3a/2 2a
(4) Region or extent Since sin θ < θ , the value of
r for the curve is less than that for the asymptote, the
Plotting these points and joining the points by
curve lies below the asymptote.
means of a smooth curve, we get the curve as shown
(5) Table of values
in Fig. 6.24.
θ 0 π/2 2π ∞ Example 6.17

Trace the curve
r ∞ 2a/π a/2π 0
r = a(1 + cos θ) [JNTU 2003S(3)]
The shape of the curve rθ = a which is known
Solution
as hyperbolic spiral or reciprocal spiral as shown in
The equation of the curve is
Fig. 6.23
r = a(1 + cos θ) (1)
Example 6.16
Trace the curve r = a(1 − cos θ) (Cardioid with its which is in polar coordinates.
axis along the negative x-axis). (1) Symmetry On changing θ to −θ, we notice that
equation remains unaltered. So, the curve is sym-
Solution metrical about the initial line θ = 0, i.e., the x-axis.
(1) Symmetry Replace θ by −θ. We find that the Changing r to −r we get a different equation. So, the
equation of the curve remains unaltered. Hence the curve is not symmetrical about the pole (the origin).
curve has symmetry about the initial line. (2) Pole Put r = 0 in Eq. (1), we get (1 + cos θ) =
(2) Pole Put r = 0. 0 ⇒ cos θ = −1 ⇒ θ = π.
θ cos θ 1 + cos θ r= φ
a(1 + cos θ)
0° 1 2 2a π/2
π/3 1/2 3/2 3a/2 2π/3
π/2 0 1 a 3π/4
2π/3 −1/2 1/2 a/2 5π/6
π −1 0 0 π
If θ = 0, φ = π/2 and r = 2a.

Thus the tangent to the curve at the point (2a, 0)
Figure 6.25 Cardioid with its horizontal axis is at right angle to the radius vector (initial line).
The shape of the curve is as shown in Fig. 6.42.
So the curve passes through the pole and the The curve is called a cardioid because of its shape in
straight line θ = π is a tangent at the pole to the the form a heart.
curve.
(3) Determination of φ From Eq. (1), we have Example 6.18
Trace the curve r = a + b cos θ (Pascal’s Limacon).
dθ 1
tan φ = r = a(1 + cos θ) ·
dr −a sin θ Case (i) a > b

θ π θ On changing the sign of θ to −θ, the equation remains
= − cot = tan + unaltered. The curve is symmetrical with respect to
2 2 2
π θ the initial line.
⇒φ= + r a + b cos θ
2 2 tan φ = dr =
dθ
−b sin θ
The tangent at the pole will be perpendicular to θ = 0 ⇒ tan φ = ∞ φ = π/2
the initial line if θ = 0 so that r = a(1 + cos θ) = 2a θ = π ⇒ tan φ = ∞ φ = π/2
(∵ cos θ = cos 0 = 1)
∴ At the point (2a, 0), the tangent will be at right (1) Symmetry On changing θ to −θ we notice that
angle to the initial line. the equation remains unaltered. So, the curve is sym-
(4) Points of intersection with the axes If θ = 0 metrical about the initial line θ = 0, that is, the x-axis.
then r = 2a so that the point (2a, 0) is the point of Changing r to −r we get a different equation. So, the
intersection of the curve and the initial line. curve is not symmetrical about the pole (the origin)
(5) Asymptote Since for any value of θ, r is finite (2) Pole Put r = 0 in Eq. (1). We get a + b cos θ =
there is no asymptote to the curve. 0 ⇒ cos θ = −a/b
(6) Tabulation of values We tabulate the values of
r and θ of any point as when φ increases from 0 to a a
|cos θ| = − = >1 (∵ a > b)
π, r decreases from 2a to 0. b b
So, for no value of θ , r equals zero.
dθ −1 Therefore, the curve does not pass through the
Also tan φ = r = a(1 + cos θ) ·
dr a sin θ pole.
2 cos2 2θ θ
=− = cot (3) Limits The least value of r is a−b (when θ = π)
2 sin 2θ cos 2θ 2 and the greatest value of r is a + b (when θ = 0)

π θ π θ Hence the curve lies entirely within a circle of
= tan + ⇒φ= + radius (a + b).
2 2 2 2
Curve-Tracing 6-17
θ=α
θ=π θ =0
x
O
r = a+ b cos θ (a < b)
Figure 6.26 Pascal’s Limacon Figure 6.27 Pascal’s Limacon

(4) Asymptotes The curve has no asymptotes For no value of θ, r equals zero. Therefore, the
because for no value of θ, r tends to infinity curve does not pass through the pole.
(5) Special points or region As θ increases from (7) Region or extent The least value of r is a − b
0 to π, r decreases from a + b to a − b and for no and the greatest value of r is a + b for θ = 0.
value of θ, r is zero.
Case (ii) a < b
θ r φ On changing θ to −θ, the equation is unaltered. The
0 a+b −∞ curve is symmetrical with respect to the initial line.
π/2 a tan−1 a/b (1) Shape of tangent We have
π a−b +∞
r a + b cos θ
tan φ = dr =
Also, dθ
−b sin θ
dθ 1
tan φ = r = (a + b cos θ) ·
dr −b sin θ θ = 0 or π tan φ = ∞ so that φ = 2π/2. At
a + b cos θ these points, the tangents are perpendicular to the
=− initial line.
b sin θ
Hence for no value of θ, a + b cos θ = 0 and (2) Tabulation of values
therefore φ = 0 at any point.
Tan φ is infinity when θ = 0. i.e., at P = (a+b, 0). θ=0 π/3 π/2 2π/3 values of θ
Thus, the tangent is perpendicular to the initial line r = a+b a+ L2 c a− 62 between 2π/3
at (a + b, 0). The shape of the curve is as shown in and π decreas-
Fig. 6.26. The name of the curve is Pascal’s limacon. ing values
(6) Tabulation of values Tabulate the values of r
corresponding to the values of θ.
θ=0 π/3 π/2 2π/3 π values of θ
The shape of the curve when a < b is shown in
from 0 to π/3 Fig. 6.27.
r = a + b a + b2 a a − b2 a−b decreasing
values Case (iii) a = b
In this case, the equation becomes r = a(1 + cos θ)
which represents a cardioid with its axis along the
r = 0 ⇒ a + b cos θ = 0
a positive x-axis. The tracing of this curve has been
⇒ |cos θ| = − > 1 discussed in Example 6.17.
b
Example 6.19
Trace the curve r = a sin 3θ (three-leafed rose).
[JNTU 1996]
Solution The equation of the curve is
r = a sin 3θ (1) O
which is in polar coordinates.

(1) Symmetry On changing θ to (π − θ) we notice
that the equation remains unaltered. So, the curve is
symmetrical about the line θ = π/2 through the pole,
(i.e., the y-axis).
Figure 6.28 Three-leafed rose
(2) Pole With r = 0 in Eq. (1) we get θ = 0. So,
the curve passes through the pole, (i.e., origin). The lines θ = 0, π/3 and 2π/3 are tangents to
Moreover, r = 0 for θ = 0, π3 , 2π
3
, π, 4π
3
and 5π
3
. So, the curve. The shape of the curve is as shown in
π 2π
θ = 0, 3 , 3 , π, 3 and 3 are tangents to the curve
4π 5π
Fig. 6.28. The curve is known as three-leafed rose.
through the pole.
(3) Asymptotes The curve has no asymptotes θ=0 π/6 π/3 π/2 2π/3
because for no value of θ, r tends to infinity. (0 · π/6) (1 · π/6) (2 · π/6) (3 · π/6) (4 · π/6)
φ=0 π/2 0 π/2 0
(4) Region or extent Since the maximum absolute
value of sin 3θ is |sin 3θ| = 1 ⇒ sin 3θ = ±1, we
have two values r = ±a for r. Hence, the curve lies Example 6.20
entirely within a circle of radius a. Trace the curve r = a cos 3θ (Three-leafed rose).
π π 2π 5π
We observe that θ = , θ= + = , θ= Solution
6 6 3 6 (1) Symmetry About the initial line θ = 0. The
π 4π 9π 3π
+ = = are lines of symmetry for the curve is symmetrical about the initial line θ = 0
6 3 6 2
curve. Also, some special points on the curve are since changing θ into −θ leaves the equation of the
when θ increases from 0 to π/6, r is positive and curve unaltered.
increases from 0 to a and when θ increases from π/6 (2) Tangents Put r = 0. We get cos 3θ = 0
to π/3, r is positive and decreases from a to 0. Thus
π 3π 5π
we get a loop entrapped between the lines θ = 0 and ⇒ 3θ = ± , ± , ±
θ = π/3. 2 2 2
π π 5π
Again when θ increases from θ = π/3 to θ = π/2, or θ=± , ± , ±
r is negative and numerically increases from 0 to a, 6 2 6
and when θ increases from π/2 to 2π/3, r is negative These straight lines are tangents at the pole to the
and numerically decreases from a to 0. So, we obtain given curve.
a second loop entrapped between the lines θ = π/3 dr
and 2π/3, bisected by the negative of the y-axis. (3) Region or extent = −a sin 3θ; equating
dθ
Finally, when θ increases from 2π/3 to 5π/6, r is dr
= 0 we get sin 3θ = 0.
positive and increases from 0 to a. dθ
dθ 1
Further tan φ = r = a sin 3θ
dr 3a cos 3θ 3θ = 0 π 2π 3π 4π
1
= tan 3θ θ=0 π/3 2π/3 π 4π/3
3
Curve-Tracing 6-19
θ=0 π/6 π/3 π/2 2π/3 5π/6 π

(0 · π/6) (1 · π/6) (2 · π/6) (3 · π/6) (4 · π/6) (5 · π/6) (6 · π/6)
r=0 a 0 −a 0 a 0
Figure 6.30 Four-leafed rose

Figure 6.29 r = a cos 3θ three-leafed rose
(4) Tabulation of values (3) Limits The least value of cos 2θ = −1 and
the greatest value of cos 2θ is 1. The corresponding
θ=0 π/6 π/3 π/2 2π/3 π/6 π values of r are −a and a. Hence the curve lies entirely
r=a 0 −a 0 a 0 −a within the circle of radius a.
(4) Asymptotes The curve has no asymptotes
On plotting the above points and joining them by because for no value of θ, r tends to infinity.
means of smooth curve, we get the shape of the curve (5) Region or extent The values of r correspond-
as depicted in Fig. 6.29. ing to some special values of θ are as follows.
When θ decreases from 7π/4 to 0, r increases from
Example 6.21
0 to a and when θ increases from 0 to π/4, r decreases
Trace the curve r = a cos 2θ. [JNTU 2003S]
from a to 0. Thus one loop is formed between the rays
Solution The equation of the curve is θ = π/4 to π/4 which is bisected by the initial line
θ = 0.
r = a cos 2θ (1)
(θ = π/4, θ = 3π/4), (θ = 3π/4, θ = 5π/4) and
(1) Symmetry Changing θ to −θ, we notice that (θ = 5π/4, θ = 7π/4).
the equation remains unaltered. So, the curve is These leaves lie along the OY , OX and OY axes.
symmetrical about the initial line θ = 0, i.e., the The shape of the curve is as shown in Fig.6.30. The
x-axis. curve is known as 4-leafed rose.
Changing r to −r, we get a different equation. So,
the curve is not symmetrical about the pole, (i.e., the Example 6.22
origin). Trace the curve
(2) Pole Putting r = 0 we get cos 2θ = 0 ⇒ θ = r 2 = a2 cos 2θ (Lemniscate of Bernoulli).
π/4, 3π/4, 5π/4, 7π/4. So, the curve passes through [JNTU 2003S, 2003]
the pole and the lines θ = π/4 and θ = 3π/4 are
tangents to the curve. Solution The equation of the curve is
θ=0 π/4 π/2 3π/4 π 5π/4 3π/2

(0 · π/4) (1 · π/4) (2 · π/4) (3 · π/4) (4 · π/4) (5 · π/4) (6 · π/4)
r=a 0 −a 0 a 0 −a
r 2 = a2 cos 2θ (1) θ=0 π/6 π/4

r=a a/2 0
(1) Symmetry Changing θ to −θ we notice that φ = π/2 5π/6 π
the equation remains unaltered. So, the curve is
symmetrical about the initial line θ = 0, i.e., the
x-axis. the loop. By symmetry, a second loop exists along
Changing r to −r, we get the same equation. So, the negative x-axis, between the lines θ = 3π/4 and
the curve is symmetrical about the pole (the origin). θ = 5π/4.
(2) Pole Put r = 0 in Eq. (1), we get cos 2θ = 0 ⇒ When θ lies between π/4 and 3π/4, cos 2θ is neg-
θ = ±π/4. Hence the curve passes through the pole. ative and so there is no real value of r. This shows
Also, the lines θ = ±π/4 are tangents to the curve that there is no part of the curve in this region.
through the pole. Also,
(3) Limits The least and greatest values of r are 0 dθ dθ a2 cos 2θ
and a, respectively, which correspond to θ = ±π/4 tan φ = r = r2 · =
dr rda −a2 sin 2θ
and θ = 0 or π. π
Hence the curve lies entirely within a circle of = − cot 2θ = tan + 2θ
2
radius ‘a’. π
⇒ φ = + 2θ
(4) Asymptotes The curve has no asymptotes 2
because for no value of θ, r tends to infinity. when θ = 0, φ = π/2 and r = ±a.
(5) Region or points Let us consider the same spe- Differentiating Eq. (1) with respect to θ
cial values of θ (0 < θ < π/4) and the corresponding
dr
values of r as the curve is symmetrical about the 2r = a2 (−2 sin 2θ)
initial line and also the pole. dθ
When θ increases from −π/4 to π/4, r increases Hence at the points (±a, 0) the tangents are
from 0 to a and then decreases from a to 0. Thus one perpendicular to the initial line.
loop of the curve lies between the rays θ = −π/4 and The shape of the curve is as shown in Fig. 6.46.
θ = π/4, the line θ = 0 (the initial line) bisecting The curve is known as the Lemniscate of Bernoulli.
6.4 Tracing of Curves —

Parametric Coordinates
6.4.1 Procedure for Tracing a
Curve Whose Equation is in
Parametric Form
Suppose the equations of a curve is given in para-
metric coordinates as
x = f (t) and y = g(t) (6.3)
By eliminating the parameter t, we can obtain the

Cartesian form of the equation for the curve which
Figure 6.31 Lemniscate of Bernoulli can then be traced as explained earlier.
Curve-Tracing 6-21
If the parameter t cannot be eliminated easily d 2y

Similarly if < 0 then the curve will be concave
between the equations, the following procedure could dx2
be followed: downwards between the points.
(1) Symmetry (a) If x = f (t) is an even function (8) Periodicity If x and y are periodic functions of
of t and y = g(t) is an odd function, then the curve t with a common period, we need to study only the
is symmetrical about the x-axis. portion of the curve for one period.
(b) If x = f (t) is an odd function of t and y = f (t) Example 6.23
is an even function, then the curve is symmetrical Trace the curve x = a(t + sin t), y = a(1 − cos t).
about the y-axis.
(c) If x = f (t) and y = g(t) are both odd functions Solution The equations of the curve are
of t, then the curve is symmetrically placed in the
opposite quadrants. x = a(t + sin t), y = a(1 − cos t) (1)
(2) The origin Put x = 0 and get values of t. For a
particular value of t if we get y = 0, then the curve where t is a parameter.
passes through the origin. (1) Symmetry Here x is an odd function of t and y
is an even function of t. So, the curve (1) is symmetric
(3) Limits for the curve Find the greatest and least
about the y-axis.
values of x and y, if any. Thus we obtain lines parallel
to the axes between which the curve lies or does not (2) Origin When x = 0, t = 0. For this, y = 0.
lie. Hence the curve passes through the origin.
(4) Intersection with the coordinate axes (a) Find (3) Intersection with the axes (a) Put y = 0. This
the points where the curve meets the x-axis. This is implies t = 0 and x = 0 so that the curve cuts the
done by solving y = g(t) = 0 for t and obtaining the x-axis at the origin.
value of x for this value of t. (b) Put x = 0. This implies t = 0 and y = 0 so
that the curve cuts the y-axis at the origin.
(b) Similarly the points of intersection of the curve
and the y-axis are obtained. (4) Limitations and the region Since | cos t| ≤ 1
(5) Region (a) Find the region in which no part of for the values of t, y = a(1−cos t) has the least value
the curve lies by finding the imaginary values of x ‘0’ if t = 0 and the greatest value 2a if t = ±π.
and y. Hence the entire curve lies, between the lines
dy y y = 0 and y = 2a.
(b) Find the values of x, y, and φ = tan−1
dx x (5) Periodicity Since y is periodic with a period 2π,
for different values of t. The angle φ determines the the curve consists of an infinite number of congruent
position of the tangent at the point (x, y). arches, each arc being of period 2π. So, we can trace
(6) Asymptotes (a) If t → a ⇒ y → k, then the an arch between the limits t = −π and t = π.
line y = k will be an asymptote parallel to the x-axis. dy
(6) Special points (a) Since = ±∞ according
(b) If t → a ⇒ x → k, then the line x = k will dx
be an asymptote parallel to the y-axis. as t = ±π, the tangents are parallel to the coordinate
axes.
(7) Special points (a) Find the points on the curve (b) As t increases from 0 to π, x and y increase.
dy
for which = 0 or ±∞ at which the tangents are Hence the curve ascends when t moves from 0 to π
dx and descends as t changes from −π to 0. Thus, 0 is
parallel to the x-axis or the y-axis, as the case may
be. a point of minimum value.
1
d 2y d 2y 1 t dt sec2 2t
(b) Find the points on the curve for which 2 > 0. Since 2 = sec2 · = 2
dx dx 2 2 dx 2a cos2 2t
The curve will be concave upwards between the 1 t
points if there is no change of sign in between. = sec4 > 0 for all t.
4a 2
This implies that the arch of the curve is concave EXERCISE 6.1
from t = −π to t = π and the concavity is same
on the either side of the tangent at t = π and also at 1. Trace the following curves the equations of which
t = −π. These points are cusps. are in polar coordinates.
t
(7) Region (i) y = a(1 − cos t) = 2a sin2 ⇒ (a) r = 2a cos θ ⇒ r 2 = 2ar cos θ
* 2 ⇒ x2 + y2 = 2ax ⇒ (x − a)2 + y2 = a2 . The
t y t
sin = when y is negative. sin has no real circle centre (a, 0) r = a.
2 2a 2
value. Thus, no portion of the curve lies below the
y
x-axis.
(ii) We now tabulate the values of t and x, y and
dy
at some special points
dx
x
O (a, 0)
dx t
= a(1 + cos t) = 2a cos2
dt 2
dy t t dy t
= a sin t = 2a sin cos ⇒ = tan = tan φ Figure 6.33 Circle touching the y-axis
dt 2 2 dx 2
at (0, 0)
(b) r 2 = a2 sec 2θ or r 2 cos 2θ = a2
⇒ r 2 cos2 θ = r 2 sin2 θ = a2
θ x y dy
dx
φ = 2t ⇒ x2 − y2 = a2 (rectangular hyperbola).
0 0 0 0 0
π/2 a π2 + 1 a 1 π/4 y
π aπ 2a +∞ π/2
2π 2aπ
0 0 π
−π/2 −a π2 + 1 0 −1 −π/4
−π −aπ 2a −∞ −π/2
x
O
(8) Asymptotes No asymptotes exist for the curve.

The shape of the curve is shown in Fig. 6.32. The
curve is called a ‘cycloid’. It is the locus of a point
on a circle of radius ‘a’ rolling on a straight line. Figure 6.34 Rectangular hyperbola
(c) r = 4a cot θ cosec θ
⇒ r 2 sin2 θ = 4ar cos θ ⇒ y2 = 4ax.
x
O S (a, 0)
Figure 6.35 Parabola with its axis along the

Figure 6.32 Cycloid (Inverted)
x-axis and touching the y-axis at O
Curve-Tracing 6-23
(d) r(1 − cos θ) = 2a. Parabola with vertex at (b) x = a cos3 t, y = b sin3 t
(a, π) and the origin is at its focus S. 2/3 2/3
⇒ cos2 t = ax sin2 t = by
2/3 y 2/3
y ⇒ ax + b = 1(Hypocycloid)
y
B(0, b)
2a
A x
(a, π) O
x
A(–a, 0) O A(a, 0)
B'(0, –b)
Figure 6.36 Parabola with its axis along Figure 6.39 Hypocycloid
the x-axis and vertex at (−a, 0)
1 − t2 2t
(e) r = aeθ cot α (a > 1) (Equiangular spiral). (c) x = a , y=b
1+t 2 1 + t2
x 1−t 2
y 2t
y ⇒ = , =
a 1+t 2 b 1 + t2
x 2 y 2
⇒ a + b = 1(Ellipse)
α
α y
x
O θ=0
B (0, b)
Figure 6.37 Equiangular spiral x

A' (-a, 0) O A (a, 0)
2. Trace the following curves, the equations of which B' (0, -b)
are in parametric coordinates.
Figure 6.40 Ellipse
(a) x = a cos3 t, y = a sin3 t
2/3 2/3 (d) x = t 2 , y = t − 13 t 3
⇒ cos2 t = ax sin2 t = ay
2/3 y 2/3 3−t 2
3−x 2
⇒ ax + a =1 ⇒y=t 3
⇒ y2 = x 3
⇒x 2/3
+y 2/3
=a 2/3
⇒ 9y = x(3 − x)
2 2
B (0, a)
o x
A'(–a, 0) O A (a, 0) 3
B' (0, –a)
Figure 6.41
Figure 6.38 Astroid
(e) x = a(1−sin t), y = a(1−cos t) (Cycloid) are two asymptotes to the hyperbola
x2 y2
2
− 2 = 1· (2)
a b
y Solution Eq. (2) can*be written in the form
b a2
2aπ y =± x 1− 2 (3)
a x
2a
x a
0 A Now as x → ∞, → 0 and the R.H.S. of (3)
t=0 t = 2π x
b
Figure 6.42 Cycloid becomes ± x. Therefore the asymptotes to the curve
a
a t (2) are
(f) x = a cos t + log tan2 , b
2 2 y = ± x.
y = a sin t (Tractrix) a
Asymptotes parallel to the axes

y
B (0, –a) Let the equation of the curve, arranged in decreasing
c A powers of x be
x
O
f x, y = a0 xn + a1 y + b1 xn−1
B' (0, a)
+(a2 y2 +b2 y+c2)xn−2 + · · · = 0 (4)
Figure 6.43 Tractrix
If a0 = 0 and y is chosen so that a1 y + b1 = 0,
6.5 Asymptotes then the coefficients of two highest powers of x in
6.5.1 Plane Curve Asymptotes (1) vanish and hence two of its roots are infinite.
Therefore
Algebraic plane curves
A plane curve which has an equation in cartesian a1 y + b1 = 0 or y = −b1 /a1 (5)
coordinates of the form f (x, y) = 0, where f is a
is an asymptote of the curve Eq. (1) and is parallel
polynomial in x and y is called as an algebraic curve. to the x-axis.
If f has degree n, then the curve is an algebraic curve Again, if a0 , a1 , b1 are all zero and y is so chosen
of degree n. that
Asymptotes a 2 y 2 + b2 y + c 2 = 0 (6)
For a plane curve, an asymptote is a line which has then three roots of Eq. (1) become infinite. Therefore
the property that the distance from a point P on the the two lines given by Eq. (3) are asymptotes of the
curve to the line approaches zero as the distance from curve Eq. (1) and these are parallel to the x-axis.
P to the origin increases without bound and P is on The asymptotes parallel to the y-axis for a curve
a suitable piece of the curve. can be determined in a similar manner.
In other words, an asymptote is a straight line Rules to find the asymptotes parallel to the x-axis/y-
which cuts the curve in two points at an infinite dis- axis are
tance from the origin and yet is not itself wholly at
(a) x-axis: Equate to zero the coefficient of the high-
infinity.
est power of x in the equation if it is not a mere
Example 6.24 constant.
The straight lines (b) y-axis: Equate to zero the coefficient of the high-
b est power of y in the equation if it is not a mere
y=± x (1) constant.
a
Curve-Tracing 6-25
Example 6.25 The straight line (6.4) is an asymptote to the curve

Find the asymptotes of the curve (6.3) if it cuts the curve at a point at infinity, i.e., if
Eq. (6.6) has two infinite roots. The conditions for
x2 y2 − x2 y − xy2 + x + y + 1 = 0. (1)
this are
Solution The highest power of x is x2 . Its φn (m) = 0 and cφn + φn−1 (m) = 0. (6.8)
coefficient is (y2 − y).
The asymptotes parallel to the x-axis are given by 6.5.3 Working Rule for Oblique
y − y = 0,
2
i.e., y = 0, y = 1 Asymptotes
(1) Put x = 1, y = m in the highest degree terms of
Next, the highest power of y is y2 . Its coefficient is the equation of the curve φn (m).
(x − x).
2
The asymptotes parallel to the y-axis are given by Solve φn (m) = 0 for m getting m1 , m2 , · · · mn .
(2) Form φn−1 (m) by putting x = 1, y = m in the
x2 − x = 0, i.e., x = 0, x = 1. (n−1)th degree terms of the equation of the curve.
6.5.2 Oblique (Inclined) (3) Find c1 , c2 , · · · cn by putting m = m1 , m2 , · · · mn in

Asymptotes turn in the formula
(Asymptotes not parallel to the axes) c = −φn−1 (m)/φn (m)
Let the equation of the curve be written in the form
If the above expression reduces to oo form then find
xn φn y/x + xn−1 φn−1 y/x c from the condition

+ xn−2 φn−2 y/x + · · · = 0 (6.4) c2
φ (m) + cφn−1 (m) + φn−2 (m) = 0
2 2! n
where φr y/x = a0 + a1 y/x + a2 y/x
r In this case, there will be two parallel asymptotes.
+ · · · + ar y/x (6.5)
(4) Substitute the value of m and c in y = mx + c in
an expression of degree r in the variable (y/x) with
turn to obtain the oblique asymptotes
different constants ar s for different functions φr .
To find the points of intersection of Eq. (6.3) with y = m r x + cr (r = 1, 2, · · · n).
the straight line Example 6.26
y = mx + c (6.6) Find the asymptotes of
y c
put = m + in Eq. (6.3) y3 −2xy2 −x2 y+2x3 +3y2 −7xy+2x2 +2y+2x+1 = 0.
x x
We obtain Solution Put x = 1, y = m in the 3rd degree
c c term. We get
xn φn m + + xn−1 φn−1 m + + xn−2
x x φ3 (m) = m3 − 2m2 − m + 2 = 0 ⇒ m = 1, −1, 2
c
φn−2 m + + ··· = 0 (6.7) Put x = 1, y = m in the 2nd degree terms, we get
x
Expanding each of the φ-functions in Taylor’s φ2 (m) = 3m2 − 7m + 2
series, we obtain φ2 (m)
Now c = −
c c2 φ3 (m)
xn φn (m) + φn (m) + φ (m) + · · · ⎧
2!x2 n ⎪ −1 for m = 1
3m − 7m + 2 ⎨
x
c
2
+xn−1 φn−1 (m) + φn−1 (m) =− 2 = −2 for m = −1

x 3m − 4m − 1 ⎪⎩
0 for m = 2
c2
+ 2 φ(n−1) (m) + · · · + · · · The asymptotes are y = x−1, y = −x−2, y = 2x.
2!x
Applications of Integration 7
7.1 Introduction Since v(1) = −1, we have
Suppose we know the acceleration of a particle and −1
want to know the velocity at a particular instant. Here −1 = v(1) = + C1 ⇒ C1 = 0
1
the derivative f is known. The problem is to find the
corresponding function f . In this context, we use the Further,
term antiderivative which we define below. ds
v(t) =
dt
−1 1
7.2 Antiderivative ⇒ s(t) = v(t) dt = 2
dt =
t t
Definition A function F is called an antiderivative 1
of a given function f on an interval I if F (x) = f (x) ∴ s(4) =
for all x in I . 4
We know that if we differentiate F (x) = 12 x2 or ∴ The particle is at 1
4
when t = 4.
F (x) = 12 x2 +C, where C is an arbitrary constant, we
get the same function F (x) = f (x) = x, therefore
7.2.1 Area as an Antiderivative
F (x) = 12 x2 + C is an antiderivative of F (x) =
f (x) = x. We call F (x) = 12 x2 + C an integral of If f is a continuous function such that f (x) ≥ 0 for all
f (x) = x. x on a closed interval [a, b], then the area bounded
This is represented as by the curve y = f (x), the x-axis and the vertical
lines x = a and x = t, viewed as a function of t, is

1 2 an antiderivative of f (t) on [a, b].
x dx = x + C. Let A(t) be area of the region bounded by the curve
2
y = f (x), the x-axis, and the vertical lines x = a,
Modelling Problem x = t for a ≤ t ≤ b as shown in Fig. 1.
The Motion of a particle: A particle moves along the The area under the curve y = f (x) between x = t
x-axis in such a way that its acceleration is a(t) = t23 and x = t + h < b, where h is small, is
for t > 0. If the particle is at x = 5 when t = 1 and A(t + h) − A(t)
has velocity v = −1 at this time, find x when t = 4. Lt = f (t)
h→0 h
Solution ⇒ A (t) = f (t)
dv b
a(t) = This implies that f (t) dt = A(b) − A(a)= Area
dt
2 −1 a
⇒ v(t) = a(t) dt = dt = 2 + C1 under the curve between the x = a and x = b.
t3 t
7.2.2 Area as a Limit of a Sum 7.2.3 Riemann Sums

Let f be a non-negative function which is contin- We have seen that we can find the area under the
uous on [a, b], partition the interval into n equal curve y = f (x), where f is continuous and f is non-
b−a negative, on the closed interval [a, b] using the sums
sub-intervals each of width x = . For r =
n by following the steps:
1, 2, 3, . . . n, the r sub interval is [a+(r −1) x, a+
th
r x ], and the approximated rectangle is constructed 1. Partition the interval into n equal sub-intervals
with width x and height f (a + r x) equal to the b−a
each of width x = .
height of y = f (x) above the right end point of the n
sub-interval. 2. Evaluate f at the right end point a + r x of the
Adding the areas of n-rectangles we get Sn = r th sub-interval, for r = 1, 2, 3, . . . n.
$n
f (a + r x) then the area under the curve x = a
r=1 3. Form the approximate sum of the areas of n-
and x = b is given by, rectangles denoted by
Lt Sn = f (x) dx
n→∞ %
n
Sn = f (a + r x)
r=1
)
f(x
y = 4. The area under the curve y = f (x) is given by
$n
A = Lt f (a + r x) x, if the limit exists.
n→∞ r=1
The above process contains the essentials of integra-

A(t) tion. But then, there is no need to take partition points
equally spaced or insist on evaluating f at right end
points only. The approximate sums that occur in inte-
gration are called Riemann sums. In the following, we
o a t b give step-by-step description of how Riemann sums
are formed.
Figure 7.1 Let f be a bounded function on a closed interval
[a, b].
Step 1: Divide [a, b] into n sub-intervals choosing
a set of points P = {a = x0 , x1 , x2 , . . . xn = b} such
y = f(x) that xr−1 < xr for, r = 1, 2, 3, . . . n, which is called a
partition. The largest sub-interval of P is called the
f(t + h) norm of P denoted by P. Thus
f(t) f(ξ)
P = max { xr }.
r=1, 2, 3,...
Step 2: Let xr∗ ∈ [ xr−1 , xr ].

$n
Step 3: Form Rn = Lt f (xr∗ ) xr
n→∞ r=1
o t ξ t+h b Rn is called the Riemann sum associated with f
and partition P
$n
If Lt f (xr∗ ) xr of this exists for all parti-
Figure 7.2 A(t + h) − A(t) is approximated by the P→∞ r=1
area of a rectangle tions of [a, b], then this limit is called the definite
Applications of Integration 7-3
integral of f from a to b, which are called the limits 7.3.1 Cartesian Coordinates
b Suppose that y = f (x) is the equation of the curve C
of integration and we denote it by I = f (x) dx.
a where f is continuously differentiable function of x.
Let P(x, y) and Q(x+δx, y+δy) be two neighbouring

7.2.4 Properties points on C. Let s be the length of the AP from a
Let f , g be integrable on [a, b], fixed point A on C and PQ = δs. Draw PM , QN
perpendicular to the x-axis and PR⊥r QN . (Refer
b Fig. 7.3)
1. Linearity: [af (x) + bg(x)]dx =
a
When δx is sufficiently small, we can assume that
b b PQR is a right-angled triangle so that the element
a f (x) dx + b g(x)dx of length of the curve is
a a
2. Dominance Rule: If f (x) ≤ g(x)on [a, b], then (δs)2 = (δx)2 + (δy)2
b b ⇒ (ds)2 = (dx)2 + (dy)2 (7.1)
f (x) dx ≤ g(x) dx
a a
Dividing throughout by (dx)2 , we get
b 2 2
3. Sub-division rule: f (x) dx = ds dy
a = 1+
c b dx dx
f (x) dx + f (x)dx 2
a c ds dy
or = 1+ (7.2)
a b dx dx
4. Reversal of limits: f (x) dx = − f (x) dx
b a Integrating Eq. (7.2) with respect to x between
x = x1 and x = x2 , we get the length of the arc of the
7.3 Length of a Curve curve C between the points (x1 , y1 ) and (x2 , y2 ):
We will now consider some geometrical applications
of integration viz., finding the lengths of curves, x2 2
dy
surface areas and volumes of solids of revolution in s= 1+ dx. (7.3)
dx
cartesian, parametric and polar coordinates. x1
Calculation of the length of an arc of a curve
is called rectification. Let us first consider the 7.3.2 Parametric Coordinates
rectification of a curve whose equation is given in
If x = x(t), y = y(t) where x(t), y(t) are continu-
cartesian coordinates.
ously differentiable functions of ‘t’ we get
2 2 2
ds dx dy
= +
dt dt dt

t2 2 2
dx dy
⇒ s= + dt. (7.4)
dt dt
t1
7.3.3 Polar Coordinates

Let r = f (θ) be the equation of the curve C in polar
coordinates where f is a continuously differentiable
function of θ. Suppose P(r, θ) and Q(r + δr, θ + δθ)
Figure 7.3 are two neighbouring points on C. Let PQ = δs.
Table 7.1 Formulae for Arc Lengths in Different Coordinate Systems

Equation of the curve Element of length of the curve Arc length
Cartesian coordinates
x2 + dy 2
(i) y = f (x) ds = (dx)2 + (dy)2 s= 1+ dx
dx
x1
*
y2 2
(ii) x = f (y) ds = (dx)2 + (dy)2 s= 1+ dx
dy
dy
y1
Parametric coordinates
+ dy 2 t2 + dx 2 dy 2
dx 2
(i) x = x(t) y = y(t) ds = dt
+ dt
dt s= dt
+ dt
dt
t1
Polar coordinates
θ2 + dr 2
(i) r = f (θ) ds = (dr)2 + r 2 (dθ)2 s= r2 + dθ
dθ
θ1
r2 + dθ 2
(ii) θ = f (r) ds = (dr)2 + r 2 (dθ)2 s= 1 + r2 dr
dr
r1
2
ds dr 2
= r +
2
dθ dθ
2
ds dr
⇒ = r2 + (7.6)
dθ dθ
Integrating with respect to ‘θ’ from θ = θ1 to θ =

θ2 , we get the arc length
2
θ2
dr
s= r +
2 dθ. (7.7)
Figure 7.4 dθ
θ1

Draw PR ⊥r OQ. PR = PR= rδθ. Formulae for arc lengths in different coordinate
systems are entered in Table 7.1.
QR = OQ − OR = r + δr − r = δr
Example 7.1
From the right-angled triangle PQR Find the perimeter of the semi-circle x2 + y2 = a2 ,
y ≥ 0, bounded by the diameter y = 0.
(δs)2 = (rδθ)2 + (δr)2
Solution The perimeter consists of two parts
(i) diameter AB = 2a (1)

∴ Element of length of the curve in polar coordi-
nates is and the (ii) semi-circular arc
(ds)2 = r 2 (dθ)2 + (dr)2 (7.5) x2 + y2 = a2 y≥0 (2)

Dividing by (dθ)2 , we get which is 2 × length of the BC .
Differentiating Eq. (2) with respect to x, we get
dy dy x
2x + 2y =0 ⇒ =−
dx dx y
,

a 2 12
Length of dy
= 2× BC= 2 1+ dx
the arc ACB dx Figure 7.5
0
a 12
x2
=2 1+ 2 dx
y , 4a *
Length of y2
0
= 1+ dy
a the arc 4a2
x2 + y2 0
=2 dx
y 4a
0 1
= 4a2 + y2 dy
a 2a
dx
= 2a √ 0
a − x2
2 1 y
0 = 4a2 + y2
2a 2
∵ y= a2 − x2 [from Eq. (2)] 4a2 4a
x a + log y + 4a2 + y2
= 2a sin−1 2 0
a 0 1 - √
π = 2a · 2 5a + 2a2
= 2a. = πa 2a
2 √ )
4a + 2 5a
Perimeter of the semi-circle log
2a
√ √
= Diameter + Semi-circular arc = a 2 5 + log 2 + 5 .
= 2a + πa = (π + 2)a. Example 7.3

√ that the length of the curve x (a −x ) = 8a y
2 2 2 2 2
Show
Example 7.2
is 2πa.
Find the length of the arc of the parabola y2 = 4ax
cut off by the straight line y = x. Solution The equation of the curve is

Solution x2 a2 − x2 = 8a2 y2 (1)
Parabola y2 = 4ax (1) Since the equation contains even powers of x and
Straight line y = x (2) y, the curve is symmetrical about both the axes.
Solving Eqs. (1) and (2), we get the points of

intersection
O = (0, 0); P = (4a, 4a) (3)
Differentiating Eq. (1) with respect to y, we get

dx dx y
4a = 2y ⇒ = (4)
dy dy 2a
Figure 7.6
The length of the curve is 4 times the length of measured from the vertex to both extremities of the

OPA in the first quadrant. latus-rectum. [JNTU 2000S] (or)
Eq. (1) can be written as Prove that the length of the arc of the parabola
y2 = 4ax cut off by its latus-rectum is
√ √
x2 (a2 − x2 ) 2a 2 + log( 2 + 1) . [JNTU 2004S(3)]
y2 =
8a2
√
x a2 − x2 Solution The equation of the parabola is
⇒y= √
2 2a y2 = 4ax (1)
dy 1
= √ 1. a2 − x2
dx 2 2a The curve is symmetrical about the x-axis. The
latus-rectum is the ordinate through S(a, 0).
1
+x · √ (−2x) Differentiating Eq. (1) with respect to x, we get
2 a2 − x2
2 dy dy 2a
1 (a − x2 ) − x2 2y = 4a ⇒ = (2)
= √ √ dx dx y
2 2a a2 − x 2
1 a − 2x2
2
Length of the required arc
= √ ·√ (2)
2 2a a2 − x2 a 2
2 2 dy
2 =2 1+ dx
dy a − 2x2 dx
1+ = 1+ 2 2 0
dx 8a a − x2
a
8a4 − 8a2 x2 + a4 − 4a2 x2 + 4x4 4a 2
= =2 1 + 2 dx
8a2 a2 − x2 y
0
9a4 − 12a2 x2 + 4x4 a *
= a
8a2 a2 − x2 =2 1 + dx (3)
2 x
3a2 − 2x2 0
= √ √ (3)
2 2a a2 − x2 Put x = a tan θ; dx = 2a tan θ sec2 θdθ
2

The limits are
Length of the curve = 4 × Length of OPA
x = 0 ⇒ θ = 0;
a π
3a − 2x
2 2
x=a ⇒ θ=
=4 √ √ dx 4
2 2a a2 − x2 π/4
0
a 2 = 2 cosecθ · 2a tan θ sec2 θdθ
4 2 a − x 2 + a2
= √ √ dx 0
2 2a a2 − x2
0 π/4
√ √ a
2 2a x a2 − x2 a2 = 4a sec3 θdθ
−1 x
= + sin 0
a 2 2 a
0 π/4
√ −1 x a = 2a [sec θ tan θ + log (sec θ + tan θ)]0
+ 2a sin √ √
√ 2
a 0 = 2a 2 + log 2+1 . (4)
2 2a π √ π √
= · + 2a = 2πa. (4) Note
a 2 2 2 y1 2
dx dx
Example 7.4 We can find and use s = 1+ dy.
dy dy
Find the length of the arc of the parabola y2 = 4ax y2

Example 7.5 8a 9 a 3/2
ex−1 = 1+ · −1
Find the length of the arc of the curve y = log ex+1
27 4a 4

from x = 1 to x = 2. [JNTU 2004, 2006, 2007(4)] 8a 125 61a
= −1 = . (3)
27 64 216
Solution
Example 7.7
y = log(ex − 1) − log(ex + 1) Find the length of a loop of the curve
dy ex ex 9ay2 = x(x − 3a)2 .
⇒ y1 = = x−1 − x+1
dx e e
2ex 2 Solution The equation of the curve is
= 2x−1 = x = cosech x
e e − e−x 9ay2 = x(x − 3a)2 (1)
Length of the arc
2+ 2 The curve meets the x-axis at x = 0, x = 3a. It is
= 1 + y1 dx =
2
1 + cosech2 xdx symmetric about the x-axis since y occurs to an even
1 1
2 power. From Eq. (1), we have
=
2
coth xdx = log sinh x 1 √
x
1 y = √ (x − 3a) (2)
ex − e−x 2 2
3 a
= log 1
= log(ex − e−x ) 1
2 Differentiating Eq. (2), we get
(e2 − e−2 ) √
= log = log(e + e−1 ) dy 1 x
e − e−1 = √ (x − 3a) + √ · 1
= log(e + 1/e). dx 6 ax 3 a
1
Example 7.6 = √ (x − 3a + 2x)
6 ax
Find the length of the arc of the semi-cubical parabola
a a 1
y3 = ax2 from the origin to , . = √ (x − a)
8 4 2 ax
Solution The equation of the curve is 2
Length of the loop = OPA
y3 = ax2 (1) 2
3a
1 dy
or x = √ y3/2 (2) =2 1+ dx
a dx
0
Differentiating Eq.(2) with respect to y, we get
dx 1 3 1/2
= √ ·y
dy a2
2
y2
dx
Arc length = 1+ dy
dy
y1
a/4*
9
= 1+ y dy
4a
0
3/2 a/4
4a 2 9
= · 1+ y
9 3 4a Figure 7.7
0
3a
1 √ a
= 2· √ x + √ dx
2 a x
0
3/2
3a
2 x
= √ +a·x 1/2
a 3
√ 0

2 3a 3a √
= √ + a · 3a
a 3
1 √ √
= √ 4a 3a = 4 3a. Figure 7.8
a
Integrating Eq. (3) with respect to ‘t’ between the
Example 7.8 limits 0 and 2π, we get
Show that the curve x2 (4−x√ ) = 32y can be formed
2 2
2π
into a circle with radius = 2. t t 2π
s = 2a sin dt = 2a −2 cos
2 2 0
0
Solution
= −4a (cos π − 1) = 8a.
x2 (a2 − x2 ) = 8a2 y2
Example 7.10
Proceeding as in Example 7.3, we obtain the length Find the perimeter of the loop of the curve 3ay2 =
√ x(x − a)2 . [JNTU 1995, 1998S, 1999S, 2003S(2)]
of the total curve = 2πa
√
If a = 2, this becomes
√ 2 2π = Circumference of Solution The equation of the given curve is
a circle with radius = 2. 3ay2 = x(x − a)2 (1)
Example 7.9
The required loop is symmetrical about the x-axis
Find the length of one arc of the cycloid and lies between x = 0 and x = a.
x = a(t − sin t) y = a(1 − cos t). Solving Eq. (1) for y
√
x
y = √ (x − a) (2)
Solution The equations of the cycloid are 3a
x = a(t − sin t), y = a(1 − cos t) (1) Differentiating Eq. (2) with respect to ‘x’, we get

dy 1 3 1 1 1 1 3x − a
Putting y = 0, we get t = 0, 2π. Differentiating =√ x −a x
2 2 = √ √
dx 3a 2 2 2 3a x
Eq. (1) with respect to ‘t’, we get
Length of the loop = 2× arc OPA
dx dy 2
= a(1 − cos t), = a sin t (2) a
dt dt dy
=2 1+ dx
dx
Element of length of the curve 0

a
2 2 (3x − a)2
dx dy =2 1+ dx
ds = + dt x
0
dt dt
+ a
1 1 −1
= a (1 − cos t)2 + sin2 t = √ 3x 2 + ax 2 dx
√ √ 3a
= 2a 1 − cos t 0
√ √ t t 2 3
1 a 4a
= 2a 2 sin = 2a sin (3) = √ x 2 + ax 2 = √ .
2 2 3a 0 3
2 2
Example 7.11 t
dx dy
Find the perimeter of the cardioid r = a(1 − cos θ). s= + dt
dt dt
[JNTU 1998, 2002] 0
t +
Solution The equation of the cardioid is = a2 t 2 (cos2 t + sin2 t)dt
0
r = a(1 − cos θ) (1)
t
t2
=a tdt = a . (3)
The curve is symmetrical about the initial line (x- 2
axis); r varies from r = 0 to r = 2a and θ varies 0
from θ = 0 to θ = π. EXERCISE 7.1

Differentiating Eq. (1) with respect to ‘θ’, we get
1. Find the length of the arc of the curve y = log sec x
dr π
= a sin θ (2) from x = 0 to x = .
dθ √ 3
Ans: log(2 + 3)
Length of the entire curve = 2× Length of the arc 2. Find the length of the arc of the curve y =
above the x-axis ex − 1
log x from x = 1 to x = 2.
e +1
θ2 2 [JNTU 2004(4)]
dr
=2 r2 + dθ 1
dθ Ans: log e +
θ1 e
π + 3. Find the length of the arc of the parabola x2 = 4ay
=2 a2 (1 − cos θ) + 2
a2 sin θdθ
2
extending from the vertex to one extremity of the
0 latus-rectum.
√ √
π π Ans: a[ 2 + log( 2 + 1)]
θ − cos 2θ
= 4a sin dθ = 4a 1 4. Find the length of the arc of the curve x =
2 2 π
0
0
et sin t, y = et cos t from t = 0 to .
π 2
= 8a − cos + cos 0 = 8a. (3) [JNTU 1987, 2004(1)]
2 √ π
Ans: 2 e 2 − 1
Example 7.12
Find the length of the arc of the curve x = a(cos t + 5. Rectify the curve r = a(1 + cos θ).
t sin t), y = a(sin t − t cos t) from t = 0 to any point [JNTU 1996S, 1993, 1990]
‘t’ on the curve. Prove that the arc of the upper half the cardioid is
π
bisected at θ = . [JNTU 2004]
Solution Differentiating the given equations with 3
Ans: 8a
respect to ‘t’, we have
6. Find the length of the curve y2 = x3 from the
dx origin to the point (1, 1). [JNTU 1989]
= a(− sin t + t cos t + sin t) = at cos t (1) 2 √
dt Ans: 13 13 − 8
dy 27
= a(cos t + t sin t − cos t) = at sin t (2) 2 2
dt x 3 y 3
7. Show that the length of the curve + =
a b
Length of the arc from t = 0 to any point ‘t’ on 4 a2 + ab + b2
the curve 1 is .
ab
8. Find the length of the spiral r = eαθ from the pole Example 7.14
to thepoint (r, θ).
If we revolve a right-angled triangle about one of its
√
1 + α2 legs, it generates a right-circular cone.
Ans: r
α
9. Find the perimeter of the curve

r =√a(cos θ + sin θ) (0 ≤ θ ≤ π).
Ans: 2πa
10. Find the length of the loop of the curve

1
x = t2, y = t − t3.
√ 3
Ans: 4 3
Figure 7.10
11. Show that the curve x = a(t − sin t), y = a(1 −
2π Example 7.15
cos t) is divided in the ratio 1:3 at t = . If we revolve a rectangle about one of its sides, it
3
generates a right-circular cylinder.
12. Find the length
of the lemniscate r 2 = a2 cos 2θ.
2
√ 1 1·3 2 1 2
Ans: 2πa 1 + +
2 2·4 2
2 2
1·3·5 1
+ + ···
2·4·6 2
Figure 7.11
7.4 Volumes of Solids of
Revolution
The volume of a solid of revolution can be obtained
Imagine a plane region R being revolved about a as shown in Example 7.16.
fixed line L lying in the plane. It generates a solid of
revolution. The line L about which the region R is Example 7.16
revolved is called the axis of revolution. Find the volume got by the revolution of the cate-
nary y = a cosh x/a bounded by the x-axis, and the
Example 7.13 ordinates x = ±a.
If we revolve a semi-circular region R about its x
diameter, it generates a sphere. Solution Equation of catenary: y = a cosh
a
a a
x
Volume = πy2 dx = π a2 cosh2 dx
−a −a a

πa2 a 2x
= 1 + cos wh dx
2 0 a
a
sinh 2xa
= πa x +
2
2/a
0
sinh 2
Figure 7.9 = πa3 1 + .
2
A plastic or card-board hand-held fan rotated Example 7.17
about its handle generates a sphere. A sphere of radius ‘a’ units is divided into two parts
by a plane distant a2 from the centre. Show that the by rotating the upper half of the curve when θ varies
ratio of the volumes of the two parts is 5:27. from θ = 0 to π
[JNTU 2003S, 2006(4)] π
2 3
∴ Volume = πr sin θdθ
Solution When a semi-circle of radius ‘a’ units 0 3
π
revolves about its diameter a sphere of radius a 2π
= a3 (1 + cos θ)3 sin θdθ
units is formed, which is divided into two spherical 3 0
segments major and minor by a plane a2 units from
2πa3 π
its centre. The equation of the generating circle is =− (1 + cos θ)3 d(1 + cos θ)
3
x2 + y2 = a2 . 0
2πa3 (1 + cos θ)4 π
Volume of the minor spherical segment =− 0
3 4

a a πa 3
πa 3
V1 = πy2 dx = π (a2 − x2 )dx = (0 − 16) = .
a/2 a/2 6 3
a
1
= π a2 x − x 3 Example 7.19
3 a/2 Determine the volume of the solid generated revolv-

a 3
1 3 a3 ing the Limacon r = a+b cos θ (a, b) about the initial
=π a − 3
− a − line. [JNTU 2004S, 2005(4)]
2 3 8
5πa3
= Solution Please refer to Ex. 7.18 for tracing of
24 curve p.
Volume of the major spherical segment
2π π 3
Volume = r sin θdθ
3 0
a/2 a/2
π
V2 = πy2 dx = π (a2 − x2 )dx −2π
a −a = (a + b cos t)3 d(cos θ)
a/2 3 0
1 2π a+b
= π a2 x − x 3 = a + b cos θ 3 a−b
3 −a 3b
3 π
a 1 a3 = (a + b)2 − (a − b)2
=π +a −
3
+ a3 6b
2 3 8
4πa 2
27a π
3
= (a + b2 ).
= 3
24
The ratio of two volumes 7.4.1 Cartesian Coordinates

Consider the plane area ABCD bounded above by the
5πa3 27πa3 curve y = f (x) and below by the x-axis and between
V1 : V2 = : = 5 : 27.
24 24 the ordinates x = a and x = b.
Example 7.18 When the plane area ABCD is revolved about the
Find the volume of solid of revolution generated by x-axis, a solid of revolution is generated.
revolution of cardioid r = a(1 + cos θ) about the Figure 7.12 depicts one-quarter of it in the first
initial line θ = 0. octant.
[JNTU 2003, 2003S, 2006S (Set 2)] The volume of an element circular disc of radius y
and thickness dx is πy2 dx. Integrating these elements,
Solution The given curve is symmetrical about we get the volume V of the solid of revolution of the
the initial line θ = 0 and the volume is generated area bounded by the above curves as
Corollary When the plane region is bounded by

two curves y = y1 (x) and y = y2 (x) y1 (x) ≤ y2 (x)
then the volume of the solid of revolution is
b

V =π y22 − y12 dx.
a
When the axis of revolution L is not a part of the

b
boundary of the region R, V = 2πx (y) dx.
Figure 7.12 x=a
b
V =π y2 dx. (7.8) 7.4.2 Polar coordinates
a Volume of the solid of revolution generated by
If the equation of the curve is x = g(y), then the revolving the region R bounded by the curve r = f (θ)
volume of the solid of revolution of the area ABCD the initial line θ = 0 and the radii vectors θ = θ1 and
bounded by the curve x = g(y), the y-axis and the θ = θ2 about the initial line θ = 0 is
abscissae y = c and y = d, is given by
d θ2
2π
V = π x2 dy. (7.9) V = r 3 sin θdθ
3
c θ1
and about the line perpendicular to the initial line is
θ2
2π
V = r 3 cos θdθ.
3
θ1
Figure 7.14
Formulae for volumes in different coordinate
systems are entered in Table 7.2.
Example 7.20
Find the volume of the solid of revolution of the
region OAP bounded by (i) y = x2 and x = a,
Figure 7.13 (ii) y = x2 and y = b.
Table 7.2 Formulae for Volume in Different Coordinate Systems

Region R Axis Volume of the solid generated
b
(i) y = f (x), the x-axis x-axis V =π y2 dx
a
and x = a,x = b
d
(ii) x = g(y), the y-axis y-axis V =π x2 dy
c
and y = c, y = d
t2
(iii) x = x(t), y = y(t) the x-axis V =π y2 dx
dt
dt
t1
x-axis, ordinates where
t = t1 , t = t2
t2
(iv) x = x(t), y = y(t) the y-axis V =π x2 dy
dt
dt
t1
y-axis abscissae where
t = t1 , t = t2
Polar coordinates
θ2
(v) r = f (θ), the initial line θ = 0 The initial line θ = 0 V = 2π
3
r 3 sin θdθ
θ1
radii vectors θ = θ1 , θ = θ2
θ2
(vi) r = f (θ), the line θ = π/2 ⊥r The line θ = π/2 ⊥r V = 2π
3
r 3 cos θdθ
θ1
to the initial line radii vectors to the initial line
θ = θ1 , θ = θ2
[since y = x2 ⇒ dy = 2xdx]
P
b
x4 π 4
= 2π · = b.
4 0 2
Example 7.21
Determine the volume of the solid of revolution of
A
the region bounded by y2 = 4x and x = 4 about the
line x = 4.
Figure 7.15
Solution OPAQO is the plane region bounded by

Solution
the parabola y2 = 4x and the vertical line x = 4.
a a A typical horizontal element PQ of length (4 −
2 2 x) and width dy when rotated about the line x = 4
(i) V = π y dx = π
2
x dx
generates a circular disc of volume π(4 − x)2 dy.
0 0
a Here y varies from −4 to 4. So, the required
x5 π
=π = a5 volume is
5 0 5
b b 4 4 2
y2
(ii) V = π x dy = π x2 2xdx
2
V =π (4 − x)2 dy = π 4− dy
4
0 0 −4 −4
b
y3 b3
= 2πa 2
y− 2 = 2πa 2
b− 2
3b 0 3b
4πa2 b
= .
3
Example 7.23
Find the volume of the solid generated by the revo-
t3
lution of the loop of the curve x = t 2 , y = t −
3
about the x-axis.
Figure 7.16
Solution The curve is symmetrical
√ about the x-
axis. The limits for t are t = 0 and 3.
√ √
3 3 2
4 t3
y4 V =π y2 dx = 2π t− tdt
= 2π 16 + − 2y2 dy 3
16 0 0
0 √
4 3
y5 2 2 5 1 7
= 2π 16y + − y3 = 2π t − t + t dt
3
80 3 3 9
0 0
= 2π 64 +
64 128
− √3
t4 2 t6 1 t8

5 3
= 2π − +
1 2 4 36 98 0
= 128π 1 − + 3

5 3 9 2 27 1 81 3π
= 2π − · + · = .
128π 1024 4 3 6 9 8 4
= (15 + 3 − 10) = π.
15 15
Example 7.24
Find the volume of the solid generated by revolving
Example 7.22
one arc of the cycloid
x2 y2 x = a(t + sin t), y = a (1 + cos t). [JNTU 1999]
the ellipse 2 + 2 = 1 about its minor axis.
a b
[JNTU 1993]
Solution The equation of the ellipse is

x2 y2
2
+ 2 =1 (1)
a b
Its minor axis is x = 0 (y-axis); y varies from

y = 0 to y = b. Figure 7.17
Volume of the solid of revolution when Eq. (1) is
rotated about the y-axis is Solution The equations of the curve are
x = a(t + sin t), y = a(1 + cos t) (1)

b b
y2 t = 0 ⇒ x = 0, y = 2a (2)
V = 2π x2 dy = 2π a2 1− 2 dy
b t = π ⇒ x = aπ, y = 0
0 0
π
V = 2π y2 dx
0
π
= 2π a2 (1 + cos t)2 · a (1 + cos t) dt
0
π
t
= 2πa · 2 3 3
cos6 dt
2
0
t 1 Figure 7.18
u= ; du = dt
2 2
π/2 0
= 16πa · cos6 udu,
3 V =− πy2 dx
6 a
5 3 1π 0
= 32πa3 · · · = 5a2 π2 . (3) 2a3
6 4 22 =π −x − 2ax − 2a +
2 2
dx
a−x
−a
Example 7.25 0
x3
Find the volume of the solid generated by revolving = π − − ax2 − 2a2 x − 2a3 log(a − x)
the cardioid r = a(1 + cos θ) about the initial line. 3 −a
3
[JNTU 1993, 1995] 2a
=π + 2a3 log 2
3
Solution
2πa3
= 1 + 3 log 2 .
π 3
2
V = π r 3 sin θdθ (1) Example 7.27
3
0 Find the volume of the solid generated by the revolu-
π tion of the astroid x2/3 + y2/3 = a2/3 about the x-axis.
2πa3
=− (1 + cos θ)3 d(cos θ) [JNTU 1990]
3
0
π Solution In parametric coordinates, the equa-
2πa 3
(1 + cos θ)4 8 3 tions of the curve are
=− = πa . (2)
3 4 0 3 x = a cos3 θ, y = a sin3 θ (1)
π
Example 7.26 The limits are θ = 0 and θ =
Find the volume of the solid generated by revolution 2
of the loop of the curve y2 (a − x) = x2 (a + x) about π/2
dx
the x-axis. V = 2π y2 · · dθ (2)
dθ
0
Solution The equation of the curve is
π/2

a+x 2 = 2π a2 sin6 θ −3a cos2 θ sin θ dθ
y2 = x (1)
a−x 0
π/2
The loop is between x = −a and x = 0 and is = 6πa 3
sin7 θ cos2 θdθ
symmetric about the x-axis. 0
y Solution
(0, a)
x
(-a, 0) (a, 0)
(0, -a)

θ2
2π
V = r 3 cos θdθ
3
θ1
6 4 2 1 32πa3
= 6πa · · · · =
3
. (3) π/2
9 7 5 3 105 2π
= a3 (1 + cos θ)3 cos θdθ
3
−π/2
Example 7.28
π/2
Find the volume of the solid generated by revolving 4πa3 4
the curve r = 2a cos θ about the initial line. = cos θ+3 cos3 θ+3 cos2 θ+cos θ dθ
3
0

π 4πa 3.1 π
3
2 1π
Solution The limits are θ = 0 and θ = . = +3· +3· +1
2 3 4.2 2 3 22
The required volume is πa 3
= (5π + 16) .
4
θ2
2π
V = r 3 sin θdθ
3 Example 7.30
θ1
π/2 the curve xy = a(x − b) about its asymptote.
2π
= (2a) 3
cos3 θ sin θdθ
3 Solution The equation of the curve is
0

π/2 x−b
16πa3 y=a (1)
=− cos3 θd(cos θ) x
3
0
π/2 Writing the equation in the form
16πa 3
cos4 θ 4πa3
=− = . −ab
3 4 0 3 y−a= (2)
x
Example 7.29 we see that as y → a, x → ∞ so that y = a is an
Find the volume of the solid generated by the region asymptote of the curve.
of the cardioid r = a(1 + cos θ) included between Let P(x, y) be any point on Eq. (1). If M is the foot
π π π
θ = − and θ = when rotated about the θ = . of the perpendicular from P on the asymptote y = a
2 2 2
the x-axis and the curve x2 = y3 is symmetrical about

the y-axis.
Solving the equations we get (0, 0) and (1, 1) as
the points of intersection. The required volume
1
2 4/3
V =π y2 − y1 dx = π
2
x − x3 dx
1
x=0

3 x4 3 1 5π
= π x7/3 − =π − = .
Figure 7.21
7 4 0 7 4 28
EXERCISE 7.2
1. Find in each of the cases (a–e) in Table 7.3, the
PM = a − y (3) volume of solid of revolution generated by the
region bounded by the given curve about
and x varies from x = b to ∞.
the indicated line L as axis.
Volume of the solid of revolution of Eq. (1) about
16 2
y=a Ans: (a) ab
∞ ∞ 15
V = π (PM ) dx = π (a − y)2 dx
2 (b) 2π2
b b 1024
(c) π
∞ ∞ 15
a(x − b) 2 1 √
=π a− dx = πa b
2 2
dx 256 2
x x2 (d) π
b

b 15

1 ∞ h
= πa b −
2 2
= πa2 b. (4) (e) πh a −
2
x b 3
Example 7.31 2. Find the volume of the solid of revolution gen-
Show that the volume of the solid obtained by revolv- erated by revolving the hypocycloid (x/a)2/3 +
ing the plane region included between the curves (y/b)2/3 =1 about (a) the x-axis, (b) the y-axis,
5π and (c) Deduce the corresponding results for the
y2 = x3 and x2 = y3 about the x-axis is .
28 astroid x2/3 + y2/3 = a2/3 .
32
Solution The curve y2 = x3 is symmetrical about Ans: (a) πab2
105
Table 7.3 Volume of Solid Revolution of the Region Bounded by the Given Curve
About the Line L as Axis
S. no. Curve Boundaries Line L
(1) (2) (3) (4)
(a) ay2 = x2 (a − x) Loop of the curve in (2) y=b
(b) y4 = x(4 − x) Loop of the curve in (2) x-axis
(c) Parabola y2 = 4x x = 4, x = 0 x=4
(d) Parabola
y + x2 + 3x − 6 = 0 x+y =3 x+y =3
(e) Arc of the circle
x2 + y2 = a2 x =a−h x-axis
x=a
32πa2 b
(b) 105
slant height PQ and the radii y and y + δy. The area
(c) 32
105
πa3 · of this element of surface is 2yδs. Divide C between
a and b into n parts. The total surface area of the
3. Find the volume of the solid generated by revolv- solid of revolution is obtained by summing up these
ing one arc of the cycloid x = a(1 − sin t), y = areas and allowing n to tend to infinity and δs → 0,
a(1 − cos t) about its base. if the limit exists. Area of the surface of the solid
Ans: 5π 2 a3 of revolution generated by revolving the AB of the
4. Find the volume of the solid generated by revolv- curve C is given as follows.
ing y = ex sin x between x = 0 and x = π about
the x-axis.

Ans: π8 e2π − 1
5. Find the volume of the solid formed when the

curve r 2 = a2 cos θ is rotated about its line of
symmetry.
8πa2
Ans: 15 Figure 7.23
6. The cardioid r = (1 − cos θ) revolves about the About any line L

initial line. Find the volume of the solid generated. S = 2π (PM ) ds
[JNTU 1989]
8πa3 where PM = ⊥r distance from P on C to the axis of
Ans: 3
revolution L.
7. Find the volume of the solid generated by revolv- Formulae for surface areas in different coordinate
ing r 2 = a2 cos 2θ about the initial line. systems are entered in Table 7.4.
[JNTU 2003]
√
πa3 3
√ √ Example 7.32
Ans: 12 3 log( 2 + 1) − 2 Find the surface area of the right circular cone formed
by revolving a right-angled triangle about a side
7.5 Surface Area of a Solid of containing the right-angle.
Revolution
Let P(x, y) and Q(x+δx, y+δy) be two neighbouring Solution Let the sides of the right-angled triangle
points on a plane curve C whose equation is y = f (x). OPM be OP = l, OM = h, PM = r.
When the part of the curve between x = a and x = b The curve y = x tan α from x = 0 to h is rotated
is rotated about the x-axis a solid of revolution is about the x-axis (y = 0). Surface area of the solid
generated. We now find its surface area. generated by the revolution
2
h
dy
y S = 2π y 1 + dx
y = f (x)
B dx
0
Q h
P δs
= 2π x tan α 1 + tan2 α dx
s 0 h
A x2 r l h2
y y + δy = 2π tan α sec α · = 2π · · · = πrl.
O
2 0 h h 2
M N
x δx
Example 7.33
Figure 7.22 Find the surface area generated by the area of the
x
When the curve C is rotated, the chord PQ (= δs curve y = c cosh (catenary) between (0, c) and
c
approximately) generates a frustum of a cone with (x, y) about the x-axis. [JNTU 1989, 1994, 1994S]
Table 7.4 Formulae for Surface Areas in Different Coordinate Systems

Equation of the curve Axis of revolution Surface area

x=b +
x=b 2
(i) y = f (x) x-axis 2πyds = 2π y 1 + dy
dx
dx
x=a x=a
*

y=d
y=d 2
(ii) x = g(y) y-axis 2πxds = 2π x 1 + dy
dx
dy
y=c y=c
t2 + 2
2
(i) x = x(t), y = y(t) x-axis 2πy(t) dx
dt
+ dy
dt
dt
t1
t2 + dy 2
dx 2
(ii) x = x(t), y = y(t) y-axis 2πx(t) dt
+ dt
dt
t1
Polar coordinates
θ2 θ2 + dy 2
r = f (θ) θ=0 ds
2πy dθ dθ = 2π r sin θ r 2 + dθ dθ
θ1 θ1
(initial line) replace r by f (θ)
θ = π/2
θ2 θ2 + dy 2
(⊥r to the initial line ds
2πx dθ dθ = 2π r cos θ r 2 + dθ
θ1 θ1
through the pole) [replace r by f (θ)]
x
x
= 2πc cosh2 dx
c
0
x
2x
= πc 1 + cosh dx
c
0

c 2x
= πc x + sinh . (3)
2 c
Figure 7.24
Solution Example 7.34
x
y = c cosh (1) Find the area of the surface of a sphere of radius a.
c
Differentiating Eq. (1) with respect to ‘x’, we get Solution The equation of the upper half of the
circle
dy x x2 + y2 = a2 , y ≥ 0 (1)
= sinh (2)
dx c
2
x Differentiating Eq. (1) with respect to ‘x’, we get
dy
S = 2π y 1 + dx dy dy x
dx 2x + 2y =0 ⇒ =− (2)
0
dx dx y
x *
x x 2
= 2πc cosh 1 + sinh2 dx ds dy
c c ∴ = 1+
0 dx dx
* a
= 1 + x2 /y2 = a/y [by Eq.(1)] (3) 4πbe x a 2 a2 −1 xe
= − x + 2 sin
2
a 2 e 2e a
The required curved surface
0
1/2
a a 2πbe a 1 − e
2 2
a 2
= + 2 sin−1 e
S = 2π yds = 2π adx a e e
−1
x=−a x=−a
= 2πab 1 − e2 + sin e /e
= 2πa [x]a−a = 4πa2 . (4)
= 2 area of the ellipse
Example 7.35
× 1 − e2 + sin−1 e /e · (4)
Prove that the surface of the prolate spheroid formed
by the revolution of the ellipse of eccentricity e about
its major axis = 2 (area of the ellipse).
Example 7.36
Solution Prolate spheroid is generated by revolv- Find the surface of the solid formed by revolution
ing an ellipse about its major axis. about the x-axis of the loop of the curve x = t 2 , y =
Let the equation of the ellipse be 1
t − t3.
x2 y2 3
2
+ 2 =1 (1)
a b Solution The parametric equations of the curve
Differentiating Eq. (1) with respect to ‘x’, we get are
1
x = t2, y = t − t3 (1)
dy b2 x 3
=− 2 (2)
dx ay dx dy
∴ = 2t, = 1 − t2
2 dt dt
ds dy b4 x 2 2
= 1+ = 1+ 4 2 ds dx 2 dy
dx dx ay ⇒ = +
2 dt dt dt
a4 1 − e 2 x 2 + 2
= 1+ ∵ b2 = a2 (1 − e2 ) = (2t)2 + t − 12 = 1 + t 2 (2)
a ·y
4 2
3 √
4 2
4 1 − e2 x 2 For the loop, t varies from t = 0 to t = 3. The
5
= 1+ 2 required surface area
b
a2 − x2 √ √

a2 3 3
ds
1 − e2 x 2 a2 − e2 x2 S = 2π yds = 2π y dt
= 1+ = (3) dt
a −x
2 2 a2 − x2 t=0
√
t=0
3
x varies from −a to a. Hence the required area 1
= 2π t − t 3 1 + t 2 dt
a a 3
b 2 ds t=0
S = 2π yds = 2π a − x2 dx √
a dx 3
x=−a x=−a 2π
= 3t + 2t 3 − t 5 dt
a 3
4πb a2 − e 2 x 2 0
= a2 − x 2 dx √
a a2 − x 2 2π 3 2 1 4 1 6 3
0 = t + t − t
a * 3 2 2 6 0
4πbe a 2
= − x2 dx 2π 9 9 9
a e = + − = 3π. (3)
0 3 2 2 2

Find the surface of the solid of revolution about the Find the area of the solid generated by the rotating of
y-axis the part of the curve ay2 = x3 from x = 0 to the loop of the curve r 2 = a2 cos 2θ about the initial
x = 4a which is above the x-axis. line (lemniscate of Bernoulli). [JNTU 2000]
Solution The equation of the curve is (semi- Solution The equation of the curve is
cubical parabola, Fig. 6.5(c) p. 3 of Ch. 6).
r 2 = a2 cos 2θ (1)
ay = x
2 3
(y ≥ 0) (1)
There are two loops of the curve. Differentiating
Differentiating Eq. (1) with respect to ‘x’, we get Eq. (1) with respect to ‘θ’, we get
dy dy 3x2 dr
2ay = 3x2 or = 2r = −2a2 sin 2θ
dx dx 2ay dθ
2 dr a2 sin 2θ
ds dy or =−
= 1+ dθ

r
dx dx 2
dr dr
9x4 = r2 +
= 1+ 2 2 dθ dθ
4a y
* a4 sin2 2θ
9x = a2 cos 2θ + 2
= 1+ ∵ ay2 = x3 (2) a cos 2θ
4a a
= √ (2)
The required area of the surface of revolution is cos 2θ

x=4a 4a The required surface area = 2 (surface generated
ds
S = 2πxds = 2π x dx by one loop)
dx
x=0 x=0
π/4 π/4
4a ds
2π √ = 2 × 2π yds = 4π r sin θ dθ
= √ x 4a + 9x dx dθ
2 a 0 0
x=0
√ π/4 √
2 10a a
π t 2 − 4a 2t = 4π a cos 2θ sin θ √ dθ
= √ t. dt cos 2θ
a √
a 9 0
2 a
π/4
π/4
Put 4a + 9x = t 2 ⇒ 9dx =√2tdt = 4πa 2
sin θdθ = 4πa2 [− cos θ]0
The limits are x = 0 ⇒ t = 2 √a 0

x = 4a ⇒ t = 2 10a 1
= 4πa 2
1− √ . (3)
√ 2
2 10a
2π
= √ t 4 dt − 4at 2 dt Example 7.39
81 a √
2 a Find the surface of the solid formed by the revolution
2√10a of the cardioid r = a(1 + cos θ) about the initial line.
2π 1 5 4 3
= √ t − at
81 a 5 3 √ Solution The equation of the cardioid is
2 a
128 2 √
= πa 125 10 + 1 . (3) r = a(1 + cos θ) (1)
1215
The curve is symmetric about the initial line. For Differentiating with respect to ‘t’, we get
the upper portion of the curve θ varies from 0 to π.
Differentiating Eq. (1) with respect to ‘θ’, we get dx dy
= −3a cos2 t sin t, (2)
dt dt
dr = 3a sin t cos t
2
= −a sin θ (2)
dθ
2
2 ds dx 2 dy
ds dr ∴ = +
= r +
2 dt dt dt
dθ dθ = 3a sin t cos t (3)
+ The required surface area

= a (1 + cos θ)2 + sin2 θ
= a 2 (1 + cos θ) π/2 π/2
ds
= 2a cos (θ/2) (3) S = 2 2πyds = 4π y dt (4)
dt
t=0 t=0
The required surface area π/2

= 4π a sin3 t · 3a sin t · cos t dt
B π
ds 0
S = 2π yds = 2π r sin θ dθ
dθ π/2
θ=0
A
= 12πa 2
sin4 t cos t dt
π
0
= 2π a (1 + cos θ) · sin θ · 2a cos (θ/2) dθ
1 π/2 12πa2
θ=0 = 12πa2 sin5 t 0 = . (5)
π 5 5
= 2πa2 2 cos2 (θ/2) · 2 sin (θ/2) cos (θ/2)
Example 7.41
0 Find the area of the surface generated by a circular
×2 cos (θ/2) dθ arc, subtending an angle 2α at the centre, when it
π revolves about its chord.
= 16πa 2
cos4 (θ/2) sin (θ/2) dθ (4)
0 Solution Let the equation of the circle be
π
θ θ
−32πa2 cos4 d cos
2 2
0
π
1 θ 32πa2
= −32πa2 cos5 = . (5)
5 2 0 5
Example 7.40
Find the surface of the solid generated by the rev-
olution of the astroid x2/3 + y2/3 = a2/3 about the
x-axis.
Solution The parametric equations of the curve

are
x = a cos3 t, y = a sin3 t (1) Figure 7.25
x 2 + y 2 = a2 (1) Solution The curve

is symmetrical about the y-
Then OA = radius = a, COA = α (2) −x x
axis since y = a cosh = a cosh . The x-axis
a a
(semi-vertical angle)
is called the directrix for the curve called catenary.
and OC = a cos α (3) The curve meets the y-axis at A = (0, a); it does not
meet the x-axis.
where ADB is the arc of the circle which subtends an x
angle 2α at the centre O. D is the mid point of the y = a cosh (1)
a
arc ADB.
dy x
Let P(a cos θ, a sin θ) be any point on the curve. ⇒ = sinh (2)
dx a
From P draw PM perpendicular to the chord ACB
about which the arc revolves. If s = AP is the arc, then
2
ds dy
Now OK = a cos θ = 1+
dx dx
and PM = OK − OC = a(cos θ − cos α) (4) *
x x
= 1 + sinh2 = cosh
a a
The parametric equations of the circle are
The required surface area
a
x = a cos θ, y = a sin θ (5) ds
= 2πy · dx
dx dy dx
∴ = −a sin θ, = a cos θ
0
a
dθ dθ x x
= 2πa cosh · cosh dx
2 a a
ds dx 2 dy a0
= + 2x
dθ dt dθ = πa cosh + 1 dx
0 a
= a2 sin2 θ + a2 cos2 θ = a (6) a
a 2x
= πa sinh +x
2 a
θ varies from 0 to α. 0
sinh 2
Hence the required area of the surface of revolution = πa2 +1 .
2
S = 2 × Surface formed by revolution
of arc AD about AB EXERCISE 7.3
α
= 2 2πPMds 1. Find the area of the surface generated by revolv-
ing the loop of 9y2 = x(x −3)2 about the x-axis.
θ=0
Ans: 3π
α
= 4π a (cos θ − cos α) adθ (7) 2. The part of the parabola y2 = 4ax cut off by the
0 latus-rectum revolves about the tangent of the
= 4πa2 [sin θ − θ cos α]a0 vertex. Find the area of the curved surface of the
solid so generated. [JNTU 1998]
= 4πa2 (sin α − α cos α) . (8) √ √
Ans: πa 3 2 − log 2 + 1
2
Example 7.42
3. Find the area of the surface of revolution formed
Find the surface area generated by rotating the arc of
x by revolving the loop of the curve 9ay2 =
the catenary y = a cosh from x = 0 to x = a about x (3a − x) 2
a √ about the x-axis.
the x-axis. Ans: 22πa 3/45
4. Evaluate the area of the surface generated by the 7. Show that the surface area of the solid gener-
revolution of the cycloid x = a(θ − sin θ), y = ated by revolving the lemniscate r 2 = a2 cos 2θ
a(1 − cos θ) about the x-axis. about the tangent at the pole is 4πa2 .
64 2
Ans: πa 8. Find the area of the surface rotated about the
3
line θ = π/2 of the arc of the cardioid r =
5. Find the curved surface of a hemisphere of a(1 +√cos θ) included between [−π/2, π/2].
radius ‘a’. Ans: 48 2πa2 /5
Ans: 2πa2
9. Prove that the surface of the solid generated by
a
6. Find the area of the surface of revolution formed the revolution of the tractrix x = a cos t +
by revolving the curve r = 2a cos θ about the 2
t
initial line. log tan2 , y = a sin t about its asymptote is
2
Ans: 4πa2 equal to the surface of a sphere of radius a.
Multiple Integrals
8
8.1 Multiple Integrals sub-regions, Ar , r = 1, 2, · · · , n. To simplify the
notation, we write Ar for the sub-region and also
8.1.1 Introduction for its area.
We know from integral calculus, the meaning of a Suppose that the function is continuous and
definite integral. Let y = f (x) be a continuous and bounded in R. We can then form lower sums with the
bounded function on [a, b]. Divide [a, b] into n sub- infimum mr in Ar and upper sums with the supre-
intervals Ir = [xr−1 , xr ], r = 1, 2, · · · , n where a = mum Mr in Ar . If the sub-division of R is refined,
x0 < x1 < x2 < · · · < xr−1 < xr < · · · < xn = b and then as n → ∞ and | Ar | → 0 the sequence of
xr = xr − xr−1 . If ξr is any point in Ir , then form lower sums, the sequence of upper sums and any
the sum of the products $n
sequence of intermediate sums f (ξr , ηr ) Ar tend
r=1
%
n
to the same limit whatever the intermediate point
sn = f (ξr ) xr (8.1) (ξr , ηr ) is chosen in Ar . The integral of the func-
r=1
tion z = f (x, y) over the region R is defined to be
If this sum tends to a finite limit as n → this common limit and is called the double integral
∞ and | xr | → 0 for all choices of ξr in Ir then because there are two variables of integration.
we say that f is integrable in [a, b] and denote it by %

f x, y dAr = n→∞
lim mr Ar
b Ar →0
R
f (x) dx (8.2) %
= n→∞
lim Mr Ar (8.3)
a
Ar →0
b The double integral exists also if z = f (x, y) is

If f (x) > 0 in [a, b], then f (x) dx represents the
a bounded and piece-wise continuous in R. If it exists
area under the curve y = f (x) and above the x-axis the function is said to be integrable over R.
and between the ordinates x = a and x = b.
Geometrical interpretation of the double
8.1.2 Double Integral integral
We will now see that double and triple integrals The simple definite integral may be regarded as the
involving two and three variables respectively are area of a plane region below a curve. Similarly the
natural generalisations of this concept. The inter- double integral of a continuous function of two vari-
val [a, b] of integration is now replaced
by a plane ables may be interpreted as the volume below a sur-
region R on which a function z = f x, y is defined, face z = f (x, y) (Fig. 8.1) provided that z = f (x, y)
in the case of a double integral. R is divided into takes only positive values in R. Ar are elements
the graph of function y = y1 (x) and ADB given

by the graph of y = y2 (x). Similarly C, D sepa-
rate the boundary curve into CAD and CBD given
by x = x1 (y) and x = x2 (y), respectively. For fixed
x = ξr , y1 (ξr ) and y2 (ξr ) are end points of an interval
y1 (ξr ) ≤ y ≤ y2 (ξr ) over which the function f (ξr )
of the single variable must be integrated:
y
2 (ξr )

ϕ (ξr ) = f ξr , y dy
y1 (ξr )
is for fixed x = ξr a constant which depends on x, i.e.,

a function ϕ (x) of x on the interval a ≤ x ≤ b. With
suitable assumptions about the boundary curve of
R, ϕ(x) is a continuous function of x and is therefore
Figure 8.1 Geometrical interpretation of double
integrable over the interval [a, b]. Similarly,
integral
x
2 (ηr )
of area in the xy-plane which after the limiting pro-
cess are denoted by dxdy in cartesian coordinates ψ (ηr ) = f (x, ηr )dx
and by rdθdr in polar coordinates. Every product x1 (ηr )
mr Ar is the volume of a cylinder with base area
Ar and height mr . Likewise for Mr Ar now the is for fixed y = ηr a constant which depends on y
cylinder height being Mr . and as a continuous function ψ y on the interval
For the volume V below the surface z = f (x, y), c < y < d is integrable over this interval.
we have It can be shown that both repeated integrations
lead to the same value, which is equal to the value of
%
n %
n
mr Ar ≤ V ≤ Mr Ar (8.4) the double integral. This agrees with the geometrical
r=1 r=1 idea that φ(x) are areas of plane sections parallel
to the yz-plane and ψ(y) areas of the plane sections
Refinement of the subdivision leads to mono-
tonically increasing sequence of lower sums and
monotonically decreasing sequence of upper sums
and both sequences have the same limit because the y2(x)
y=d
sequence of differences of corresponding upper and
lower sums tends to zero. y2(ξr)
Calculation of the double integral
A double integral may be calculated by making two

successive integrations over each variable in turn. y1(x)
y1(ξr)
Suppose that the region R of integration has a y=c
simple boundary which meets the rectangle
O x=a x=b
a ≤ x ≤ b, c≤y≤d (8.5)
at the points A, B, C, D. The points A, B separate the
boundary curve of R into two arcs ACB which is Figure 8.2
Multiple Integrals 8-3
parallel to the xz-plane of the solid body lying below and x = b = 2. The value of the integral is
the surface z = f (x, y) and above the plane region R.
b 2
⎛ y (x) ⎞
x=b 2 F (x) dx = 4 − x − x2 /2 dx

f (x, y) A = ⎝ f x, y dy⎠dx a 0
2
R x=a y1 (x) = 4x − x2 /2 − x3 /6 0
⎛ ⎞
y=d x2 (y) 4 14
⎜ ⎟ = 8−2− =
= ⎝ f x, y dx⎠dy. 3 3
2 3−x
y=c x1 (y) 14
∴ x + y dydx = .
3
0 1
Example 8.1

Evaluate x + y dA where R is the region Example 8.2

R Evaluate xydA, the region R of the integration
between the lines x = 0, y = 1 and x + y = 3.
R
is enclosed by the curves (x − 1)2 = 2y and y = 2.
Solution For fixed x, the y-integration goes from
the constant limit y1 = 1 up to the variable limit
Solution Solving the equations
y2 = 3 − x.
(x − 1)2 = 2y (1)
3−x
1 2 3−x y=2 (2)
F (x) = x + y dy = xy + y
2 1
1
we get P1 (−1, 2), P2 (3, 2) as the points of intersec-
(3 − x)2 1 tion. The boundary of R is given by
= x (3 − x) + − x+
2 2
4−x−x 2 y1 = 2, y2 = (x − 1)2 /2 or x =1± 2y
=
2
The calculation is simpler if the x-integration is
This function F(x) must now be integrated in the performed first.
x-direction. The limits of integration are x = a = 0 √
2y
1+ 1+√2y
x2
F y = xydx = y
2 √
√ 1− 2y
1− 2y
y=2
P1 (–1, 2) P2 (3, 2)
o x
(–1, 0) (1, 0)


y 2 2 b 1
= 1+ 2y − 1− 2y 1
2 F (x) dx = 1 − 3x + 6x2 − 4x3 dx
3
= 2 2y3 x=a x=0
1
y=d √ 3/2
2 1 3 2
= x − x + 2x − x
3 4
Now F y dy = 2 2 y dy 3 2
0

y=c 0 1 3 2 1
= 1− ·1 +2·1 −1 =
3 4
√ 2 2 32 3 2 6
= 2 2 y5/2 0 =
5 5 2 1
∴ x + y2 dxdy = .
32 6
∴ xydxdy = . R
5
R Example 8.4
Example 8.3
Evaluate y2 dxdy where R is the region bounded
Evaluate x + y dxdy in the positive quadrant
2 2 R
by the parabolas y2 = 4x and x2 = 4y.
R
for which x + y ≤ 1. [JNTU 1996] Solution Equations of the parabolas
y2 = 4x (1)
Solution The region R is the triangle OAB for
which x + y ≤ 1. The boundary for R : x = 0, y = x2 = 4y (2)
0, x + y ≤ 1 y varies from 0 to 1 − x and x varies Solving Eqs. (1) and (2) we get the points of
from 0 to 1 intersection

y=1−x
2
F (x) = x + y2 dy O = (0, 0), P = (4, 4)
√
y=0
1−x The limits for y are y = x2 /4 to 2 x
1 The limits for x are x = 0 to x = 4
= x2 y + y3 √
3 0 x
y=2
1 1 3 2√x
= x (1 − x) + (1 − x)3
2 F (x) = y2 dy = y x2 /4
3 3
y=x2 /4
1−x 2
= 3x + 1 − 2x + x2 1 1
3 = 8x3/2 − x6
1 3 64
= 1 − 3x + 6x2 − 4x3
3

b x=4 x=b 1
1 1 6 1 3/2 4 3
F (x) dx = 8x − x dx
3/2
F (x) dx = x + x − x dx
5/2
3 64 3 3
x=a x=0 x=a 0
x=4 1
1 2 1 1 1 7 2 7/2 1 2 5/2 4 1 4
= · 8 · x5/2 − · · x = x + · x − · x
3 5 3 64 7 x=0 7 3 5 3 4 0
8 2 1 1 1 2 1 2 1
= · · 25 − · 3 · · 47 = + · −
3 5 3 4 7 7 3 5 3

44 2 1 30 + 14 − 35 3
= − = = .
3 5 7 105 35
44 9 Example 8.7
= ·
3 5·7 1 1
768 dxdy
= . Evaluate √ .
35 1 − x2 1 − y2
0 0
Example 8.5
1 x Solution Since the limits are constant, we can
x+y evaluate the integrals in any order
Evaluate e dydx. [JNTU 2000]
⎛ ⎞
0 0 1 1 1
dx −1 1 dy
⎝ √ ⎠ dy = sin x 0
Solution ⎛ ⎞
1 x 1 1−x 2 1−y 2 1 − y2
0 0 0
⎝e x
e dy⎠ dx =
y
ex (ex − 1)dx 1
π dy
0 y=0 0
=
1 2 1 − y2
e2x 0
= − ex π −1 1
2 0 = sin y 0
e2 1 2
= − e1 + π π π2
2 2 = × = .
1 2 2 4
= (e − 1) .2
2 Example 8.8
√
Example 8.6 1 1+x2
√
1 x dydx
Evaluate .
Evaluate x + y dxdy.
2 2
1 + x2 + y2
0 0
0 x
Solution Here the limits of the inner integral are
Solution The limits of the inner integral are func- functions of x. They are the limits for y. The outer
tions of x. These are limits for y. The outer integral integral has to be evaluated with respect to x between
has constant limits 0 and 1 for x. x = 0 and 1.
√
x ⎛ √ ⎞

F (x) = x2 + y2 dy 1 y= 1+x 2
⎜ dy ⎟
⎜ ⎟dx
y=x ⎝ √ 2 ⎠
√ x=0 y=0 1+x 2 +y 2
1 3 x
= x y+ y 2
√
3 1 1+x2
x 1 y
1 4 = √ tan−1 √ dx
= x5/2 + x3/2 − x3 1 + x2 1 + x2 0
3 3 x=0
1 Example 8.11
1 π
= √ dx Evaluate (x2 + y2 )dxdy in the positive quadrant
1+x 4
2
for which x + y ≤ 1.
x=0
π 1
= sinh−1 x 0 Solution
4
π π √ x=1 y=1−x
= sinh−1 1 or log 1 + 2 . (x2 + y2 )dxdy = dx (x2 + y2 )dy
4 4
R x=0 y=0
Example 8.9
x=1 y=1−x
1
= x y + y3
2
dx
Evaluate ydxdy where R is bounded by the x=0 3 y=0
x=1
R 1
parabolas y = 4x and x = 4y.
2 2
[JNTU 1995] = x2 − x3 + (1 − x)3 dx
x=0 3
3 4
x=1
Solution x x 1
= − − (1 − x)4
⎛ √ ⎞ 3 4 2 x=0
4 2 x 4
⎜ ⎟ 1 2 2√x 1 1 1
⎝ ydy⎠ dx = y x2 /4 dx = − −0+
2 3 4 12
0 x2 /4 0 1
= .
4 6
1 1 4
= 4x − x
2 16
0
4 y
1 x5
= 2x2 −
2 16.5 0 B (0, 1)

1 45
= 2.42 −
2 16.5 Q
48
= . R
5
x=0 x+y=1
Example 8.10

Evaluate xydxdy over the positive quadrant of
0 P A (1, 0) x
x +y =a .
2 2 2 y=0
Figure 8.6
Solution√ The region is bounded by x = 0, y = 0, Example 8.12

and y = a2 − x2
Evaluate (x2 + y2 )dxdy over the area bounded
⎛ √ ⎞
a a2 −x2 a √ x2 y2
⎜ ⎟ 1 2 a2 −x2 by the ellipse + = 1.
⎜x ydy ⎟ dx = x y dx a2 b2
⎝ ⎠ 2 0
0 y=0 0 x2 y2
Solution Solving + 2 = 1 for y, we get
a √ a2 b
1
= x a2 − x2 y = ± ba a2 − x2 .
2 So the region of integration is
0
a b 2 b 2
1 2 x2 x4 a4 R : −a ≤ x ≤ a, − a − x2 ≤ y ≤ a − x2
= a − = . a a
2 2 4 0 8
Since the integrand is even in x and y a b

dydx
(2) Evaluate .
yx
I = (x2 + y2 )dxdy 1 1
R
⎡ √ ⎤ Ans: log a log b
a ba a2 −x2 √
=4 ⎣ (x2 + y2 )dy⎦ dx a a −y
2 2
x=0 y=0
(3) Evaluate a2 − x2 − y2 dxdy.
√
a y= ba a2 −x2 0 0
1
= 4· x2 y + y3 dx Ans: πa3 /6
x=0 3 y=0
a 2
b 2 (4) Evaluate x + y dxdy over the area
=4 x2 · a − x2
x=0 a
x y 2 2
1 b3 2 bounded by the ellipse 2 + 2 = 1.
+ 3 (a − x )
2 3/2
dx a b
3a πab 2
Ans: a + b2
b 4 b3 4
= 4 I1 + · 3 I2
a 3 a
a (5) Evaluate e2x+3y dxdy over the triangle
where I1 = x2 a2 − x2 dx,
bounded by x = 0, y = 0 and x + y = 1.
0
a 1
I2 = (a2 − x2 )3/2 dx Ans: (e − 1)2 (2e + 1)
0
6
b πa4 4b3 3πa4 1 x +
= 4· · + 3·
a 16 3a 16 (6) Find the value of x/y · dydx.
πa3 b πb3 a πab 2
= + = (a + b2 ) 0 x2
4 4 4 Ans: 1/5
π/2
I1 = a4 sin2 t cos2 tdt 1 ey
dydx
0 (7) Evaluate .
1 1 π xy
= a4 · · · ; −1 1
4 2 2 Ans: 2
x = a sin t, dx = a cos t dt
1 y2 +1
when x = 0, t = 0
(8) Evaluate x2 ydxdy.
πa4
when x = a, t = π/2 = 0 0
16 Ans: 67/120

π/2
I2 = a 4
cos t · dt
4 (9) Evaluate xy x + y dxdy over the area
0
between y = x2 and y = x.
3 1 π 3πa4
=a · · · = 4
. Ans: 3/56 (Hamirpur 1994S, Karnataka 1988,
4 2 2 16 Poona)
EXERCISE 8.1
(10) Evaluate xydxdy where A is the domain
1 1 A
(1) Evaluate dx ey/x dy. bounded by the x-axis, the ordinate x = 2a and
0 0
the curve x2 = 2ay.
Ans: 1
(e − 1) Ans: a4 /3 (Marathwada 1993, A.M.I.E. 1990)
2
8.2 Change of the Order of Since there are two different functions of x defining
Integration the limits for y in Eq. (1) we have to divide R into
There are two ways of evaluation of double integrals, two regions R1 (QPBQ) and R2 (OPQO):

depending on the order of integrations R1 QPBQ : x varies from x = 0 to x = 2 − y
⎛ ⎞
x=b y=f
2 (x) y varies from y = 1 to y = 2
⎜ ⎟ √
⎝ f x, y dy⎠ dx (8.7) R2 OPQO : x varies from x = 0 to x = y
x=a y=f1 (x)
⎛ ⎞ y varies from y = 0 to y = 1
x=f2 (y)
y=d
⎜ ⎟
⎝ f x, y dx⎠ dy (8.8)
y=c x=f1 (y)
In some cases, evaluation of the integral in the

given order is impossible (please see Example 8.2)
or difficult and/or change of the order of integration
makes it possible or easy for evaluation. So, we con-
sider the procedure for the change of the order of
integration.
Identify the boundaries of the region of integration
by drawing the curves
y = f1 (x) , y = f2 (x) , x = a, x = b (8.9)
Figure 8.7
While changing the order, we have to fix the limits
for x in terms of y and then find the constant limits
for y. Changing the order we have to evaluate
⎛ 2−y ⎞ ⎛ √ ⎞
x may have two functions in y as two limits (as in
√ 2 1 y
Example 8.1 where x = 2 − y and x = y as limit ⎝ ⎜ ⎟
xdx⎠ ydy + ⎝ xdx⎠ydy
functions). Then we have to subdivide the region R
y=1 x=0 y=0 x=0
into sub-regions R1 and R2 .
2 1
Example 8.13 1 2 2−y 1 2 √y
= x 0 ydy + x 0 ydy
Change the order of integration and evaluate 2 2
1 2−x y=1 y=0
xydydx. [JNTU 2003, 1998S, 1998, 1997S], 2 1

1 1 2
o x2
= 4y − 4y + y dy +
2 3
y dy
2 2
[Madras 1993, Andhra 1991, Kerala 1990, Ranchi y=1 y=0
1990, Mysore 1994]
1
4 2
1 4 y 1 y3
= · 2y2 − y3 + +
Solution The inner integral has functions of x as 2 3 4 1 2 3 0
limits so it has to be evaluated with respect to y. The
1 2 4 3
outer integral has constant limits for x: = 2 2 − 12 − 2 − 13
2 3
y = x2 , y =2−x (1)
1 1 1
x = 0, x=1 (2) + 24 − 14 + ·
4 2 3
The region of integration R is bounded by the y-axis, 2 1 1 3
the parabola and the straight line x + y = 2. =3− · 7 + .15 + = .
3 8 6 8
Example 8.14
∞ ∞
e−y
Evaluate dy by changing the order of
y
0 y=x
integration.
Solution The inner integral is to be evaluated

with respect to y between y = x and ∞. The outer Figure 8.9
integral has to be evaluated with respect to x between
x = 0 and ∞. integration
x varies from x = y to x = ∞
y varies from y = 0 to y = ∞
⎛ ⎞
∞ ∞
The given integral = ⎝ xe−x /y dx⎠dy
2
0 x=y
∞ ∞
1 2 /y
=− ye−x dy
2 x=y
Figure 8.8 0
∞
The region of integration is R above the straight 1
line y = x to the right of the y-axis. Changing the = ye−y dy
2
order of integration, the limits for x are x = 0 to x =
o

y and for y are y = 0 to y = ∞ 1 e−y e−y ∞
= y − 1.
⎛ ⎞ 2 −1 (−1)2 0
∞ y −y ∞ −y 1
⎝ e e
dx⎠dy = y dy = ·
y y 2
0 0 0
∞ [Note that e−z → 0 as z → ∞.]
= −e−y 0 = 1.
Example 8.16
√
Example 8.15 1 1−x2
∞ ∞ Evaluate y2 dxdy by changing the order of
2 /y
Evaluate xe−x dydx by changing the order of
0 0
0 0 integration. [JNTU 1994S, 1998]
integration.
Solution The inner integral has the following
Solution The given double integral is limits for y:
⎛ ⎞
∞ y=x y=0 and y= 1 − x2
⎝ e−x /y dy⎠xdx
2
0 y=0
The outer integral has the following limits for x:
x=0 and x=1

Inner integral y varies from y = 0 to y = ∞
Outer integral x varies from x = 0 to x = ∞ So the region of integration R is bounded by
The region of integration R is bounded by y =
0 and straight line y = x. Changing the order of x = 0, y=0 and x2 + y2 = 1
Now changing the order of integration, we have Changing the order

√ of integration, the limits for y
the following limits for x: are y = 0 and y = a2 − x2 . The constant limits for
x are x = 0 and x = a. The given integral when
x=0 and x= 1 − y2 we change the order of integration is
⎡ √ ⎤
The limits for y are a y= a2 −x2
⎢ 2 ⎥
⎢ x + y2 dy⎥
⎣ ⎦dx
y=0 and y=1
x=0 y=0
√
Changing the order of integration, the given double a 3
a2 −x2
y
integral is = yx2 + dx
⎛√ ⎞ 3 y=0
x=0
2
1 1−y 1 √ a a
⎜ ⎟ 2 x= 1−y2 2 1 3/2
⎜ ⎟
dx⎠ y dy = (x)x=0 y dy = 2
a2 − x2 dx + a2 − x2
⎝ x
3
dx
y=0 x=0 y=0 x=0 x=0
a
2 1/2
a −x
2
Put y = sin θ ⇒ dy = cos θdθ = 3x2 + a2 − x2 dx
Limits y = 0 ⇒ θ = 0 3
x=0
y = 1 ⇒ θ = π/2 a
1 1/2
1 = a2 a2 − x2 dx
3
= y2 1 − y2 dy x=0
a
y=0 2 1/2
+ x2 a2 − x2 dx
π/2 3
x=0
= sin2 θ · cos θ · cos θdθ
θ=0 Put x = a sin θ ⇒ dx = a cos θdθ
π/2 Limits x = 0 ⇒ θ = 0
π
= sin2 θ cos2 θdθ = . x = a ⇒ θ = π/2
16 a
θ=0 a2 x 2 a2 −1 x
= a −x + 2 sin
Example 8.17 3 2 2 a 0
√
a
2 2
a −y
π/2
2
Evaluate x2 + y2 dydx. [JNTU 1989] + a2 sin2 θ · a2 cos2 θ · dθ
3
0
0 o
a2 a2 π 2 4 π πa4
Solution The inner integral has limits = · · + a · = .
3 2 2 3 16 8
x=0 and x= a2 − y 2 for x EXERCISE 8.2
Evaluate the following integrals by changing the
and the outer integral has limits order of integration:
y=0 and y = a for y a a
xdxdy
(1) .
The region of integration R is bounded by x2 + y2
0 y
1
x = 0, y=0 and x +y =a
2 2 2 Ans: πa
4
a a 8.3 Transformation of
y2 dydx
(2) . Coordinates
√ y 4 − a2 x 2
0 ax If a region G in the uv-plane is transformed into the
1 region R in the xy-plane by differentiable equations
Ans: πa2
6 of the form
√
1 2−x2 x = f (u, v), y = g(u, v) (8.10)
xdydx
(3) .
x2 + y2 then a function F(x, y) defined on R can be thought
0 x
1 of as a function
Ans: √
2
F [f (u, v), g(u, v)] (8.11)
√
4a 2 ax
defined on G. By a theorem from Advanced Calculus,
(4) dydx.
if all the functions involved are continuous and
0 x2 /4a
have continuous first derivatives then the integral of
Ans: 16a2 /3 F(x, y) over R and the integral of F [f (u, v), g(u, v)]
√ over G are related by
a x/a

(5) x2 + y2 dydx. [JNTU 1996]
0 x/a
a3 a
Ans: +
28 20
1 2−x
(6) xydxdy. [JNTU 1997, 1998, 2000]
0 x2
3
Ans:
8
√
3 4−y

(7) x + y dxdy.
0 1
181
Ans:
60
a 2a−x

(8) xy2 dydx. [JNTU 2002]
0 x2 /a
47a5 Figure 8.10 Transformation of coordinates
Ans:
120
a
√
b b b −y
2 2
F(x, y)dxdy = F [f (u, v), g(u, v)]
(9) xydxdy. [JNTU 2001S] R G
0 0 ∂(x, y)
Ans: a b /82 2 du dv (8.12)
∂(u, v)
∂x ∂x In this case, we have

∂ x, y ∂u ∂v
where = (8.13)
∂ (u, v) ∂y ∂y θ2 r2
∂u ∂v F(r, θ)rdrdθ = dθ F(r, θ)rdrdθ (8.16)
R θ1 r1
which is called the Jacobian of the coordinate trans-
formation.
θ1 = XOK, θ2 = XOL;
r1 = f1 (θ) and r2 = f2 (θ)
8.3.1 Polar Coordinates
are functions of θ representing boundary arcs ACB
In the case of polar coordinates, we have u = r and
and ADB, respectively.
v = θ x = r cos θ, y = r sin θ
In particular, these functions (one or both) can be
∂x ∂x constant numbers (Fig. 8.12).
∂ x, y ∂r ∂θ
=
∂ (r, θ) ∂y ∂y
∂r ∂θ
cos θ −r sin θ
=
sin θ r cos θ
= r(cos2 θ + sin2 θ) = r (8.14) Figure 8.12
Hence Eq. (8.12) becomes Case (ii) The pole is interior to the contour and
each polar ray meets the contour once.

In this case, we have to put r1 = 0 and θ1 = 0 and
F(x, y)dxdy = F (r cos θ, r sin θ) rdrdθ θ2 = 2π
R G
This corresponds to
2
θ=θ 2 (θ)
r=f
F(r, θ)dA = F(r, θ)rdrdθ (8.15)

θ=θ1 r=f1 (θ)
We discuss below the various cases.

Figure 8.13
Case (i) The pole is exterior to the contour and each
polar ray intersecting the contour meets twice (Fig. Case (iii) If the pole lies on the contour, then r1 = 0
8.11) θ1 = XOA, θ2 = XOB
If every circle with centre at the pole intersects the
contour at most twice (Fig. 8.14), then
L
B
dr Q
s D
M rdθ R
P
c
r K
A
N
o θ1 x
θ2

r2 θ2 π/2

a3
F(r, θ)rdrdθ = rdr F(r, θ)dθ (8.17) =− cos3 θd (cos θ)
3
R r1 θ1 0
Here r1 = OC, r2 = OD, and θ1 and θ2 are a 4 π/2

3
=− cos θ 0
functions of r representing the boundary arcs CBD 3.4
and CAD, respectively. a3
= .
12
Example 8.18
Example 8.19
Evaluate r sin θdA if the region R is the π a sin θ
R Evaluate rdrdθ.
semicircle of diameter a. 0 0
Solution
π a sin θ π sin θ
r=a
rdrdθ = dθ rdr
0 0 θ=0 r=0
π
2 a sin θ π/2
r a2 πa2
= dθ = 2 · sin2 θdθ = .
2 0 2 4
θ=0 0
Figure 8.15 Example 8.20

Solution Since OPA is right angled at P π θ)
a(1+cos
OP Evaluate r 2 cos θdrdθ.

= cos θ 0 0
OA
⇒ r = a cos θ (∵ OA = a) Solution
θ2 r2
π θ)
a(1+cos
f (x, y)dA = dθ F(r, θ)rdr
r 2 cos θdrdθ
R θ1 r1
0 0
Here r1 = 0, r2 = a cos θ π a(1+cos θ)
θ1 = 0, θ2 = π/2 r3
= cos θdθ
3 0
0
π/2 acos θ
π
∴ We have r sin θdA = dθ r sin θrdr a3
= (1 + cos θ)3 cos θdθ
R 0 0 3
0
π/2 acos θ
π
3
= sin θdθ r 2 dr a
= (1 − cos θ)3 (− cos θ) dθ
0 0
3
0
π/2 π
1 a cos θ a 3
= sin θ r 3 0 dθ = cos θ (1 + cos)3 − (1 − cos θ)3 dθ
3 6
0 0
π/2 a a
a3
= sin θ cos3 θdθ ∵ f (x)dx = f (a − x)dx
3
0 0 0
π
a3 Example 8.22

= 2 cos θ 3 cos θ + cos3 θ dθ
6 Calculate r 3 drdθ over the area included
0
a between the circles r = 2 sin θ and r = 4 sin θ.
1 [JNTU 1997]
= [f (x) + f (a − x)] dx
2
0
π/2 π/2 Solution The equations of the given circles are

a3 a3 r = 2 sin θ, (1)
= 2·3 cos θdθ +
2
· 2 cos4 θdθ
3 3
0 0 r = 4 sin θ (2)
π 2a 3 1 π 3 ⇒ r 2 = 2r sin θ, r 2 = 4r sin θ
= 2a3 · + · · ·

4 3 4 2 2
⇒ x2 + y2 = 2y, x2 + y2 = 4y
πa3 1 ⇒ x2 + (y − 1)2 = 1, (3)
= 1+
2 4 ⇒ x2 + (y − 2)2 = 22 (4)
5πa3
= . The circles are with
8
c1 = (0, 1), r1 = 1; c2 = (0, 2); r2 = 2
Example 8.21
√
π/4 cos 2θ r varies from r = 2 sin θ and r = 4 sin θ
r θ varies from θ = 0 and θ = π to cover the region R
Evaluate drdθ.
(1 + r 2 )2 lying between the two circles. Now
0 0
Solution y
√
π/4 cos 2θ
r r = 4sin θ
drdθ
1 + r2 R
(0, 2) Q
0 0
r = 2sin θ (0, 1)
P
⎛ √ ⎞ x
π/4 r= cos 2θ o
1 ⎜ 2rdr ⎟
= ⎝ 2 ⎠ dθ Figure 8.16
2 1 + r2
r=0 r=0
π r=4 sin θ
π/4 r=√cos 2θ r drdθ =
3
r 3 dr dθ
1 1
= − dθ r=2 sin θ
2 1 + r2 r=0
R θ=0
0 π
1 4 4 sin θ
π/4 = r r=2 sin θ dθ
1 1 4
= 1− dθ θ=0
2 1 + cos 2θ
0 π
4 − 24
4
π/4 π/4 = sin4 θdθ
1 1 1 4
= (dθ) − sec2 θdθ 0
2 2 2
0 0 π/2

π 1 π/4 = 60.2 sin4 θdθ
= − (tan θ)0
24 4 0
π 1 3 1 π 45
= − . = 60 · 2 · · · = π.
8 4 4 2 2 2
EXERCISE 8.3
Example 8.23
Find the area of the cardioid r = a(1 + cos θ).
1. Find the area of one loop of the curve r = a cos 3θ.
Solution The curve is symmetric about the initial πa2
line (the x-axis). Ans: sq.u.
12
The area of the loop
2. Find area lying between the cardioid r = a(1 −
= 2 Area OABO π
π cos θ) and its double tangent θ = .
1 2 2
=2 r dθ √ a2
2 Ans: 15 3 − 8π sq.u.
0 16
π 3. Show that the area included between the cardioids
=a 2
(1 + cos θ)2 dθ r = a(1 + cos θ) and r = a(1 − cos θ) is
0 (3π − 8)a2 /2 sq.u.
π
θ dθ 4. Find the area outside the circle r = 2a cos θ and
= 2×4a 2
cos4
2 2 inside the cardioid r = a(1 + cos θ).
0
πa2
3 1 π 3πa2 Ans:
= 8a2 · · · = . 2
4 2 2 2
5. Find the area of the loop of the curve x3 + y3 =
Example 8.24
3axy by transformation to polar coordinates.
Find the whole area of the lemniscate r 2 = a2 cos 2θ. 3a2
Ans:
Solution The curve is symmetric about both the 2
π
axes. The limits for θ are θ = 0 to for the upper
4 8.4 Double Integral—Volume
half loop.
Enclosed by a Cylindrical
Surface
y
3π
θ= 4 π
π
θ= 4 Let z = f (x, y) be a surface above the xy-plane whose
θ= 2
projection on the xy-plane is the surface with area S.
Divide S into elementary rectangles with area δxδy
θ=0 by drawing lines parallel to the x and y coordinate
o axes. Suppose we construct a prism of height z on
the rectangular element. The volume of this prism
between the xy-plane and the surface z = f (x, y) is
δv = zδxδy = f (x, y)δxδy. The total volume of the
cylinder with S as base is obtained from
Figure 8.17 Lemniscate of Bernoulli

π/4 v= zdxdy = f (x, y)dxdy (8.18)
1 2
The required area = 4 r dθ S S
2
0
in cartesian coordinates and
π/4

=a 2
cos 2θdθ
V = F(r, θ)r dr dθ (8.19)
0
S
π/4
= a2 (sin 2θ)0 = a2 · 1 = a2 . in polar coordinates
straight line y = x and x = 1 and where top lies in

the plane.
z = f (x, y) = 3 − x − y.
Solution For any x between 0 and 1, y may vary
twice y = 0 to y = x (Fig. 8.18)
1 x
v= (3 − x − y)dydx
0 0
1 y=1
1
= 3y − xy − y2 dx
2 y=0
0
1
3 3 2 1 3 1
= 3x − x2 dx = x − x
2 2 2 0
0
=1
When the order of the integration is reversed, the
integral for the volume is
⎛ 1 ⎞
1
V = ⎝ (3 − x − y)dx⎠ dy
0 x=y
1 x=1
x2
= 3x − − xy dy
2 x=y
0
1
12 y2
= 3.1 − − 1 · y − 3y − −y 2
dy
2 2
0
1
5 3
= − 4y + y2 dy
2 2
0
1
5 1
= y − 2y2 + y3 = 1.
2 2 0
Example 8.26
Find the volume of the cylinder x2 + y2 = 1 bounded
above by the plane x + y + z = 4 and below by the
Figure 8.18
xy-plane.
Solution The equation of the plane

Example 8.25
Find the volume of the prism whose base is the tri- x+y+z = 4
angle in the xy- plane bounded by the x-axis and the or z = 4−x−y (1)
1
= [4θ − r(sin θ − cos θ)]2π
0 rdr
r=0
1 1
r2
= 4.2π rdr = 4 · 2π = 4π
2 0
0
But in the following example, reversal of the order

results in reducing the problem to a case of not being
able to express it in elementary functions.

sin x
Equations of the base of the cylinder Evaluate dA where A is the triangle in
x
x2 + y2 = 1, z=0 (2) A
the xy-plane bounded by the x-axis, the x-axis the
For any fixed values of x and y in the xy-plane, z line y = x and the straight line x = 1.
varies from z = 0 to z = 4 − x − y.
Also x and y vary within x2 + y2 ≤ 1. Hence the Solution The region of integration is the same
required volume is as in the above examples. If we first integrate with
respect to y and then with respect to x, we have
V = (4 − x − y)dxdy (3) ⎛ ⎞
1 x 1
⎝ sin x ⎠ sin x y=x
A
dy dx = y dx
Put x = r cos θ, y = r sin θ 0 ≤ r ≤ 1 and x x y=0
x=0 y=0 0
0 ≤ θ ≤ 2π and dxdy = rdrdθ substituting these in
1
Eq. (3)
= sin xdx = − cos 1 + 1
2π 1 0
v= (4 − r(cos θ + sin θ)rdr)dθ = 0.45 approx.
θ=0 r=0
If we reverse the order of the integration and try
2π 1 1 1
1 3 sin2 x
= 2r − r (cos θ + sin θ)
2
dθ to calculate dxdy, we are restricted by
3 r=0 x
θ=0
0 y=1
2π the fact that this cannot be expressed in elementary
1
= 2 − (cos θ + sin θ) dθ terms.
3
θ=0
2π
1
= 2θ − (sin θ − cos θ) 8.5 Multiple Integrals—Triple
3 θ=0 Integrals
= 4π cu.u. Integration of functions of two variables leads to dou-
Note ble integrals. Similarly, the integration of functions of
Changing the order of integration, we can evaluate three or more variables leads to triple or multiple inte-
the integral as easily: grals. Consider a function of three variables defined
in a bounded three-dimensional region R. Subdi-
1 2π vide R into subregions Ri and form lower sums
V = (4 − r cos θ − r sin θ)dθ · rdr $n $n
mi Ri , upper sums Mi Ri and intermediate
r=0 θ=0 i=1 i=1
$
n
problem instead of the rectangular cartesian coor-
sums f (ξi , ηi , ζi ) Ri , where mi is the infimum and
i=1 dinates. So we study the procedure for transforma-
Mi is the supremum of the function f in the subre- tion of coordinates from cartesian to cylindrical and
gion Ri . Also, f (ξi , ηi , ζi ) is a function value at some spherical polar coordinate.
point Pi in Ri . If the sequences of lower and upper As already stated in the case of transformation of
sums tend to a common limiting value as n → ∞ coordinates in double integral, if rectangular coordi-
and Ri → 0 so do the sequence of intermediate nates x = x(u, v, w), y = y(u, v, w), z = z(u, v, w)
sums and the common limit is defined to be the triple are one-to-one continuously differentiable functions
integral of the function f (x, y, z) over the region R. of the coordinates u, v, w then the volume element dv
is multiplied by the absolute value of the functional
determinant called the Jacobian
8.5.1 Triple Integral
∂x ∂x ∂x
∂u ∂v ∂w
%
n

f (x, y, z)dR = n→∞
lim mi Ri x, y, z ∂y ∂y ∂y
J = . (8.22)
R
Ri →0 i=1 u, v, w ∂u ∂v ∂w
%
n ∂z ∂z ∂z
= n→∞
lim Mi Ri (8.20) ∂u ∂v ∂w
Ri →0 i=1
Any function f that is bounded and continuous in 8.5.3 Cylindrical Coordinates

R is integrable in this sense, whereas the region of
intergration can still be thought of geometrically as a x = ρ cos φ, y = ρ sin φ, z = z (8.23)

region in space. A geometrical interpretation of the x, y, z
J = (8.24)
integral is no longer possible. If f (x, y, z) ≡ 1, we ρ, φ, z
get on evaluation of the triple integral the volume of cos φ −ρ sin φ 0
the solid enclosed between the limits of integration. = sin φ ρ cos φ 0 = α
In the context of mechanics, if f (x, y, z) denotes the 0 0 1
density at a point P(x, y, z) in the region R then it
could be interpreted as the total mass of the solid. f (x, y, z)dv
Triple (and multiple) integrals can be evaluated
V
by repeated integrations with respect to the variables ⎡ ⎛ ⎞ ⎤
x, y, z · · · in turn within the limits of variations, as ρ=b φ=f
2 (ρ) z=g2 (ρ,φ)
⎢ ⎝ ⎥
shown below. = ⎣ F(ρ, φ, z)dz ⎠ dφ⎦ ρdρ
ρ=a φ=f1 (ρ) z=g1 (ρ,φ)
f (x, y, z)dR (8.25)

R
⎧ ⎡ ⎤ ⎫
x=b⎪ y=f (x) z=g2 (x,y)
⎨ 2 ⎪
⎬ 8.5.4 Spherical Polar Coordinates
⎢ ⎥
= ⎣ f (x, y, z)dz⎦ dy dx. (8.21)
⎪
⎩ ⎪
⎭ x = rsin θcos φ, y = rsin θsin φ, z = rcos θ (8.26)
x=a y=z1 (x) z=g1 (x,y)
x, y, z
J
r, θ, φ
8.5.2 Transformation of Triple sin θ cos φ r cos θ cos φ −r sin θ sin φ
Integrals = sin θ sin φ r cos θ sin φ r sin θ cos φ
In many cases it would be more convenient to use a cos θ −r sin θ 0
coordinate system suitable for the geometry of the = r 2 sin θ (8.27)
r=b θ=f
2 (r) 3 2 1
f (x, y, z)dv = ∴ dx dy ex+y+z dz
V r=a θ=f1 (r) 0 0 0
⎡⎛ φ=g (r,θ) ⎞ ⎤ 3
2
⎣⎝ F(r, θ, φ)dφ⎠ sin θdθ ⎦ r 2 dr. (8.28) = (e − 1)(e2 − 1)ex dx
φ=g1 (r,θ) 0
= (e − 1)(e2 − 1)(ex )30

Example 8.28
1 2 3 = (e − 1)(e2 − 1)(e3 − 1).
Evaluate dx dy xyz dz. Example 8.30
0 1 1 a x (x+y)

Evaluate e(x+y+z) dzdydx.
Solution
0 0 0
3 3 [JNTU 2003S, 1989]
z2
xyz dz = xy
z 1 Solution
1
1 2 x+y
= 3 − 12 = 4xy x+y
2 ex+y+z dz = ex+y (ez )0
2
2 0
4xy dy = 2xy2 1 = ex+y (ex+y − 1)
⎛ ⎞
1
x x+y x

= 2x 22 − 12 = 6x ⎝ ex+y+z dz ⎠ dy = ex+y ex+y − 1 dy
1 2 3 1 0 0 0
∴ dx dy xyz dz = 6x dx x

0 1 1 0 = e2x e2y − ex ey dy
1
= 3x2 0 = 3. 0
x
e2y
= e2x − ex (ey )x0
2 0
Example 8.29 e4x e2x
3 2 1 = − − e2x + ex
2 2
Evaluate dx dy ex+y+z dz.
e4x 3
0 0 0 = − e2x + ex
2 2
Solution
a x x+y
1 ex+y+z dz dy dx
ex+y (ez dz) = ex+y (ez )10 0 0 0
0 a
e4x 3
= e (e − 1) x+y
= − e2x + ex dx
2 1 2 2 2
0
dy ex+y+z dz = ex+y (e − 1)dy 4x a
e 3
= − e2x + ex
0 0 0 8 4 0
= (e − 1)ex (ey )20
e4a 3 2a 1 3
= (e − 1)(e2 − 1)ex = − e + ea − − +1
8 4 8 4
e4a 3 3 Example 8.32

= − e2a + ea − .
8 4 8 Evaluate the triple integral xy2 z dx dy dz
Example 8.31
e log yex taken through the positive octant of the sphere x2 +
y2 + z 2 = a2 . [JNTU 2000]
Evaluate log z dz dx dy. [JNTU 1998S]
1 1 1 Solution The equations of the sphere in the first
octant are
Solution We know that
x2 + y2 + z 2 = a2 x ≥ 0, y ≥ 0, z ≥ 0. (1)
1
log t dt = log t · t − tdt The limits of integration are
t
= t · log t − t (1) z = 0 and √ a2 − x2 − y2 , for z
y = 0 and a2 − x2 , for y
t2 1 t2
t log t dt = log t − dt x = 0 and a, for x,
2 t 2 √
t2 t2 a2 −x2 −y2
√
= log t − (2) xy2 2 a2 −x2 −y2
2 4 xy2 z dz = z 0
2
ex 0
∴ log z dz = (z log z − z)e1

x
1 2 2
= xy a − x2 − y2
1 2
= ex log ex − ex − 0 + 1 1 1
= x a2 − x2 y2 − xy4
2 2
= xex − ex + 1 √ ⎛√ ⎞
a2 −x2 −y2
a2 −x2
log y ⎜ ⎟
⎜ xy2 zdz ⎟
log y
[(x − 1)ex + 1]dx = [(x − 1)ex − ex + x]1 ⎝ ⎠ dy
y=0 z=0
1
√ √
= (log y − 1)elog y − elog y a2 −x2 a2 −x2
x 2 1 4
+ log y − (0 − e1 + 1) = a − x2 y2 dy − xy dy
2 2
= y log y − y − y + log y y=0 y=0
√
+e − 1 √ a2 −x2
a2 −x2
= y log y − y − 2y + log y x a2 − x 2
1 3 1 1
= y − x y5
+e − 1 2 3 2 5
y=0
y=0
e
1 5/2 x 2 5/2
(y log y − 2y + log y + e − 1)dy = x a2 − x2 − a − x2
6 10
1 x 2 5/2
e = a − x2
y2 y2 15
= log y − − y2 + y log y1 − y + (e − 1)y
2 4 1 √ √
a2 −x2 −y2
e2 e2 a a2 −x2
= log e − − e2 + e log e − e + e(e − 1) − ∴ xy2 z dz dy dx
2 4
1 0 0 0
0− −1+0−1+e−1 a
4 1 5/2
1 2 13 = x a2 − x2 dx
= e − 2e + . 15
4 4 0
a
1 5/2 2 Solution The required area is
=− a2 − x2 d a − x2
30 π/4
0 π/4
A = (cos x − sin x)dx = (sin x + cos x)0
1 2 2 7/2 a
a7
= − a − x2 = . 0
30 7 105
0
1 1 √
= √ + √ − 0 − 1 = ( 2 − 1) sq.u.
2 2
8.6 Applications of Example 8.35
Integration—Quadrature √ √
Find the area bounded by the curve x+ y = 1 and
By quadrature we mean the calculation of areas of the coordinates axes. [JNTU 1993]
plane regions with curved boundaries.
Solution The equation of the curve
√ √
8.6.1 Area Under a Curve y = f (x) x+ y = 1
√ √
The area under a curve y = f (x) and above the x-axis or y = (1 − x)2 = 1 + x − 2 x (1)
and between the ordinates x = a and x = b is found
The limits for x are x = 0 and x = 1.
from the integral
The required area
b 1 1
A= f (x)dx (8.29) √
A= ydx = (1 + x − 2 x)dy
a
0 0
1
If f (x) ≥ 0 in the interval a ≤ x ≤ b. x2 2 3/2
If f (x) changes sign in [a, b] we may split the inter- = x+ −2 x
2 3 0
val into subintervals in which f (x) takes positive and
1 4 1
negative values and the integral into a sum of positive = 1 + − (1) = .
and negative contributions corresponding to oriented 2 3 6
areas. If we ignore orientation, the total area is the Example 8.36
sum of the absolute values of these contributions. Find the area included between the cycloid and its
The area included between two curves y = f1 (x) base
and y = f2 (x) is given by
x = a(θ + sin θ)y = a(1 − cos θ).
b
A= (f1 − f2 )(x)dx [f1 (x) ≥ f2 (x)]. (8.30)
a
Example 8.33
Find the area under the curve y = cos x in (0, π/2).
Solution The required area is
π/2
π/2
A = cos xdy = (sin x)0 = 1 sq.u.
0
Figure 8.20
Example 8.34
Solution The equations of the curve are
Find the area included between the curves y = cos x
and y = sin x in (0, π/4). x = a(θ + cos θ), y = a(1 − cos θ)
The base of the cycloid is the line segment AB.

y2
The required area = 0ABO = 2OAB = 2 xdy
y1
At y = 2a, x = aπ. The limits for θ are θ = 0 to π
π π
dy
= 2 x dθ = 2 a(θ + sin θ)a sin θdθ Figure 8.22
dθ
θ=0 0
⎡ ⎤
π/2 Solution The required area is
= 2a2 ⎣(−θ cos θ + sin θ)π0 + 2 sin θdθ ⎦
2
b
0
A= ex dx = (ex )ba = eb − ea
a
1π
= 2a 2
(−π cos π + sin π) + (2) If a → −∞, A tends to a finite limit
22
π
= 2a2 π+ = 3πa2 .
2 A = eb ∵ ea → 0 as a → −∞
Example 8.37
The improper integral converges and the region
Find the quadrature of the parabola y = ax3/2 (Neil’s
extending to infinity has finite area.
parabola).
Example 8.39
Find the quadrature of the rectangular hyperbola
y = k 2 /x from x = a to x = b.
Figure 8.21
Solution The required area is OPAB =
x1
2a 5/2 Figure 8.23
a x3/2 dx = x
5 1 Solution The required area is
0
2
If y1 = ax1
5/2
then A =
x1 y1 so that the area A b
5 dx
is less than that of the straight-angled triangle with A=k 2
= k 2 (log b − log a)
x1 y1 x
base x1 and height y1 is less by . a
10 = k log(b/a)
2
Area under the positive branch of Neil’s parabola
1 2 5x1 y1 − 4x1 y1 x1 y1 Here we have a rational curve for which the calcu-
x1 y1 − x1 y1 = = sq.u. lation of area involves a transcendental function.
2 5 10 10
Find the quadrature of the exponential curve y = ex Find the quadrature of the curve y = sin x from x = 0
from x = a to x = b. to x = π.
The required area S = s1 + s2 where
s1 = 2OAB, s2 = 2OA B
as the curves are symmetric about the x-axis.

Figure 8.24
a
Solution The required area is s1 = 2OAB = 2 ydx
π 0
A= sin xdx = (− cos x)π0 a
√
0 = 2×2 a (a − x)1/2 dx
= − cos π + cos 0 = −(−1) + 1 = 2 sq.u. 0
√ 2 a 8a2
Here we have a transcendental curve with rational = −4 a, (a − x)3/2 = (3)
3 0 3
area.
0

Example 8.41 s2 = 2OA B = 2 ydx
Find the area enclosed between the parabolas y2 = −a
4a(a−x) and y2 = 4a(x+a). [JNTU 1999] 0
√
= 2×2 a (a + x)1/2 dx
−a
y
√ 2 a 8a2
= 4 a (a + x)3/2 = (4)
3 0 3
S2 : y2 = 4a (a + x) The required area is
B (0, 2a) 8a2 8a2 16a2

S = s1 + s 2 = + = . (5)
3 3 3
(–a, 0) A' 0 A (a, 0) x
(0, –2a) B'
Example 8.42
Find the area between the parabola y2 = 4ax and the
S1 : y2 = 4a (a – x) straight line y + x = 3a. [JNTU 1990]
Figure 8.25
Solution The equation of the curves are
y2 = 4a(a − x) (1)
y2 = 4a(a + x) (2)
The points of intersection are Figure 8.26
B = (0, −2a), B = (0, 2a) Solution Solving the equations of the
The curves cut the x-axis at A = (−a, 0) and A = parabola y2 = 4ax (1)
(a, 0) and the straight line y + x = 3a (2)
We have (3a − x)2 = 4ax ∴ The required area = R1 + R2 + R3 + R4

⇒ x2 − 10ax + 9a2 = (x − a)(x − 9a) = 0 (3) √ √
8a2 4
⇒ x = a, x = 9a = + 2a2 + 4 3a2 − a2 + 54a2 − 4 3a2
A = (9a, −6a), B = (a, 2a) (4) 3 3
4 2
= 54 + a sq.u.
Also C = (3a, 0) 3
The region R consists of the subregions R1 , R2 , R3 EXERCISE 8.4

and R4 .
a √ 1. Find the area under the curve x2 = 4 − y and above
R1 = 2 4axdx (5) the x-axis.
32
x=0 Ans: sq.u.
3 a 3
√2 x 8a2
= 2×2 a = sq.u. 2. Find the area between the straight line y = 4 and
3 2 0 3
the curve x2 = 4 − y for x = −2 to x = 2.
( since the curve is symmetric about the x-axis) 16
3 3a Ans: sq.u.
x2 3
R2 = (3a − x)dx = 3a× −
2 a 3. Find the area enclosed between the parabola x2 = y
a
and y2 = x.
9a2 5a2 1
= − = 2a2 sq.u. Ans: sq.u.
2 2 3
1
(R2 = BDC = 2a · 2a = 2a2 ) 4. Find the area common to the parabola y2 = x and
2
thecircle x2+ y2 = 2.
3a √ 3 3a
√ 2 x π 1
R3 = 4axdx = −2 a Ans: + sq.u.
3 2 a 2 3
a
4a2 √ 5. Show that the area common to the ellipses a2 x2 +

= 3 3 − 1 sq.u. b2 y2 = 1 and b2 x2 + a2 y2 = 1 (0 < a < b) is
3
4
9a √ tan−1 a/b sq.u.
ab
R4 = 4ax − 3a + x dx
3a 6. Find the area enclosed between the circles x2 +y2 =
x + y −4x = 0.
2 2
4 and
16π √
9a Ans: − 3 sq.u.
√ 2 x2 3
= 4a · x3/2 − 3ax +
3 2 3a
7. Find the whole area between the form infi-
√ 2 √ √
= 4a · (9a 9a − 3a 3a) − 3a(9a − 3a) nite branches of the tractrix x = a cos t +
3 1 t
a log(tan2 ), y = a sin t.
1 2 2
+ [(9a)2 − (3a)2 ] Ans: πa2 sq.u.
2
√ 2 √ √
= 4a · a a(27 − 3 3) − 18a2 + 36a2 8. Find the large of the two areas into which the circle
3 x2 + y2 = 64a2 is divided by the parabola y2 =
4a2 4a2 √ 12ax.
= 27 − 3 3 + 18a2 √
3 3 16a2
√ Ans: 8π − 3 sq.u.
= (54 − 4 3)a2 sq.u. 3
8.6.2 Physical Applications Moment of inertia of a plane lamina

—Multiple Integrals
The moment of inertia of an elemental mass ρdxdy
Mass of a plane lamina
at P(x, y) of a plane lamina in the xy-plane about the
If ρ(x, y) be the density at any point P(x, y) of a plane x-axis is y2 ρdxdy. Therefore, moment of inertia of
lamina in the xy-plane, then the mass of an elemental the lamina about the x-axis is
area surrounding P is dm = ρ(x, y)dxdy and the total
mass of the lamina is given by ρy2 dxdy (8.35)
A
M = dm = ρ(x, y) dxdy (8.31)
where the integration is carried over the area of the
A
lamina. Similarly moment of inertia about the y-axis
integrated over the area of the lamina. is
ρx2 dxdy
Centre of mass of a plane lamina
A
The centre of mass x̄, ȳ of a plane lamina is given
by The moment of inertia of the lamina about the axis
perpendicular at 0 to the xy-plane is
⎫
x dm x ρ(x, y)dxdy ⎪
⎪
x̄ = = ⎪
dm ρ(x, y)dxdy ⎬ r x2 + y2 dxdy.
(8.32)
y dm y ρ(x, y)dxdy ⎪
⎪ A
ȳ = = ⎪
⎭
dm ρ(x, y)dxdy
Moment of inertia of a solid
Mass of a solid
Let us consider an elemental mass ρdxdydz around
P(x, y, z). Then moment of inertia of the solid about
M= dm = ρ(x, y, z)dxdydz (8.33) the x-axis
ρ y2 + z 2 dxdydz
integration being carried out throughout the volume and two more similar formulae. The following two
of the solid. theorems give moment of inertia about any general
axis.
Centre of mass of a solid
The centre of mass of a solid is given by Theorem of perpendicular axis

xdm xρ dxdydz Let Ix and Iy be the moments of inertia of a plane
x̄ = = (8.34) lamina about the x and y-axes, respectively. Moment
dm ρ dxdydz
of inertia of the lamina about an axis oz perpendicular
Similar formulae for ȳ and z̄. to it is given by Iz = Ix + Iy .
Moment of inertia Theorem of parallel axis
Let a particle P of mass m be at a distance r from a Let the moment of inertia of a body of mass M about
given line L. The moment of inertia of the particle an axis through the centre of gravity of the body be
P at a distance r from the given line L is mr 2 . For a I . Then the moment of inertia I of the body about
system of particles this is MK 2 where M = m is an axis l parallel to l at a distance d, is given by
the mass of the system of particles. I = I + Md 2 .
Transformation of coordinates—Triple The Jacobian of transformation is

integrals
∂x ∂x ∂x
Depending upon the geometry of the problem ∂r ∂θ ∂φ

under consideration, we may have to use x, y, z ∂y ∂y ∂y
J =
cylindrical or spherical polar coordinates in r, θ, φ ∂r ∂θ ∂φ
place of rectangular cartesian coordinates. We now ∂z ∂z ∂z
consider transformation of cartesian coordinates to ∂r ∂θ ∂φ
cylindrical and spherical polar coordinates in three sin θ cos φ r cos θ cos φ −r sin θ sin φ
dimensions. = sin θ sin φ r cos θ sin φ r sin θ cos φ
Cylindrical coordinates r cos θ cos φ −r sin θ sin φ
= cos θ
r cos θ sin φ r sin θ cos φ
x = ρ cos φ, y = ρ sin φ, z=z (8.36) sin θ cos φ −r sin θ sin φ
+r sin θ
sin θ sin φ r sin θ cos φ
The Jacobian of transformation is = r2 cos2 θ sin θ · 1 + r 2 sin3 θ · 1

= r 2 sin θ(cos2 θ + sin2 θ)
∂x ∂x ∂x = r 2 sin θ.
∂ρ ∂φ ∂z
∂y ∂y ∂y Transformation of coordinates
x, y, z
J = ∂ρ ∂φ ∂z
ρ, φ, z
∂z ∂z ∂z Example 8.43
∞ ∞
∂ρ ∂φ ∂z − x2 +y2
Evaluate e dxdy.
cos φ −ρ sin φ 0 0 0
= sin φ ρ cos φ 0 Solution Here it is convenient to transform the
0 0 1 cartesian coordinates into polars by putting
2
= ρ cos φ + sin φ = ρ
2
x = r cos θ, y = r sin θ,
dxdy = rdrdθ
y −1
f (x, y, z)dR ⇔ x + y = r ; θ = tan
2 2 2
x
R
⎧ ⎫ x and y ranges from 0 to ∞ (in the first quadrant)

⎛z ⎞ ∴ r varies from r = 0 to r = ∞
ρ=b⎪ φ=f
⎨ 2 (ρ)
2 =g2 (ρ,φ) ⎪
⎬
= ⎝ F(ρ, φ, z)dz⎠ dφ ρdρ. (8.37) and θ varies from θ = 0 to θ = π/2
⎪
⎩ ⎪
⎭
ρ=a φ=f1 (ρ) z=g1 (ρ,φ)
⎛ ⎞
∞∞ π/2 ∞
2
− x +y 2
dxdy = ⎝ e−r rdr⎠dθ
2
∴ e
Spherical polar coordinates
0 0 0 0
⎛ ⎞
π/2 ∞
x = r sin θ cos φ, y = r sin θ sin φ, 1 ⎝
e−r d(r 2 )⎠dθ
2
=−
z = r cos θ (8.38) 2
0 0
π/2 ∞ limits for θ are θ = 0 and θ = 2π.

1 2
=− e−r dθ
2 0 x2 y2
0 ∴ dxdy
x + y2
2
π/2 R
1
=− (0 − 1)dθ 2π b
2 r 4 cos2 θ sin2 θ
0 = rdrdθ
r2
1 −π π θ=0 r=a
=− = . ⎛ ⎞
2 2 4 2π b
= ⎝ r 3 dr ⎠ cos2 θ sin2 θ dθ
Example 8.44
√ 0 r=a
a a −y
2 2
2π
1 4 b
Evaluate x2 + y2 dxdy. = r r=a cos2 θ sin2 θ dθ
4
0 0 0
Solution Here it is convenient to transform the 2π

b4 − a4
cartesian coordinates into polar coordinates. = sin2 2θ dθ
16
Put x = r cos θ, y = r sin θ, dxdy = rdrdθ. 0
The limits for x are x = 0 and x = a2 − y2 and 2π
the limits for y are y = 0 and y = a. The region of b4 − a4
= (1 − cos 4θ) dθ
integration is the quarter of the circle in the quadrant. 32
0
The limits for r are 0 and a and the limits for θ are 0 2π
and π/2. b4 − a4 1
= θ − sin 4θ
32 4 0
√ ⎛ ⎞ b4 − a4
a −y
2 2
a π2 a = π.
16
∴ x2 + y 2
dxdy = ⎝ r 3 dr ⎠ dθ
0 0 θ=0 r=0 Example 8.46
π/2 1 a
1 4 a xdxdy
= r 0 dθ Evaluate the integral ·
4 x2 + y2
y=0 x=y
0
π/2 Solution The region of integration is bounded

a4
= dθ by the curves
4
0
x = y, x = a, y = 0 and y = 1
πa 4
= ·
8
Example 8.45

x2 y2
Evaluate dxdy over the annular region
x2 + y2
R
enclosed between the circles x2 + y2 = a2 and x2 +
y2 = b2 (b > a) by changing to polar coordinates.
Solution Put x = r cos θ, y = r sin θ, dxdy =

rdrdθ. The limits for r are r = a and r = b and the Figure 8.27
x = y ⇒ r cos θ = r sin θ Put x = r cos θ, y = r sin θ, dxdy = rdrdθ

π
⇒ tan θ = 1 ⇒ θ =
4 The circles equation become
y=0⇒θ=0
x = a ⇒ r cos θ = a ⇒ r = a sec θ C1 : r 2 = ar cos θ ⇒ r = a cos θ
C2 : r 2 = a2 ⇒ r=a
a a π/4 asec θ
xdxdy r cos θ
∴ = rdr dθ The limits for r are a cos θ and a, and those for θ
x2 + y2 r2 are 0 and π/2.
y=0 x=y y=0 r=0
π/4 asec θ √
= cos θdr dθ a a2 −x2
dxdy
0 r=0 ∴
√ a2 − x2 − y2
π/4 0
ax−x2
πa
= adθ = · π/2 a
4 rdr
θ=0
= √ dθ
Example 8.47 a2 − r 2
√ 0 r=a cos θ
a a2 −x2 π/2 a
dxdy
Evaluate the integral by = − a2 − r 2 dθ
√ a2 − x 2 − y 2 r=a cos θ
0 0
ax−x2
changing into polar coordinates. π/2
= 0+ a2 − a2 cos2 θ dθ
0
π/2
= a sin θ dθ
0
π/2
= −a (cos θ)0 = a · 1 = a.
Areas between curves—Polar coordinates

Example 8.48
Figure 8.28 Find the area of the lemniscate r 2 = a2 cos 2θ.
Solution
√ The region √
of integration is bounded
by y = ax − x2 , y = a2 − x2 Solution The equation of the curve is
⇒ x2 + y2 − ax = 0, x2 + y2 = a2 (1) r 2 = a2 cos 2θ (1)

2
a 2 a
⇒ x− + y2 = , x2 + y2 = a2 (2)
2 4 In cartesian coordinates
There are circle of centres C1 , C2 and radii r1 , r2 , 2
respectively, where x2 + y2 = a2 x2 − y2 (2)
a a ⎫
⎬
C1 = ,0 r1 =
2 2 (3) The curve is symmetric about both the axes. When
⎭ θ = 0, r = a and when θ = π/4r = 0. The required
C2 = (0, 0) r2 = a
Figure 8.29
area is 4 × area of OABO Figure 8.31
√ r 2 = 2ar cos θ ⇒ x2 + y2 = 2ax

π/4 r= cos 2θ π/4 2 r=√cos 2θ
r ⇒ (x − a)2 + y2 = a2 (2)
=4 rdr dθ = 4 dθ
2 r=0 c2 = (a, 0) , r 2 = a
θ=0 r=0 0
π/4 r=√cos 2θ The required area = 2× area of OAB (along the

r2
= 2a 2
dθ x-axis)
2 r=0
0 π π
θ= π
2 cos θ
r=2a 2 2a cos θ
=a 2
(sin 2θ)0=θ4 =a .
2 r2
=2 rdrdθ = 2 dθ
2 a cos θ
θ=0 r=a cos θ θ=0
Example 8.49 π π
2 2
Find the area enclosed between the curves r = a cos θ 3a2
and r = 2a cos θ. = 3a2 cos2 θdθ = 1 + cos2 θ dθ
2
0 0
π2
3a 2
sin 2θ 3a2 π 3πa2
= θ− = · = .
2 2 0 2 2 4
Example 8.50
Find the required area by double integration and the
cardioid r = a(1−cos θ) meet at θ = 0 and θ = π/2.
Solution The circle r = a sin θ and the cardioid

r = a(1 − cos θ) meet at θ = 0 and θ = π/2.
r = a sin θ (1)
Figure 8.30
r = a(1 − cos θ) (2)
Solution In cartesian coordinates, the equations
From Eqs. (1) and (2), we have sin θ = 1 − cos θ
of the circles are
θ θ 2θ
⇒ cos = 2 sin
2 sin
r = ar cos θ ⇒ x + y = ax
2 2 2
2 2 2
a 2 a2 θ θ θ
⇒ x− + y2 = (1) ⇒ sin cos − sin =0
2 4 2 2 2
a a π
c1 = , 0 , r1 = ⇒ θ = 0 or θ =
2 2 2
Hence the limits for θ are θ = 0 to θ = π/2. Hence 2. Areas

the limits for r are r = a(1 − cos θ) to r = a sin θ.
Hence the required area (a) Find the area lying inside the cardioid r = a(1+
cos θ) by double integration.
2
π
asin θ 2
π
Ans: 3πa2 /2 sq. u.
1 2 r=a sin θ
= rdrdθ = r r=a(1−cos θ) dθ (b) Find the area lying inside the cardioid r = a(1+
2
θ=0 r=a(1−cos θ) θ=0 cos θ) and outside the circle r = a by double
π integration.
2
2
a 1
= sin2 θ − (1 − cos θ)2 dθ Ans: πa2 sq. u.
2 2
θ=0
π (c) Find the area bounded by the circle r =
2
2
a 2 sin θ, r = 4 sin θ.
= −2 cos2 θ + (2 cos θ) dθ Ans: 3π sq. u.
2
θ=0
(d) Find the area outside the circle r = a and inside
1 π π
= a2 1 − · = a2 1 − . the cardioid r = a(1 + cos θ).
2 2 4
a2
Ans: (π + 8) sq. u.
4
EXERCISE 8.5
3. Volumes
1. Transformation of coordinates
Evaluate the integral by changing to polar coordi- (a) Find the volume of the sphere x2 + y2 + z 2 = 9
nates: bydouble integration.

4
Ans: 3π − cu. u.
1 − (x2 + y2 ) π
(a) dxdy over the positive
1 + x2 + y2 (b) Find the volume bounded by the xy-plane, the
quadrant of the circle x2 + y2 = 1. paraboloid x2 + y2 = 2z and the cylinder x2 +
π2 π y2 = 4. (Hamirpur 1995)
Ans: − Ans: 4π cu. u.
8 4
(c) Find the volume bounded by the xy-plane,the
(b) a2 − (x2 + y2 ) over the semi-circle x2 + paraboloid 2z = x2 + y2 and the cylinder x2 +
y2 − ax y2 = 4.
= 0 inthe positive quadrant. Ans: 4π
a3 π 2
Ans: − (d) Find the volume bounded by the paraboloid x2 +
3 2 3
y2 = az, the cylinder x2 + y2 = 2ay and the
a a plane z = 0. (Poona 1987, Mysore 1995)
x2
(c) dxdy. Ans: 3πa3 /2 cu. u.
x2 + y2
0 y (e) Find the volume of the paraboloid of revolution
a3 √
Ans: log(1 + 2) x2 + y2 = 4z cut off by the plane z = 4 by triple
3 integration.
√ Ans: 32π cu. u.
a a −y
2 2
(f) Find the volume of the region bounded by
(d) y2 x2 + y2 dxdy. paraboloid az = x2 + y2 and the cylinder
0 y x2 + y2 = b2 by triple integration.
πa5 πb4
Ans: Ans: cu. u.
20 2a
(g) Find the volume of a sphere of radius a by triple Put x = a sin3 θ, dx = 3a sin2 θ cos θdθ. When
integration. x = 0, θ = 0; when x = a, θ = π/2
4πa3
Ans: cu. u.
3 π/2
3
M = 2μb 2
a sin3 θ 1 − sin2 θ ·3a sin2 θ cos θdθ
8.7 Physical Applications
0
Example 8.51 π/2
Find the mass of a plate in the shape of the curve = 6μa b2 2
sin5 θ cos7 θdθ
x 2/3 y 2/3
+ = 1, the density being given by 0
ρ a= μxy. b 42 · 6 · 4 · 2 μa2 b2
= 6μa2 b2 = .
12 · 10 · 8 · 6 · 4 · 2 20
B (0, b)
Example 8.52
Find the mass of a lamina in the form of the cardioid
r = a(1 + cos θ) whose density at any point varies
A' (–a, 0) A (a, 0)
as the square of its distance from the initial line.
Solution The curve is symmetric about the initial

B' (0, –b) line. So, its mass M will be twice that above the initial
(Cartesian form)
line. Distance of P(r, θ) from the initial line = r sin θ.
The density at P(r, θ) is given by
ρ = μ (r sin θ)2 = μr 2 sin2 θ.
The limits of variations for θ are θ = 0 and θ = π.
Figure 8.32
The limits of variation for r are 0 and a(1 + cos θ)
Solution The curve is symmetric about both the
axes. The required mass is 4 times the mass of the
θ=π θ=a(1+cos
θ)
plate in the first quadrant.
∴ M =2 ρrdrdθ
x 2/3 3/2
y =b 1− = y1 (say) θ=0 r=0
a
θ=π r=a(1+cos
θ)
x varies from x = 0 to x = a
=2 μr 3 sin2 θdrdθ
y varies from y = 0 to y = y1 .
θ=0 r=0
a y1 a y1 π a(1+cos θ)
r4
M =4 ρdydx = 4 μxydydx =2 μ sin θ 2
dθ
4 0
0 0 0 0 0
a y1 π
y 2
μa4
=4 μx dx = sin2 θ (1 + cos θ)4 dθ
2 0 2
0 0
a π 2 4
4a 4
θ θ θ
= 2μ 2
xy dx = 2 sin cos 2 cos2
dθ
2 2 2 2
0 0
a π
x 2/3 θ θ
= 2μ xb 2
1− dx = 32μa 4
sin2 cos10 dθ
a 2 2
0 0
√
π/2 Put 2 sin θ = sin t
θ 1 √
= 32μa · 2 sin2 t cos10 tdt, t = , dt = dθ
4
2 cos θdθ = cos t dt
2 2
0 π/2
1·9·7·5·3·1 π 21 4a 1
= 64μa4 · = μπa4 . x̄ = (1 − sin t)3/2 √ cos tdt
12 · 10 · 8 · 6 · 4 · 2 2 32 3 2
0
Example 8.53 When θ = 0, t = 0
Find the centroid of a loop of the lemniscate r 2 = π π
θ= , t=
a2 cos 2θ. 4 2
Solution The curve is symmetric about the initial √ π/2
2 2a
line. So, the centre of gravity lies on the initial line so = cos4 tdt
3
that ȳ = 0. Let the density be k. For the loop θ varies
π π √ √
0
√
from − to and r varies from 0 to a cos 2θ. 2 2a 3 1 π πa 2
4 4 √ = · · · =
a cos 2θ
π/4
3 √ 4 2 2 8
r cos θρrdrdθ πa 2
−π/4 0 G= ,0 .
x̄ = √ 8
a cos 2θ
π/4
ρrdrdθ
−π/4 0
Example 8.54
√ Find the moment of inertia of a quadrant of the ellipse

π/4 a
cos 2θ
x2 y2
kr cos θdrdθ
2
+ = 1 of mass M about the z-axis, if the density
−π/4 0 a2 b2
= √ at a point is proportional to xy.
a cos 2θ
π/4
krdrdθ
−π/4 0
√

π/4
r3
a cos 2θ y
cos θ 3
dθ
−π/4 0
= √ B (0, b)

π/4
r2
a cos 2θ
2
dθ
−π/4 0
b

π/4
a3
3
cos θ (cos 2θ)3/2 dθ
−π/4 0 a A (a, 0) x
=

π/4
a2
2
cos 2θdθ
−π/4 Figure 8.33

π/4
Solution Here the region of integration is the
2 a
cos θ (cos 2θ)3/2 dθ a 2
0
3
quadrant OAB bounded by x = 0, x = b − y2
= b

π/4
1 and y = 0, y = b. Let ρ = μxy. Then the mass M of
2 cos 2θdθ
0
2 the quadrant is

π/4 3/2 a
√
b b b −y
2 2
2a
3
cos θ 1 − 2 sin2 θ dθ
=
0
1 π/4 M = ρdxdy = μxydxdy
2
sin 2θ 0 R 0 0
π/4 b √
4a 1 a b2 −y2
= cos θ 1 − 2 sin2 θ = μy x2 0b dy
3 2
0 0
b By the theorem of perpendicular axis, the moment

μ a2
= y b2 − y2 dy of inertia about the z-axis is
2 b2
0

b 1 2
μa 2y
2 2
y4 Iz = Ix + Iy = M a + b2 .
= b − 3
2b2 2 4 0

μa2 b4 b4 μa2 b2
= − = Example 8.55
2b2 2 4 8 Find the moment of inertia of one loop of the
8M lemniscate r 2 = a2 cos 2θ about the initial line.
⇒μ= 2 2
ab (Punjab 1980, Ranchi 1989)
The moment of inertia of OAB about the x-axis
a
√
b b b −y
2 2

Ix = ρy2 dxdy = μxy3 dxdy
A 0 0
b √
μ a b2 −y2
= y3 x2 0b dy
2
0
b
μa2
= b2 y3 − y5 dy
2b2
0
b
μa2 b2 y4 y6
= −
2b2 4 6 0 Figure 8.34 Lemniscate of Berrnoulli
μa b
2 6
μa b
2 4
1
= = = μb2 Solution The equation of the lemniscate is
2b2 12 24 3
The moment of inertia of OAB √ about the y-axis

a
b b −y
b
2 2 r 2 = a2 cos 2θ (1)
Iy = ρx2 dxdy = μx2 ydxdy
A 0 0 It is symmetric about the initial line.
b √
μ a b2 −y2
= y x4 0b dy For one loop, θ varies from θ = π/4 to π/4 (2)
4 √
0 and r varies from 0 to a cos 2θ (3)
b
μ a4 2
= b2 − y2 ydy
4 b4 Let M be the total mass and ρ be the density of the
0
lemniscate’s loop.
b
μa4 1
= − b2 − y2 d b2 − y2 √
4b4 2 π/4 a cos 2θ

0
∴ M = ρrdrdθ
3 b
μa 4 b − y2
2
θ=−π/4 r=0
=− 4
8b 4 π/4
0 ρ 2 a√cos 2θ
μa 6 4
8μ a b 1 4 2 = r 0 dθ
= b = 2 2· = Ma2 2
24b 4 ab 24 3 θ=−π/4
π/4
ρ
= a2 cos 2θdθ
2
θ=−π/4
π/4
= ρa 2
cos 2θdθ
θ=0
ρa2 1
= (sin 2θ)π/4
0 = ρa2 (4)
2 2
The moment of inertia about the initial line (θ = 0)
Figure 8.35
is given by
√ Solution Let us take the axes as shown in the
π/4 a cos 2θ
figure. The region R is bounded by the sides
√ of OAB
Ix = ρy dxdy = ρ
2
r 2 sin2 θrdrdθ
x varies from 0 to a, y varies from 0 to 3x√ and when
R
⎛
−π/4
⎞
0 x varies from 0 to 2a, y varies from 0 to − 3(x−2a).
√
π/4 a cos 2θ ∴ The product of inertia of the triangle about the
⎜ 3 ⎟ 2 coordinate axes is given by
=ρ ⎝ r dr ⎠ sin θdθ
θ=−π/4 r=0 Ixy = ρxydxdy
π/4 R
ρ 4 a√cos 2θ 2 √ √
= r 0 sin θdθ a 3x 2a − 3(x=2a)
4
θ=−π/4 = ρxydydx + ρxydydx
π/4 0 0 a 0
ρa4
= .2 sin2 θ cos2 θ2θdθ a 2
√3x 2a −√3(x−2a)
4 y y2
0 = ρx dx + ρx dx
2 0 2 0
π/4 0 0
ρa4
= (1 − cos 2θ) cos2 2θdθ a 2a
4 3ρ 3ρ
0 = 3
x dx+ x3 − 4ax2 + 4a2 x dx
2 2
0 0
⎛ π/2 ⎞ a 2a
π/2 3ρ x4 3ρ x4 x3
ρa 4
⎝ cos2 φdφ − cos3 φdφ⎠ ; = + − 4a + 2a2 x2
= 2 4 2 4 3
4×2 0 0
0
π
0
= ρa4
2θ = φ, θ = π/4 ⇒ φ =
2 Example 8.57

ρa4 1 π 2 x2 y2 z 2
= · − Find the volume of the ellipsoid + + = 1.
8 2 2 3 a2 b2 c 2
[JNTU 1998] (or)
πa 4
2M a4
= (3π − 8) = 2 · (3π − 8) Find the volume of the greatest rectangular parallel
9 6 a 96 0 piped that can be inscribed in an ellipsoid
Ma2 x2 y2 z 2
= (3π − 8) . + + = 1. [JNTU 1999]
48 a2 b2 c2
Example 8.56
Solution The coordinate planes cut the solid
Find the product of inertia of an equilateral triangle
given by
of side 2a about two perpendicular axes in its plane x2 y2 z2
vertex one of the axes being along a side. + + =1 (1)
a2 b2 c2
into 8 octants. The volume of the solid is therefore x2 + y2 + z 2 = a2 lying inside the cylinder
equal to 8 times the volume of portion of the solid in x2 + y2 = ay. [Kerala 1990; JNTU 2001S]
the first octant bounded by x = 0, y = 0, z = 0 and
the surface Eq. (1). For fixed x and y on the xy-plane,
z varies from Solution The volume can be found by changing
* the problem to cylindrical coordinates (ρ, φ, z).
x2 y2
z = 0 to z = c 1 − 2 − 2 .
a b ∴ x = ρ cos φ, y = ρ sin φ, z = z

x, y, z
Consider the quadrant of the ellipse in the first and J =ρ (1)
quadrant of the xy-plane. For a fixed x, y varies from ρ, φ, z
* The equation of the sphere is ρ2 + z 2 = a2 (2)
x2 The equation of the cylinder is ρ = a sin φ (3)
y = 0 to b 1 − 2 .
a
The volume of the portion of the cylinder cut off by
Then x varies from x = 0 to x = a. Hence, the
required volume is
z
a y1 z1
V =8 dzdydx
x=0 y=0 z=0
where y1 = b 1 − x2 /a2 ,
*
x2 y2
z1 = c 1 − 2 − 2
a b P (ρ, φ, z)
⎛ ⎞ O y
ay1 * 2 2
φ
Q
x y
∴ V = 8 ⎝ c 1 − 2 − 2 ⎠dx
a b
x=0 y=0
⎛ y ⎞ a/2
a 1 +
8c ⎝
= y12 − y2 dy⎠ dx x
b
x=0 y=0
a + y
8c y 2 y2 y 1
= y1 − y2 + 1 sin−1 dx
b 2 2 y1 0 Figure 8.36
x=0
a the sphere is 4 times the volume in the first octant. z
8c y12 π varies from 0 to a2 − ρ2 , ρ varies from 0 to a sin φ
= · dx
b 2 2 and φ varies from 0 to π/2. (Refer Fig. 8.36).
x=0
The required volume is
a
2πc 2 x2 √
= b 1 − 2 dx
π/2 asin φ a −ρ
2 2
b a

0
V =4 ρ dz dρ dφ
a3 4πabc
= 2πbc a − 2 = . φ=0 ρ=0 z=0
3a 3
π/2 asin φ
Example 8.58 =4 ρ a2 − ρ2 dρ dφ
Find the volume of the portion of the sphere φ=0 ρ=0
π/2 asin φ (b) Using double integral, find the moment of

1 1/2
=4 − (a2 − ρ2 ) dρ dφ inertia about the x-axis of the area enclosed by
2
φ=0 ρ=0 the straight lines.
π/2 a sin φ x = 0, y = 0; x/a + y/b = 1 [Mysore 1987S]
1 2 3/2 Ans: ab3 /12
=4 − a − ρ2 dρ
3 0
φ=0 (c) Find the moment of inertia of a circular plate
π/2 about a tangent. [Gorakhpur 1991]
4a3
= 1 − sin3 φ dφ [Hint: Use the theorem of parallel axes]
3 Ans: 5Ma2 /4
φ=0
2a3 3. Mass
= (3π − 4).
9
(a) Find the mass of the lamina x = a(1 + cos θ)
EXERCISE 8.6 whose density varies as the square of its distance
from the initial line.
1. Volumes
Ans: 21 Kπa4 /32
(a) Find the volume bounded by the cylinder
(b) Find the mass of the tetrahedron bounded by the
x2 + y2 = 4 and the planes y + z = 4 and z = 0.
coordinate planes and the plane
[JNTU 2000]
Ans: 16 π cu. u. x y z
+ + = 1,
a b c
(b) Find the volume of the solid generated by the
revolution of the cardioid r = c(1−cos θ) about the density at any point (x, y, z) K.
its axis. Kabc
Ans:
Ans: 8 πa3 /3 cu. u. 6
(c) Find the volume bounded by the xy-plane, the 4. Centroid

paraboloid 2z = x2 + y2 and the cylinder
x2 + y2 = 4. (a) Find the centre of gravity of the area of the car-
Ans: 4π cu. u. dioid r = a(1 − cos θ).
Ans: (−5a/6, 0)
2. Moment of inertia
(b) Find the centroid of the area enclosed by the
(a) Find the moment of inertia about the initial line parabola y2 = 4ax, the axis of x and its latus-
of the cardioid
0 r = a(1 + cos θ). rectum.
Ans: 21πa4 32 Ans: (3a/5, 0)
Sequences and Series 9
9.1 Introduction coordinate, the term an , is a real number. Since the
In the case of certain differential equations, correspondence is single-valued, sequences can also
solutions are obtained in the form of infinite series. be defined as functions.
Infinite series occur in Fourier and Fourier–Bessel
expansions. Further, exponential, logarithmic and 9.2 Real Sequence
other transcendental functions are expressed as
Definition A real sequence is a function whose
infinite series. Many engineering problems are solved
domain of definition is a set of natural numbers and
numerically using infinite series. So, there is a need
whose range is a set of real numbers. Thus, a function
for a student of engineering to study infinite series.
s : N → S where S ⊆ R is called a real sequence.
Since the behaviour of an infinite series is related to
If s is a sequence, then s(n), the image of n
that of an infinite sequence, we begin our study with
under s, is called the nth term (or general term) of
sequences.
the sequence and is sometimes denoted by sn . The
Let S be a non-empty set of real numbers. We
sequence itself is denoted by
can form real sequences by choosing from S, a first
number a1 , a second number a2 , a third number a3 and ∞
so on. Here a1 , a2 , a3 , · · · , an are called the first, the < s1 , s2 , s3 , · · · , sn , · · · > or < sn >
n=1
second, the third, · · · , the nth term of the sequence. or s(1), s(2), s(3), · · · , s(n), · · · .
In the formation of a sequence, an element of S
may be chosen more than once as in To denote the sequence sn by the set notation
{sn } is not appropriate since a set contains distinct
2, 4, 2, 6, 2, 8, 2, 10, · · · elements whereas a sequence may contain elements
which are equal.
or the same number a may alone be chosen every
time in which case we have the constant sequence {1, 1, 1, · · ·, 1, · · · } = {1} is a singleton set while
1, 1, 1, · · ·, 1, · · · is an infinite constant sequence.
a, a, a, · · · , a, · · · But the notation {sn } is also used. The distinction
between a set and a sequence is to be made from the
To every natural number n ≥ 1 if there corre-
context.
sponds exactly one real number an , then we have an
Sequences are described by listing the terms, or
infinite sequence (Examples 9.1–9.6).
by stating a defining law or laws by means of an
If this correspondence exists only for each natural
analytical expression (Examples 9.1–9.6) or through
number n between 1 and N (1 ≤ n ≤ N ), then we
recursive formulas (Example 9.7) as illustrated.
obtain a finite sequence (Examples 9.8 and 9.9).
Thus, we can regard every sequence as a set of Example 9.1
ordered pairs of numbers (n, an ) whose first coor- The sequence of even positive integers.
dinate n is a natural number and whose second sn = 2n = 2, 4, 6, 8, · · · , 2n, · · · .
Example 9.2 (2) tn = −n|n ∈ N

The sequence
< = of
< perfect squares. = = · · ·, −6, −5, −4, −3, −2, −1
sn = n2 = 12 , 22 , 32 , · · · , n2 , · · · . > ?
1
(3) un = n∈N
Example
> 9.3? > ? >n ?
n 1 2 3 n 1 1 1 1
sn = = , , ,··· , ,··· · = 1, , , , · · · , , · · ·
n+1 2 3 4 n+1 2 3 4 n
Example 9.4< > ?
= 1
sn = 2n = 21 , 22 , 23 , · · · , 2n , · · · · (4) vn = K + K ∈ R, n ∈ N
> n+1 ?
1 1 1
= K + ,K + ,K + ,···
2 3 4
9.2.1 Alternating Sequences
Example 9.5 None of the terms of sn is greater than 1 and
none of the terms of tn is greater than –1. These
@ 1 A sequences are said to be bounded above and the
, if n is even numbers 1 and –1 are called upper bounds of sn
sn = n
1 and tn , respectively.
− , if n is odd
> n ? Again, none of the terms of un is less than 0
1 1 1 1 1 and none of the terms of vn are less than K. These
= −1, , − , , − , , · · · ·
2 3 4 5 6 sequences are said to be bounded below and the
numbers 0 and K are called lower bounds of un
Example
< 9.6 = and vn , respectively.
sn = (−1)
n−1
n = 1, −2, 3, −4, 5, −6, · · · ·
Definition A sequence sn is said to be bounded
above (below) if its range set is bounded above
9.2.2 Fibonacci Sequence (below), i.e., if there exists a real number A (B) such
Example 9.7 that
sn where s1 = 0, s2 = 1, sn ≤ A (sn ≥ B) for all n ∈ N .
sn = sn−1 + sn−2 (n ≥ 3). A sequence which is bounded above as well as
below is called a bounded sequence, i.e., a sequence
The sequence of perfect cubes below 100 is bounded if there exist real numbers A and B such
that
Example
< 9.8
= B ≤ sn ≤ A for all n ∈ N .
13 , 23 , 33 , 43
or, equivalently, if there exists a positive real number
Last element sN = 43 = 64 for N = 4.
M such that
The sequence of primes below 50 |sn | ≤ M for all n ∈ N .
Example 9.9
2, 3, 5, 7, 11, 13, 17, 19, 23, 29, 31, 37, 41, 43, 47 If A is an upper bound of a sequence, then any
number A1 greater than A is also an upper bound of
Last element sN = 47 for N = 15. the sequence.
If u is an upper bound of a sequence sn and any
9.3 Bounded Sequences real number less than u is not an upper bound of
Observe the following sequences: sn , then u is called the least upper bound (l.u.b.) or
supremum of sn . We write sup sn = l.u.b. sn = u.
> ?
1 2 3 n Similarly, if l is a lower bound of a sequence sn
(1) sn = , , , · · · , ,··· and any real number greater than l is not a lower
2 3 4 n+1
Sequences and Series 9-3
bound of sn , then l is called the greatest lower bound Proof: Let sn converge to the limit l. Then, for
(g.l.b.) or infimum of sn . every > 0 there exists a positive integer m such that
We write inf sn = g.l.b. sn = l. |sn − l| < for all n ≥ m
⇒ l − < sn < l + for all n ≥ m
Example 9.10
sn = 1 + (−1)n |n ∈ N Consider the finite set of numbers
s1 = 0, s2 = 2, s3 = 0, s4 = 2, · · · {s1 , s2 , · · · , sm−1 , l − , l + }. If k and K be
the least and the greatest of numbers in a set, then
Range of sn is the set {0, 2} which is a finite set. k ≤ sn ≤ K for all n ∈ N
sup sn = 2 and inf sn = 0. ∴ The sequence sn is bounded.
The converse of the theorem is not, in general,
Example
> 9.11 ? true. A bounded
1 < sequence
= may not be convergent.
sn = n∈N For example, (−1)n−1 = 1, −1, 1, −1, · · · is not
n convergent to a unique limit.
1 1
s1 = 1, s2 = , s3 = , · · ·
2 3 Example 9.13 > ?
Clearly, 0 < sn ≤ 1 for all n ∈ N . 1 ∈ sn and is 1
therefore the greatest number of sn ; 0 ∈ sn . So, Show that the sequence converges to the limit 0.
n
sn has no least number.
sup sn = 1 and inf sn = 0. Solution For any given > 0, we have
1 1 1
− 0 < whenever < , i.e., when n > .
Example
> 9.12 ? n n
n 1
sn = n∈N Let us choose a positive integer m > . Then for
n+1
1 2 3 all n ≥ m > ?
s1 = , s2 = , s3 = , · · · 1 1 1 1
2 3 4 we have −0 = < < ∴ converges
1 1 n n m n
Clearly, ≤ sn < 1 for all n ∈ N . ∈ sn and
2 2 to the limit 0. A sequence whose limit is zero is called
is therefore the least number of sn ; 1 ∈ sn so sn a null sequence.
has no greatest number.
Example 9.14
9.4 Convergent Sequences Show that the sequence sn
2n2 + 1
Definition Let l be a real number. A sequence sn where sn = 2 (n ∈ N ) converges to 1.
is said to converge to the number l if for every > 0 2n − 1
there exists a positive integer m such that |sn −l|< Solution Let > 0 be given. We have
or l − < sn < l + for all n ≥ m 2n2 + 1 2
The number l is called the limit of the sequence |sn − 1| = −1 = 2 <
2n2 − 1 2n − 1
sn , and symbolically we write 12
2+
sn → l as n → ∞ or lim sn = l if n >
n→∞ 2
or simply lim sn = l or sn → l
If we choose a positive integer m > (2 + )/2 ,
The positive integer m depends upon . The phrase
then for all n ≥ m we have |sn − 1| <
"|sn − l| < for all n ≥ m" implies that the absolute
value of the difference between sn and l can be made ∴ lim sn = 1.
smaller than from some stage onwards.
Example 9.15
Theorem 9.1 Every convergent sequence is Show that the sequence sn where sn = (−1)n−1 /n
bounded. converges to 0.
1 (3) oscillatory: (i) finitely or (ii) infinitely.

Solution We have sn = , if n is odd and
n
1
sn = − , if n is even. The sequences s2n−1 and A sequence sn is said to diverge to +∞ (−∞) if
n for every real number K > 0 there is a positive integer
s2n are subsequences of sn
m such that
> ? > ?
1 1 1 1 1
s2n−1 = 1, , , · · · , s2n = − , − , − , · · · . sn > K (sn < −K) for all n ≥ m
3 5 2 4 6
> ? > ? We write lim sn = ±∞ (or) sn → ±∞ as n → ∞.
1 1 1 1 1 n→∞
But 1, , , · · · and , , , · · · are sub-
3 5 2> ?4 6 Example 9.17
1 The sequence 3n diverges to +∞.
sequences of the sequence which converge to
n
0. ∴ Both s2n−1 and s2n converge to the same limit Solution Let K > 0 be given. We choose an m
0. Hence sn converges to 0. such that 3m > K. Now
Example 9.16 log K

√ 3m > K ⇔ m log 3 > log K ⇔ m >
Show that lim n
n = 1. log 3

log K
Solution Let Take m > + 1 where [x] is the largest
√ log 3
limn
n = 1 + hn where hn ≥ 0 integer not greater than x. Then3n > K for all n ≥ m
and hence 3n diverges to +∞.
⇒ n = (1 + hn )n
n(n − 1) 2 Example 9.18
= 1 + nhn + hn + · · · + hnn
2! The sequence −en diverges to −∞.
(Binomial theorem)
n(n − 1) 2 Solution Let K > 0 be given. We choose an m
> hn for all n ∈ N (∵ hn ≥ 0)
2 such that
2
⇒ h2n < for n ≥ 2 ⇒ hn −em < −K ⇔ em > K ⇔ m > log K
n−1
* Take m > [log K] + 1. Then − en < −K for all
2
= for n ≥ 2 n ≥ m and hence −en diverges to −∞.
n−1
Let > 0 be given. Then |hn | < 2/(n − 1) < 9.5.1 Oscillatory Sequence
2 2
if < 2 , i.e., if n > 2 + 1. If we choose m ∈ N Definition A sequence sn is said to oscillate if
n−1 it does not (1) converge to number l, (2) diverge to
2
such that m > 2 + 1, then we have |hn | < for all +∞ or (3) diverge to −∞.
√
n ≥ m, i.e., n n − 1 < for all n ≥ m.
√ Example 9.19
∴ lim n n = 1. 5 + (−1)n oscillates finitely between 4 and 6.
9.5 Divergent Sequences Example 9.20

(−1)n n = −1, 2, −3, 4, ... oscillates infinitely
A sequence which is not convergent is said to be
divergent. A divergent sequence has one of the between −∞ and +∞.
following types of behaviour:
Example 9.21
(1) divergent to +∞ (−1)n 3 = −3, 3, −3, 3, ... oscillates finitely
(2) divergent to −∞ between −3 and +3.
9.6 Monotonic Sequences Proof: Let sn be a monotonic increasing

sequence which is bounded above and let S =
9.6.1 Monotonic Increasing {sn |n ∈ N }. Then S is a non-empty bounded set.
Sequence By the order completeness axiom 1 of real numbers
Definition A sequence sn is said to be mono- sup S = l exists. We will show that lim sn = l.
tonically increasing (m.i.) or non-decreasing if Let > 0 be given. By the definition of l.u.b. S,
there exists an sk such that
sn ≤ sn+1 for all n ∈ N
The phrase strictly increasing is used if sn < sn+1 l − < sk
for all n.
Otherwise, l − could be an upper bound less than
Example 9.22 l. By monotonicity we see that if n ≥ k then sk ≤ sn .
1, 2, 2, 3, 3, 3, 4, 4, · · · is m.i (or) non-decreasing. Hence
Example 9.23 l − < sn ≤ l< l + for all n ≥ k
1, 2, 3, 4, 5, 6, · · · is monotonically strictly
⇒ lim sn = l
increasing.
Similarly, we can prove that a bounded m.d.
9.6.2 Monotonic Decreasing sequence sn converges to the g.l.b. of sn .
Sequence
Example 9.26
Definition A sequence sn is said to be mono- Prove that the sequence sn where
tonically decreasing (m.d.) or non-increasing if
1 1 1
sn ≥ sn+1 for all n ∈ N sn = 1 + + + · · · + is convergent.
1! 2! n!
The phrase strictly decreasing is used if sn > sn+1
Solution We have
for all n.
1
Example 9.24 sn+1 − sn = >0 for all n
> ? (n + 1)!
1 1 1 1 1
1, , , , , , · · · is monotonically decreasing ∴ sn is strictly increasing.
2 2 3 4 4
or non-increasing. 1 1 1 1 1 1
Also, < 2 , < 3 , · · · , < n−1 for n ≥ 3
Example 9.25 3! 2 4! 2 n! 2
> ? 1 1 1
1 1 1 1 ∴ sn < 1 + 1 + + 2 + · · · + n−1
1, , , , , · · · is monotonically strictly decreas- 2 2 2
2 3 4 5
ing. 1 − 21n 1
= 1+ = 3 − n−1 < 3
1 − 12 2
9.6.3 Monotonic Sequence
sn is a m.i. sequence which is bounded above.
Definition A sequence which either monotoni- Since a monotonic increasing sequence which is
cally increases or monotonically decreases is called a bounded above is convergent, sn is convergent.
monotonic sequence.
Example 9.27
9.6.4 Monotone Convergence n = sn
Prove that the<sequence
Theorem where sn = 1 + 1n is convergent.
A monotonic increasing sequence sn which is 1 Order Completeness Axiom: Let S ⊆ R, S = φ (empty set)
bounded above converges to the l.u.b. sn . and S be bounded above. Then S has an l.u.b S (or Sup S) in R.
Solution 1 1 1
< 1+1+ + + ··· +
2! 3! n!
1 n % n 1 (by binomial theorem)
1+ =
n r nr 1 1 1
< 1 + 1 + + 2 + · · · + n−1
%
n
n(n − 1) · · · (n − r + 1) 1 r 2 2 2
= 2+ 1 − 21n 1
r=2
r! n = 1+ = 3 − n−1 < 3
1− 2 1
2
%
n
1 1 2 r−1
= 2+ 1− 1− · · · 1− (1)
r=2
r! n n n sn is monotonically increasing which is bounded
above. Hence sn is convergent to the limit ‘e’.
We also note that
Theorem 9.2 A monotonically increasing
1 1 −k −k sequence which is not bounded above diverges to
> ⇒ < +∞ and a monotonically decreasing sequence which
n n+1 n n+1
is not bounded below diverges to −∞.
k k
⇒ 1− < 1−
n n+1
9.7 Cauchy Sequences
Putting k = 1, 2, 3, · · · (r − 1) in succession and
multiplying, we get A sequence sn is called a Cauchy sequence if for
each > 0 there exists an m ∈ N such that

1 2 r−1
1− 1− ··· 1 − |sn − sm | < for all n > m
n n n

1 2 or |sn+p − sn | < for all n > m and every p > 0
< 1− 1−
n+1 n+1 or |sp − sq | < whenever p ≥ m, q ≥ m(p, q ∈ N ).

r−1 or |sm+p − sm | < for all p > 0.
··· 1 − (2)
n+1
Example
> 9.28 ?
Using result (2) in (1), we have 1 1 1
1, , , · · · , , · · · is a Cauchy sequence.
2 3 n
n %n
1 1 1
1+ < 2+ 1− Let > 0 be given. If n > m then,
n r=2
r! n+1 1 1 n−m 1
− = <
2 r−1 n m nm m
1− ··· 1 − 1 1 1
n+1 n+1 Choosing m so that < , we have − <
n+1
1 >m ? n m
= 1+ 1
n+1 for all n > m. Hence is a Cauchy sequence.
n
Example 9.29
< 2=
∴ sn < sn+1 for all n ∈ N , i.e., sn is monotonically
The sequence n is not a Cauchy sequence.
increasing
If n > m then n2 − m2 = (n − m)(n + m) > 2m > 1
1 n n(n − 1) 1 n(n − 1)(n − 2)
1+ = 1+1+ + for any m.
n 2! n2 2!
1 n(n − 1)(n − 3) 1 Taking = 1 we cannot find a positive integer m
3
+ + ··· + n such that n2 − m2 < for all n > m.
n 3! n
9.8 Cauchy’s General Principle of (iv) We have to show that to each > 0 there
Convergence corresponds an m such that
(or Cauchy’s Convergence
|xn | < for n > m
Criterion)
Since 1/n → 1 while |x| < 1 there is a positive
Theorem 9.3 A sequence converges if and only
integer m such that |x| < 1/m
if it is a Cauchy sequence.
Important Limits ⇒ |xn | < |x|m < for n > m
Theorem 9.4 Prove that
x n
log n (v) Let an = 1 +
(i) lim =0 n
n x n x
√ 1 log an = log 1 + = n log 1 + →x
(ii) lim n n = 1 or lim n = 1
n n n
1
(iii) lim x = 1
n (x > 0) By the application of L’Hospital’s rule
(iv) lim xn = 0 (|x| < 1) x log (1 + x/n)
lim n log 1 + = lim
x n
n 1/n
(v) lim 1 + = ex (any x)
n 1
1+x/n
− nx2
xn = lim
(vi) lim =0 (any x) −1/n2
n!
x
log x = lim =x
Proof: (i) The function f (x) = is defined 1 + x/n
x
for all,x ≥ 1, and agrees with the sequence Thus (1 + x/n)n = an = elog an → ex ·
/
log n log n
on the positive integers. So, lim equals
n n |x|n xn |x|n
log x (vi) Since − ≤ ≤ , we have to show that
lim if the limit exists. n! n! n!
x |x|n
x→∞
log n log x 1/x 0 → 0.
Hence lim = lim = lim = =0 n!
n x→∞ x 1 1 First we choose an integer N > |x| so that
(by L’Hospital’s rule) n
|x| |x|
Note < 1 and →0
N N
When we use L’Hospital’s rule to find the limit of a
sequence we often treat n as a continuous real variable For n > N , we can write
and differentiate directly with respect to n.
|xn | |x|n
=
(ii) Let an = n 1/n
n! (N + 1)(N + 2) · · · n
1.2, · · · N ·
1 (n − N ) factors
log an = log n1/n = log n → 0
n |x|n |x|n N N N N |x| n
≤ = =
∴ n1/n = an = elog an → e0 = 1. N !N n−N N !N n N! N
N
|x|n
N N
|x|
(iii) Let an = x1/n then Thus 0 ≤ ≤ →0
1 n! N! N
log an = log x1/n = log x → 0
n |x|n
⇒ → 0.
x1/n = an = elog an → eo = 1. n!
Example 9.30 1
··· + does not converge, by showing that it is not
Apply Cauchy’s general principle of convergence to n
1 1 1 a Cauchy’s sequence.
show that sn where sn = 1 + 2 + 2 + · · · + 2
converges. 2 3 n Solution We have
1 1 1
Solution We have sn+p − sn = + + ··· + .
n+1 n+2 n+p
1 1 1 1
sn = + 2 + 2 + ··· + 2; Take n = p = m. For n ≥ m
12 2 3 n
1 1 1
let p ∈ N sn+p − sn = + + ··· +
m+1 m+2 m+m
1 1 1 1 1 1
Then sn+p = + 2 + 2 + ··· + 2 > m· = (1)
12 2 3 n 2m 2
1 1
+ + ··· + If the sequence sn is convergent, we must have
(n + 1)2 (n + p)2
sn+p − sn < for all n > m and p ∈ N
Now sn+p − sn 1
1 1 1 which fails to hold if < because of Eq. (1).
= + + ··· + 2
(n + 1) 2
(n + 2) 2
(n + p)2 ∴ sn is not a Cauchy sequence and hence is
1 1 not convergent. We state the following theorems
< + + ···
n (n + 1) (n + 1) (n + 2) without proof.
1
+ Theorem 9.5 (Cauchy’s first theorem on
(n + p − 1) (n + p) s1 +s2 +· · ·+sn
limits) If lim sn = l, then lim =l
1 1 1 1 n
= − + − + ···
n n+1 n+1 n+2
Theorem 9.6 (Cauchy’s second theorem on
1 1 limits) If {sn } is a sequence of positive terms such
+ −
n+p−1 n+p that lim sn = l then lim (s1 · s2 · · · sn )1/n = l.
1 1 1
= − < (1) Theorem 9.7 If {sn } is a sequence of positive
n n+p n sn+1
terms and lim = l then lim (sn )1/n = l.
1 sn
Let > 0 be given. Take m ∈ N such that m >
Example 9.32
1 1 Prove that
n≥m⇒ ≤ < (2)
n m 1 + 21/2 + 31/3 + · · · + n1/n
lim = 1.
From (1) and (2) for every > 0 there exists an n
m ∈ N such that
Solution Let sn = n1/n , we know that
sn+p − sn < for all n > m and p ∈ N lim n = 1.
1/n
By Cauchy’s first theorem on limits,

By Cauchy’s general principle of convergence, sn s1 + s2 + · · · + sn
converges. lim
n
1
Example 9.31 = lim 1 + 21/2 + 31/3 + · · · + n1/n
1 1 n
Prove that the sequence sn where sn = 1 + + + = 1.
2 3
Example 9.33 1
3. Show that sn defined by sn = 2 − n−1 is
Prove that 2
convergent.
1/n
2 3 2 4 3 n+1 n √
lim ··· = e. 4. Show that the sequence sn defined by s1 = 2
1 2 3 n
and sn+1 = 2 + sn converges to 2.
Solution Let 5. Show that the sequence sn defined by s1 = 1 and
n n sn+1 = 2 + sn for all n ∈ N is monotonically
n+1 1
sn = = 1+ → e. increasing and bounded, and hence convergent
n n
with the limit of 2.
By Cauchy’s second theorem on limits
6. Show that the sequence an
lim (s1 · s2 · · · sn )1/n

2 3 2 4 2 n+1 n (a) diverges to +∞ if a > 1,
= lim ··· = e.
1 2 3 n (b) converges if −1 < a ≤ 1,
(c) oscillates finitely if a = −1 and
Example 9.34
Prove that (d) oscillates infinitely if a < −1.
n
lim = e. √
(n!)1/n 7. Show that n
a = 1 if a > 0.
Solution √
nn (n + 1)n+1 8. Prove that the sequence sn defined by s1 = 7
Let sn = , sn+1 = and sn+1 = 7 + sn is bounded above and mono-
n! (n + 1)!
tonically increasing. Hence it converges to the
sn+1 (n + 1) n+1
n!
= . positive root of x2 − x − 7 = 0.
sn (n + 1)! nn

1 n n+1
= 1+ 9. If sn = , then show that sn converges to 1.
n n
By the above theorem, 10. Show that if sn = n2 + 5n then sn diverges to

+∞.
sn+1 1 n
lim sn1/n = lim = lim 1 + = e.
sn n 11. Show that
> ?
EXERCISE 9.1 n+1
(a) is convergent to 1,
n
> ?
1. Show that the sequence sn defined by 3n + 1 3
(b) is convergent to ,
5n + 3 5
1 1 1
sn = + +···+ converges. (c) n! is divergent to +∞,
n+1 n+2 n+n < =
(d) 1 − n2 is divergent to −∞,
2. Show that the sequence sn where < =
(e) 1 + (−1)n−1 oscillates (finitely) between
1 1 1 0 and 2,
s1 = 2, sn = 1 + + + ··· + < =
1! 2! (n − 1)! (f) (−1)n−1 n oscillates (infinitely) between
(n ≥ 2) converges. −∞ and +∞.
9.9 Infinite Series %

∞ %
∞ %
∞
(aun + bvn ) = a un + b vn .
We know how to find the sum of a finite number of n=1 n=1 n=1
real numbers. But how to find the sum of an unending
sequence of numbers? The theory of infinite series
deals with this impossible task by introducing the 9.9.1 Telescoping Series
concept of convergence of series. Let un and vn be two sequences $ such that un =
An infinite series is merely a special kind of an vn+1 − vn for n = 1, 2, 3, · · · . Then un converges
infinite sequence. Let un be a given sequence of if lim vn exists in which case, we have
real numbers. Form a new sequence sn as follows: n→∞
%
∞
s1 = u1 uk = lim vn − v1 .
n→∞
k=1
s2 = u 1 + u 2
s3 = u 1 + u 2 + u 3 Example 9.35
.. .. %∞
1
. . Show that the series n
converges.
n=0
3
sn = u1 + u2 + · · · + un
% n
Solution Here,
= uk (n = 1, 2, 3, · · · ) (9.1)
1
k=1 un =
3n
Definition The ordered pair of sequences sn = u 0 + u 1 + u 2 + · · · + un

(un , sn ) is called an infinite series. The number 1 1 1 1
sn is called the nth partial sum of the series. The = (3 − 1) + 1 − + −
2 3 3 32
series is said to converge or diverge according as sn
1 1
converges or diverges. The series defined by Eq. (9.1) +··· + − n
3n−1 3
is symbolically written as
1 1 3
= 3− n → as n → ∞
u 1 + u 2 + u 3 + · · · + un + · · · 2 3 2
or u1 + u2 + u3 + · · ·
%∞
1
%n
∴ converges to 3/2.
or uk n = 1, 2, 3, · · · 3n
n=0
k=1
% Example 9.36
or simply as uk
%
∞
1
Show that the series converges.
If the sequence sn defined by Eq. (9.1) converges n=1
n (n + 1)
to s, the number s is called the sum of the series and
we write Solution Here,
%∞
1 1 1
s= uk un = = −
k=1
n (n + 1) n n+1
$ sn = u 1 + u 1 + · · · + un
Thus for convergent series uk denotes both the
1 1 1 1 1
series and its sum. = 1− + − + −
2 2 3 3 4
$ $
Theorem 9.8 Let u = un and v = vn be 1 1
+··· + −
convergent series.
$ Then for every pair of constants a n n+1
and b, the series (aun + bvn ) converges to the sum 1
(au + bv). That is, = 1− → 1 as n → ∞
n+1
%
∞
1 The sequence sn and hence the series (9.2)
∴ converges to 1. 1
n=1
n (n + 1) converges to when −1 < r < 1.
1−r
9.10 Convergent, Divergent and Case (ii) r=1

Oscillatory Series
sn = 1 + 1 + · · · + 1 = n; lim sn = +∞
Let sn = u1 + u2 + u3 · · · + un be the nth partial sum
of the series The sequence sn and hence the series (9.2)
%∞ diverges to +∞ if r = 1.
un = u1 + u2 + u3 + · · · + un + · · ·
n=1 Case (iii) r = −1
/
If, as n → ∞, the sequence sn tends to 0 if n is even
$ sn = 1 − 1 + 1 − 1 + · · · =
(1) a finite limit l then the series un is said to be 1 if n is odd
convergent and l is called its sum;
The sequence sn and hence the series (9.2) oscil-
$ lates between 0 and 1 depending upon whether n is
(2) ±∞ then the series un is said to diverge to
±∞, as the case may be; even or odd if r = −1.
$
(3) more than one limit, the series un is said to Case (iv) r < −1
oscillate % % %
(i) finitely if the limits are finite and un = r n−1 = (−1)n−1 pn−1
(ii) infinitely if the limits are ±∞. where p = −r > 1
1 1
Note sn = (1 − r n ) = 1 − (−1)n pn
The sum of an infinite series is its limit sum and not 1−r 1+p
/
sum in the ordinary sense. Any one of the type of +∞ if n is odd
lim sn =
behaviour of a series is called the nature of the series. −∞ if n is even
% %
un = r n−1 oscillates infinitely between
9.11 Geometric Series
−∞ and +∞.
Theorem 9.9 Prove that the geometric series
(G.S.)
Case (v) r>1
% %
∞
1
un = r n−1 sn = (r n − 1)
r−1
n=1
1
= 1 + r + r 2 + · · · + r n−1 + · · · (9.2) lim sn = lim (r n − 1) = +∞
r−1
converges if 1 < r < 1 and diverges in all other cases. The sequence sn and hence the series (9.2)
diverges to +∞ if r > 1.
Solution Let sn denote the sum of the first n terms
of the series Example 9.37
Find sn and discuss the nature of the series.
Case (i) −1 < r < 1
% 1
1 − rn 1 1.
sn = , lim sn = n (n + 1) (n + 2)
1−r 1−r
%
[∵ lim r n = 0 (−1 < r < 1)] 2. (2n − 1)
%
3. (1 − n) (1 + n) s1 = 1 − 12 = 0,

% s 2 = 1 − 1 2 + 1 − 2 2 = 2 − 12 + 2 2 ,
4. (−1)n−1
s3 = 1 − 12 + 1 − 22 + 1 − 32

% = 3 − 1 2 + 2 2 + 32
5. (−1)n−1 n
sn = 1 − 1 2 + 1 − 2 2 + 1 − 3 2
Solution (Telescoping series)
% 1 + · · · + 1 − n2
1.
n (n + 1) (n + 2) = n − 12 + 22 + 32 + · · · + n2
% 1 n (n + 1) (2n + 1)
un = = n−
n (n + 1) (n + 2) 6
1 1 1
= + + + ···
1·2·3 2·3·4 3·4·5 n3
1 When n is large sn behaves like − → −∞ as
+ + ··· (1) n → ∞. 3
n + 1) (n + 2)
(n
∴ lim sn = −∞.
1 1 1
s1 = , s2 = + ,··· Hence the series (1) diverges to −∞.
1·2·3 1·2·3 2·3·4
1 1 %
sn = + + ··· 4. (−1)n−1
1·2·3 2·3·4
+
1 % %
n (n + 1) (n + 2) un = (−1)n−1 = 1 − 1 + 1 − 1 + · · · (1)

1 1 1 1 s1 = 1, s2 = 1 − 1 = 0, s3 = 1 − 1 + 1 = 1,
=2 − + −
1·2 2·3 2·3 3·4 sn = 1 − 1 + 1 − 1 + · · ·
/
1 1 1 if n is odd
+··· + − =
n(n + 1) (n + 1) (n + 2) 0 if n is even
/
2 1 if n is odd
= 1− lim sn = 1 lim sn =
(n + 1) (n + 2) 0 if n is even
Hence the series (1) converges to 1.
% Hence the series (1) oscillates between 0 and 1.
2. (2n − 1) %
5. (−1)n−1 n
%
un = 1 + 3 + 5 + · · · + (2n + 1) + · · · (1) % %
un = (−1)n−1 n
s1 = 1, s2 = 1 + 3 = 2 , 2
s3 = 1 + 3 + 5 = 3 2 , · · · = 1−2+3−4+· · ·+(−1)n−1 n+· · · (1)

sn = 1 + 3 + 5 + · · · + (2n − 1) = n2 ; 1+1
s1 =1=
2
lim sn = ∞
2
s2 = 1 − 2 = −1 = − ,
Hence the series (1) diverges to +∞. 2
% 3+1
3. (1 − n) (1 + n) s3 = 1−2+3=2= ,
% % % 2
un = (1 − n) (1 + n) = 1 − n2 4
s4 = 1 − 2 + 3 − 4 = −2 = −
= 1 − 12 + 1 − 22 + 1 − 3 2 2
5+1
+ · · · + 1 − n2 + · · · (1) s5 = 1−2+3−4+5=3=
2
6 Note
s6 = 1 − 2 + 3 − 4 + 5 − 6 = −3 = −
⎧ 2 The converse of the zero test is false. That is,
⎪
⎨ n + 1 lim un = 0 does not imply the convergence of the
if n is odd
sn = 2 series. The condition that lim un = 0 is only
$ neces-
⎪ n
⎩ − if n is even sary but not sufficient for convergence of un ; the
2 series may converge or diverge.
$1
lim sn = +∞ or −∞ according as n is odd or The harmonic series is a counter-example
n
even. The series (1) oscillates between −∞ and +∞. since lim un = lim 1n = 0 but we know that the
$1
series n
diverges to +∞ proved in Theorem on
9.12 Properties of Series hyper-harmonic series in Section 9.16.
The nature of a series remains unaffected by
However, the zero test is useful in establishing
$
(1) addition or deletion of a finite number of terms; divergence of a series. Thus, if lim un = 0 then un
is divergent. For example, consider
(2) multiplying or dividing each term by a finite non-
zero number, and in particular, by changing the % %n+1
un =
sign of each term of a series with appropriate n
change of signs of limits. 2 3 4 n+1
= + + + ··· + + ···
$ $ 1 2 3 n
$ un →
(i) If $ A, vn → B then
n+1
a un + b vn → aA + bB Here lim un = lim
$ n
(ii) $
If un diverges and c ∈ R, c = 0 then 1
cun diverges = lim 1 + = 1 = 0
n
$ $
(iii) If both un and vn diverge to +∞ or The condition for convergence fails to hold.
−∞ each series$ having all its terms with the %n+1
∴ is divergent.
same sign then (un + vn ) diverges to +∞ n
or −∞, as the case may be.
9.14 Cauchy’s General Principle
9.13 Zero Test—A Necessary of Convergence
Condition for Convergence In most cases, it is not possible to find the sum of
$ the nth partial sum sn and its limit. Cauchy’s gen-
Theorem 9.10 If un converges then un → 0
as n → ∞ eral principle of convergence gives a criterion for the
(or) A$necessary condition for the convergence of convergence of a series, without a knowledge of the
a series un is that lim un = 0. limit of the sequence of its partial sums sn .
$
(or) For every convergent series un , we must
Theorem 9.11 $ A necessary and sufficient con-
have lim un = 0.
dition for a series un to converge is that for each
$
Proof: Let un be convergent and suppose ∈> 0 there exists a positive integer m such that
sn = u1 + u2 + · · · + un |um+1 + um+2 + · · · + un | < ∈ for all n > m
is the nth partial sum of the series so that or up+1 + up+2 + · · · + uq < ∈
lim sn = l exists lim un = lim (sn − sn−1 ) for all q ≥ p ≥ m
= lim sn − lim sn−1 = l − l = 0 or un+1 + un+2 + · · · + un+p < ∈

(∵ both sn and sn−1 tend to l as n → ∞) for all n ≥ m, p ≥ 0
$
Proof: $ sn be the sequence of partial sums of
Let Theorem 9.12 The series un of positive
the series un terms, converges or diverges to +∞ according as
% the sequence sn of its nth partial sum is bounded
un converges ⇔ sn converges or not.
⇔ |sn − sm | <∈ for all n > m
Proof: By hypothesis, un > 0 for all n ∈ N
⇔ |um+1 + um+2 + · · · + un | < ∈
for all n > m ∴ sn+1 − sn = un+1 > 0 for all n ∈ N
⇒ sn+1 > sn for all n ∈ N
Hence the theorem.
⇒ sn is an increasing sequence .
9.15 Tests for Convergence and If sn is bounded then by monotone $ conver-
Divergence of Series gence theorem lim sn = l exists and hence un is
convergent.
With a view to discuss tests for convergence and
divergence of infinite series, we classify series as $If sn is not bounded then lim sn = +∞ and hence
un diverges to +∞.
follows.
Note
Numerical series In the case of series of positive terms, diverges means
(1) Series with same sign (series with positive or neg- diverges to +∞.
ative terms). Discussion is confined to positive
(term)
% series 9.16.1 Hyper harmonic series
un = u1+u3+u3+· · · with un > 0 for all n ∈ N (p-series or auxiliary series)
Theorem 9.13 The series
(2) Series
% with alternating signs (alternating series) % 1 1 1 1 1
(−1)n−1 un = u1 −u2 −u3 −u4 +· · · un > 0 = p + p + p + ··· + p + ··· (9.3)
n p 1 2 3 n
for all n ∈ N
(p ∈ R) converges if p > 1 and diverges if p ≤ 1.
(3) Series with mixed signs
Proof: Let sn denote the nth partial sum of the
% +1 for some n ∈ N series (9.3) so that
ain un where ain =
−1 for other n ∈ N 1 1 1 1
sn = + p + p + ··· + p
1p 2 3 n
Series involving Functions
$ $ Case (i) p > 1
(1) Power series: cn xn or cn (x − a)n , n ∈ N.
1
= 1 (1 term)
1p
1 1 1 1 2 1
p + p < p + p = p = p−1 (2 terms)
9.16 Series of Positive Terms
2 3 2 2 2 2
Series of positive terms are of practical importance.
$ 1 1 1 1 1 1 1 1
If sn be the nth partial sum of the series un of + p + p + p 0 for all n ∈ N 4 2p−1
⇒ sn+1 > sn for all n ∈ N ··· ··· ···

1 1 1
+ n + ··· + p
⇒ sn is an increasing sequence. (2n )p (2 + 1)p 2 −1
n+1
1 % 1 %1
< n (2n terms) dominates the series which is diver-
2p−1 np n
gent.
% 1
2n+1 − 1 ∴ (p < 1) is divergent by comparison test.
Number of terms 1 + 2 + 22 + · · · + 2n = np
= 2n+1 − 1 2−1
9.17 Comparison Tests (Type I)

Adding the (2n+1 − 1) terms on either side of the
inequality, An important and useful technique for investigating
1 1 1 the nature of a series of positive terms is to com-
s(2n+1 −1) < 1 + p−1 + 2 + · · · + p−1 n pare it with a series whose convergence or divergence
2 2p−1 2
n+1
is known. With an appropriate choice of r and p,
1− 1 $ n $ 1
2p−1 1 the geometric series r and the p-series
= < np
1− 2p−1
1
1 − 2p−1
1
discussed above are generally used for comparison
for all n ∈ N series.
Since n < 2n < (2n+1 − 1) for each n $ $
Theorem 9.14 Let un and vn be two series
1 of positive terms and suppose there exists a positive
Therefore sn < s2n < s(2n+1 −1) < real number k and an m ∈ N such that
1 − 2p−1
1
1 (i) un ≤ kv$n for all n ≥ m $

Since > 0 for each n, the sequence sn is mono- Then vn is convergent ⇒ un is
np
tonically increasing which is bounded above. By convergent.
monotone convergence theorem, sn is convergent. (ii) un ≥$ kvn for all n ≥ m $
% 1
Hence is convergent. Then vn is divergent ⇒ un is divergent.
np
Case (ii) p = 1 Example 9.38
%1 %∞
1
Suppose is convergent. Then Test for convergence the series .
n n=1
nn
lim (s2n − sn ) = lim s2n − lim sn = l − l = 0
1
(∵ lim s2n = l and lim sn = l) Solution Let un = >0 for all n
Now, nn
1 1 1 We have nn > 2n for all n ≤ 3

s2n − sn = + +· · ·+ (n terms) 1 1
n+1 n+2 n+n ⇒ < n
1 1 1 1 1 nn 2
> + + ··· + =n· =
2n 2n 2n 2n 2 1 % % 1
Take vn = n
, vn = is a geometric
1 2 2n
s2n − sn > for any n. 1
2 series with common ratio = < 1 and is therefore
2
This violates Cauchy’s general principle of con- convergent.
vergence. % 1
%1 ∴ By comparison test, is also convergent.
Hence diverges. nn
n
Example 9.39
Case (iii) p < 1 % 1
1 1
< 2 for all n ∈ N and being
1 1 3n2 +1 n n2
p
> for all n ∈ N p-series with p = 2 > 1, is convergent.
n n
% 1
∴ is convergent by comparison test. 9.18 Comparison Tests (Type II)
3n2 +1
$ $
Example 9.40 Theorem 9.15 Let un and vn be two
1 1 %1 series of positive terms. Suppose that there exists an
> for all n ≥ 2 and being p-series un+1 vn+1
log n n n m ∈ N such that < for all n ≥ m. Then
with p = 1 is divergent. un vn
% %
% 1 (i) vn converges ⇒ un converges
∴ is divergent by comparison test. % %
log n
(ii) un diverges ⇒ vn diverges
Example 9.41 (proof omitted)
n2 1 % 1
< for all n ∈ N and being p-series $
n4 + 1 n2 n2 Theorem 9.16 If un ≥ 0 and un is
% un
with p = 2 > 1 is convergent. convergent then is convergent.
% 1 + un
n2
is convergent by comparison test. Proof: un ≥ 0 ⇒ 1 + un > 1 ∀ n
n4 + 1
1
Example 9.42 <1 ∀n
1 + un
Discuss the convergence of the series un
% n3 ⇒ 0< < un ∀ n
· % un un
1 + %
n5 + 5n4 + 7 ⇒ converges since un converges.
1 + un
Solution
n3 n3 1 Example 9.44
< = 2 % 1 % 1
n + 5n + 7
5 4 n5 n converges ⇒ converges
% 1 n
2
1 + n2
2
for all n and converges 1/n 1
n2 ∵ = (by Theorem 2).
% 3 1/n + 1
2 1 + n2
n
⇒ converges.
n + 5n4 + 7
5 Example 9.45
% 1 · 3 · 5 · · · (2n − 1)
Example 9.43 Test for convergence .
$ 2 · 4 · 6 · · · (2n)
If n ≥ 0 for all n and
$ u√ un converges then
un /n converges. Solution
1 · 3 · 5 · · · (2n − 1)
Solution Let un =
un 2un un 1 2 · 4 · 6 · · · 2n
< < un
2
+ 2 + un+1 1 · 3 · 5 · · · (2n + 1) 2 · 4 · 6 · · · (2n)
n2 n2 n2 n4 = ·
2 un 2 · 4 · 6 · · · (2n + 2) 1 · 3 · 5 · · · (2n − 1)
1
= un + 2 for all n 2n + 1
n =
√ 2n + 2
un 1
or 0 ≤ < un + 2 (∵ un ≥ 0)
n n 1 vn+1 2n
% % 1 Take vn = ⇒ =
Since un and converge therefore the n vn 2n + 2
%
n2 un+1 vn+1
1 > ∀n
sum series un + 2 converges. By compari- un vn
n $ $
% √un Since vn diverges un also diverges (by
son test, converges. Theorem 9.1).
n
9.19 Limit Form Of Comparison 1

(iii) un = >0 ∀n
Theorem nn
un $ We know that nn > 2n ∀ n ≥ 3
Theorem 9.17 lim = l = 0 ⇒ both un 1 1
$ v n ⇒ n < n ∀n≤3
and vn converge or diverge together. n 2
This theorem is most useful for testing series of 1
positive terms for convergence. Take vn = n . We have un ≤ vn for all n and
$ $ 2
For a given
$ series un to choose the compari- vn is convergent, being a geometric series
son series vn with known behaviour, the following 1
with common ratio = < 1.
points may kept in mind. 2 $
1 By Comparison Test I, the series un is con-
(1) Express un as a series in powers of and
n vergent.
1
retain the term of the highest power of as a first Example 9.48
n
approximation for vn . %
∞
2n + 1
Test for convergence .
(n + 1) (n + 2)
Example 9.46 n=0
1/2
un = n2 + 1 − n = n 2 + 1 −n Solution Let
1/2
1 2n + 1
=n 1+ 2 −n un = >0 for all n ∈ N
n (n + 1) (n + 2)
1 1 1 1
= n 1 + · 2 − · 4 + ··· − 1 1 % %1
2 n 8 n Take vn = . We know that vn = is
expanding by Binomial theorem n n
divergent.
1 1 1 1 1 1
= · − · 3 + · · · ∼ , for large n un 2n + 1
2 n 8 n 2 n Now lim = lim n.
1 un 1 vn (n + 1) (n + 2)
Take vn = lim = = 0
n vn 2 2 + 1n
$ $ = lim
By Comparison Test, un diverges since vn 1 + 1n 1 + 2n
diverges. = 2 = 0
Example 9.47 ∴ By comparison test,

% % 2n + 1
1 1 1 1 un = is divergent.
(i) un = sin = − + ··· ∼ (n + 1) (n + 2)
n n 3!n3 n
for large n
Example 9.49
1 un
Take vn = lim = 1 = 0 %
∞
n vn Test for convergence n3 + 1 − n3 − 1 .
$ $
By Comparison Test, un diverges since vn n=1
diverges.
Solution Let
f (n) np + · · · 1
(ii) If un = = q ∼ q−p un = n3 + 1 − n3 − 1
g (n) n + ··· n
1 un n3 + 1 − n3 − 1
Take vn = q−p lim =1 = √ √
n vn
$ $ n3 + 1 + n3 − 1
By Comparison Test, $un and vn converges
or diverges together. vn converges for q > 2
(p + 1) and diverges for q ≤ (p + 1). = √ √ >0 (for all n)
n3 + 1 + n3 − 1
1 % 1 % 1
Take vn = . Note that is convergent is convergent being a geometric series with
n3/2 n3/2 3n
(p-series with p = 3/2 > 1) 1
common ratio < 1.
3
un 2n3/2
lim = lim √ √ un 1
vn n3 + 1 + n3 − 1 lim = lim = 1 = 0
vn 1 + (2/3)n
2 $
= lim + + = 1 = 0 By Comparison Test, un is convergent.
1 + n13 + 1 − n13
Example 9.52
∞ √
$ √ % 1 n + 1 n
By Comparison Test, n3 + 1 − n3 − 1 is Test for convergence .
n=1 n2 n + 2
convergent.
Solution Let
Example 9.50
%
∞ 1 n+1
1 un =
Test for convergence sin . n2 n+2
n
1 % 1
n=1
Take vn = , is convergent being p-series
Solution Let n2 n2
with p = 2 > 1.
1 1
un = sin > 0 for all n since ∈ (0, 1] ⊂ (0, π/2) un n+1 n
n n lim = lim
vn n+2
1
1 1 1 1 = lim
un = − 3
+ ··· 1 n
1 + n+1
n 3! n n
1 1 1
∵ is very small for large n = lim · 1 +
n3 1+ 1
n+1
n+1
n+1
1 % %1 1
Take vn = , vn = is divergent = = 0
n n e
$
sin 1 By Comparison Test, un is convergent.
un
lim = lim 1 n
vn Example 9.53
n

sin p 1 %
= lim = 1 = 0 =p n2 + 1 − n2 − 1
p→0 p n
% 1 Solution Let
By comparison theorem, sin is divergent.
n
un =
n2 + 1 − n2 − 1
Example 9.51 2
%
∞ = √ √
1 n2 + 1 + n2 − 1
Test for convergence the series . √
2n + 3n √ √ √
n=1
∵ n2 +1+ n2 −1 n2 +1− n2 −1 = 2
Solution Let
2
1 1 = + +
un = = n
n 1+ 1
+ 1− 1
2 +3
n n
3 1 + (3/2)n n2 n2
1 1
Take vn = n Take vn =
3 n
un 2 n Example 9.56
lim = lim + + ·
n→∞ vn n→∞ 1
n 1+ 1
n2
+ 1− 1
n2 % 1
= 1 = 0 2n + 3n
% %1 $ Solution Let
Since vn =is divergent so is un by
n 1 1
limit form of comparison test. un = = n
2n + 3n 3 [1 + (2/3)n ]
Example 9.54 1 % 1
% Take vn = n
; is a geometric series with
3 3 3n
n3 + 1 − n 2 $
common ratio < 1 and so convergent. ∴ un is
3
Solution Let convergent.
un =
3
n3 + 1 − n > 0 ∀ n EXERCISE 9.2
1 Test for convergence:
1 3
Also, un = n 1+ 3 −n
n % n√n
1 1. .
1 3 n4 + 1
= n 1+ 3 −1 Ans: convergent
n
% n2 + 1
1 2. .
= n 1+ 3 + · · · − 1 (n + 1)2
3n
% % 1 Ans: divergent
1 √
Take vn =
n 2
; vn =
n2
is convergent % n + 1 − √n
3. .
n 3n13 + · · · n
un 1 Ans: convergent
lim = lim = = 0
n→∞ vn 1
3 % 1/2
n2
4. n3 + 1 − n3/2 .
$
∴ un is convergent. Ans: convergent
% (n + 1) (n + 2)
Example 9.55 5. .
n (n + 3)
% 1 Ans: divergent
tan
n % (n + 1)2
Solution Let 6. .
n7/2
Ans: convergent
1
un = tan 1 3 5
n 7. + + + ···.
1 1 1·2·3 2·3·4 3·4·5
∈ (0, 1) ⊂ (0, π/2) ∀ n, tan > 0 ∀ n Ans: convergent
n n * * *
1 % 1 2 3
Take vn = ; vn is divergent 8. 3
+ 3
+ + ···.
n 2 3 43
Ans: divergent
√
% n + 1 − √n
1 1
un tan tan
lim = lim 1 n
= lim 1
n
= 1 = 0 9. .
n→∞ vn n→∞ 1 →0
n n n np
$ 1 1
∴ un is divergent. Ans: convergent if p > , divergent if p ≤ .
2 6
%
1 1
10. log 1 + 2 . (d) log 1 + .
n n
Ans: convergent Ans: divergent
EXERCISE 9.3 1
(e) .
1. Test for convergence of the series whose nth term nα+β/n
/
converges if α>1
un = √ Ans: , and
n diverges if α≤1
(a) 2 .
n +1 np
Ans: convergent (f) .
(n/+ 1)q
n converges if q > (p + 1)
(b) 2 √ . Ans: , and
n + n diverges if q ≤ (p + 1)
Ans: divergent
√ √
n (n + 1) (g) /np + 1 − np − 1.
(c) . converges if p > 2)
(n + 2)2 Ans:
Ans: divergent diverges if p≤2
(n + 2) (n + 3)
(d) . 9.20 Cauchy’s nth Root Test
(n + 1)2
Ans: divergent
$
2. Test for convergence of the series Theorem 9.18 If
un is a series of positive
$
1
terms such that lim un = l then un
n
1 3 5 n→∞
(a) + + + ···.
1·2·3 2·3·4 3·4·5 (i) converges if l < 1,
Ans: convergent
2·1 4·3 6·5 (ii) diverges if l > 1 and
(b) + 2 2 + 2 2 + ···.
3 ·4
2 2 5 ·6 7 ·8 (iii) the test fails if l = 1.
Ans: convergent
* * *
1
1 2 3
(c) + + + ···. Proof: Since un > 0 for all n ∈ N and unn is the
23 33 43 1
Ans: convergent positive nth root of un , lim un = l ≥ 0. n
√ √ √ n→∞
2−1 3−1 4−1 For each > 0 there exists an m ∈ N such that,
(d) + 2 + 2 + ···. for all n ≥ m,
32 − 1 4 −1 5 −1
Ans: convergent 1
unn − l < or (l − )n < un < (l + )n (9.4)
3. Test for convergence of the series whose nth term
un = √ √
n+2− n+1 1−l
(a) . Case (i) l < 1 Take = so that r = l + =
np 2
1 1 1−l 1+l
Ans: p > convergent, p ≤ divergent
l+ = < 1.
2 2 2 2
From the R.H.S. inequality of (9.4), for n ≥ m,
1 1
(b) √ sin .
n n un < (l + )n = r n
Ans: convergent
$ $
1 By comparison test, un is convergent since r n
(c) . is a geometric series with common ratio r < 1 and
n2
log n
Ans: convergent is convergent.
l−1 1 1
Case (ii) l > 1 Take = so that =
= <1
1 n
2 lim 1 + n e
n→∞
l−1 l+1 $
r = (l − ) = l − = > 1. By Cauchy’s root test, un is convergent.
2 2
From the L.H.S. inequality of Eq. (9.4), for n ≥ m,
1n
xn
(ii) lim un1/n = lim
n→∞ (n + 1)n
un > (l − )n = r n lim un ≥ lim r n = 0 n→∞
x
$ = lim =0<1
n→∞ n + 1
By zero test, un is divergent. $
By Cauchy’s root test, un is convergent.
1
Case (iii) l = 1 Let un = ,
np −n3/2 n1
1
1n (iii) lim un1/n = lim 1+ √
1 1 n→∞ n→∞ n
lim un = lim
n
n→∞ n→∞ np 1 1
= 1
= <1
1 e
= 1 p = 1 = 0 for p = 1, 2 lim 1 + 1
n2
lim n n n1/2
n→∞ $
$1 $ 1
We know that is divergent and is
n $ n2 n+1 −n ) n1
convergent. Thus if lim un = 1, un may diverge
1/n
n + 1 n + 1
n→∞ (iv) lim un1/n = lim −
or converge. In this case, the test fails and a further n→∞ n→∞ n n
investigation is required to settle the question of 1
=
convergence of the series. n+1
lim n+1 n
− n+1
n
n→∞
Example 9.57 1
$ = n+1 n
Test for convergence of the series un lim n
1 + 1n − 1
where un = 1
= < 1 (∵ e > 2)
−n2 e−1
(i) 1 + 1n [JNTU 2001S, 1998S, 1996] $
xn
(ii) [JNTU 1994] −n
(n + 1)n 1 1
(v) lim unn = lim 1− =e>1
3/2 n→∞ n→∞ n
1 −n $
(iii) 1 + √ [JNTU 1997S] By Cauchy’s root test, un is divergent.
n
−n
n + 1 n+1 n + 1 EXERCISE 9.4
(iv) − [JNTU 1985S]
n n Test the following series for convergence:
%∞
−n2 1. 2n /n3
n−1
(v) n=0
n Ans: divergent
Solution %∞
1
−n2 1n 2.
1 n n
(i) lim un1/n = lim 1+ n=1
n→∞ n→∞ n Ans: convergent
%
∞
1 9.21 D’Alembert’s Ratio Test
3. n
log n $
n=2 Theorem 9.19 Let un be a series of positive
Ans: convergent un+1 $
terms such that lim = l. Then un
∞
% 2 un
n−1 n (i) converges if l < 1,
4.
n
n=2
Ans: convergent (ii) diverges to +∞ if l > 1,
x x2 x3 and (iii) the test fails if l = 1.

5. 1 + + 2 + 3 + · · · (x > 0). un+1
2 3 4 Proof: We have un > 0 for all n ∈ N , >0
x xn un
[Hint: u0 = 1, u1 = , · · · un = un+1
2 (n + 1)n and lim =l≥0
x un
un =
1/n
→ 0 < 1]
n+1 For each > 0, there exists an m ∈ N such that
Ans: convergent for all n ≥ m,
%∞
nx n un+1 un+1
6. − l < or (l − ) < < (l + ) (9.5)
n=1
1+n un un
Ans: x < 1, convergent x > 1, divergent, x = 1,
Replacing n by (n − 1), (n − 2), · · · , (m + 1), m
divergent (zero test)
and multiplying the corresponding members we get
%
∞
{(n + 1) x}n from Eq. (9.5), for all n ≥ m,
7.
nn+1 un
n=1 (l − )n−m < < (l + )n−m
1 (n+1) x 1+ 1n x um
Hint: un =
n
1
= 1
→ x, x = 1, or a (l − )n < un < b (l + )n (9.6)
n·n
nn nn
1 1 1 where a = um / (l − ) and b = um / (l + ) .
m m
un = . 1 + , compare with vn = ,
n n n
un 1 n % Case (i) l < 1 Take = 1−l 2
so that r = (l + ) =
divergent = 1+ → e = 0 un l + 1−l
2
= 1+l
2
< 1. From the R.H.S. inequality of Eq.
vn n
(9.6),$ we have un 1 divergent, x = 1
test, un is convergent since r n is convergent,
divergent (comp. test)
being a geometric series with common ratio r = (l +
% 1 + nx n ) < 1.
8.
n Case (ii) l > 1 Take = l−1 so that r = (l − ) =
1 1 + nx 1 2
Hint: un =
n
= + x → x x = 1, l − l−1
2
= 1+l
2
> 1. From the L.H.S. inequality of Eq.
n n 1 n n (9.6),$we have un > ar n for n ≥ m.
un = 1+n = 1 + n → e = 0 $By comparison
n test, un is divergent to +∞ since r n is divergent,
Ans: x < 1, convergent 1 x > 1, divergent x = 1
being a geometric series with common ratio r = (l −
divergent (zero test)
) > 1.
% 2
1 −n Case (iii) l = 1 We know that
$ 1
is divergent
9. 1+ np
n for p = 1 and convergent for p = 2.
Ans: convergent un+1 np 1
But = = p → 1 as n → ∞ for
un (n + 1)p 1+ 1n
% 3/2
1 −n both p = 1 and 2. Therefore, the test is not conclusive
10. 1+ √
n if l = 1 and a further investigation is required to
Ans: convergent settle the question of convergence of the series.
$
Theorem 9.20 If un is a series of positive Example 9.60
un+1 $ Discuss the convergence and divergence of the series
terms such that → ∞ as n → ∞, then un
un
diverges to +∞. %
∞
xn
un+1 (x > 0, a > 0).
Proof: Since → ∞ as n → ∞, therefore xn + an
un n=1
un+1
there exists an m ∈ N such that > 2 for all
un Solution Here,
n ≥ m.
Replacing n by n − 1, n − 2, · · · , (m + 1), m and xn xn+1
multiplying the respective terms of the (n − m) un = > 0, ∀ n; un+1 =
xn + an xn+1 + an+1
inequalities, we get
x
um um Case (i) x < a ⇒ 0 < < 1, lim (x/a)n = 0
un > m 2n or un > k2n where k = m a
2 2
$ n un+1 xn+1 x n + an
Now 2 is a geometric series with common$ ratio ∴ = n+1 ·
2 > 1 and is divergent. By comparison test, un un x +a n+1 xn
x n
diverges. x +1 x
= · an+1 → <1
a x +1 a
Example 9.58 a
%∞
n!
Test the series for convergence. By ratio test, the series is convergent.
nn
n=1
an 1 1
Case (ii) x = a un = = un → = 0
Solution Here, + anan 2 2
By zero test, the series is divergent.
n! (n + 1)!
un = , un+1 = a a n
nn (n + 1)n+1 Case (iii) x > a 0 < < 1 and →0
un+1 (n + 1)! nn x x
∴ = ·
un (n + 1)n+1 n! xn 1
n un = = n → 1 = 0
n 1 1 x +a
n n
1 + ax
= = n → < 1.
n+1 1 + 1n e
By zero test, the series is divergent.
By ratio test, the series is convergent.
Example 9.61
Example 9.59 %
∞
(10 + 5i)n
%
∞
n2 Test the series for convergence.
Test the series for convergence. n=0
n!
n=1
2n [JNTU 1993]
Solution Here, Solution
n2 (n + 1)2 10 + 5i
un = , un+1 = u0 = 1, u1 = ,··· ,
2n 2n+1 1!
un+1 (n + 1) 2
2 n (10 + 5i) n
∴ = · 2 un = ,
un 2n+1 n n!
2 (10 + 5i)n+1
1 1 1 un+1 =
= 1+ → < 1. (n + 1)!
2 n 2
un+1 10 + 5i
By ratio test, the series is convergent. = →0<1
un n+1
By ratio test, the series is convergent. Solution

x x2
Example 9.62 u0 = 1, u1 = , u2 = ,··· ,
2p 3 p 4 p np 2 22 + 1
Show that the series 1+ + + +· · ·+ +· · · xn
2! 3! 4! n! un = > 0 for all n
is convergent for all p. [JNTU 1993] n2 + 1
Solution un+1 n2 + 1 n2 + 1
np (n + 1)p ∴ = · x = x
Here un = > 0, un+1 = un (n + 1)2 + 1 n2 + 2n + 2
n! (n + 1)!
p 1 + n12
un+1 1 1 = x→x
= 1+ → 0 < 1 for all p 1 + 2 · 1n + 2 · n12
un n+1 n
$
By ratio test, the series is convergent. By ratio test, un converges if x < 1 diverges if
x > 1 and the test fails if x = 1.
$ $ 1
Example 9.63 When x = 1, the series becomes un = .
% 2n − 2 $ $ 1
n2 +1
Test the series xn (x > 0) for conver- Compare the series with vn = , which is
2n + 1 n2
gence. convergent, being p-series with p = 2 > 1.
Solution Here un n2 1
lim = lim 2 = lim = 1 = 0
vn n +1 1 + 1/n2
2n − 2 n $
un = x > 0 for all n; By comparison test, un is also convergent.
2n + 1
2n+1 − 2 n+1
un+1 = n+1 x Example 9.65
2 +1 Discuss the convergence of the series
un+1 2n+1 − 2 2n + 1
∴ = n+1 · ·x 1 x x2 xn−1
un 2 + 1 2n − 2 p + p + p + ··· + p + · · · (x > 0).
1 3 5 (2n − 1)
1 − 21n 1 + 21n
= · ·x →x
1 + 2n+1 1 − 2n−1
1 1 Solution Here
$ xn−1
By ratio test, un converges if x < 1, diverges to un = > 0 for all n;
(2n − 1)p
+∞ if x > 1, and the test is inconclusive if x = 1.
xn
% 2n − 2 un+1 =
(2n + 1)p
When x = 1 the series becomes with
2n + 1 un+1 2n − 1 p
2 −2
n ∴ = ·x
un = un 2n + 1
2n + 1 p
2n − 2 2 − 1n
lim un = lim = p · x → x
2n + 1 2 + 1n
1 − 2n−1
1
$
= lim = 1 = 0 By ratio test, un converges if x < 1 and diverges
+ 21n
1$
to +∞ if x > 1.
By zero test, un diverges if x = 1.
The test fails if x = 1. In this case, the series is
Example 9.64 1 1 1 1
Test for convergence, the following series p + p + p + ··· + p + ···
1 3 5 (2n − 1)
x x2 x3 1
1+ + + +· · · (x > 0) . [JNTU 1996S, 1998S] Apply comparison test with vn =
2 5 10 np
un np 1 EXERCISE 9.5
= p =
vn (2n − 1) (2 − 1/n)p Test for convergence, the following series:
1
→ p , a finite non-zero number for all p. %
∞
2
$ $ 1. n2 /2n . [JNTU 2000S]
By comparison test, $ un and vn behave alike. n=1
Hence, when x = 1, un converges if p > 1 and Ans: convergent
diverges if p ≤ 1.
a + 1 (a + 1)(2a + 1)
2. 1 + +
Example 9.66 b + 1 (b + 1)(2b + 1)
% n!2n (a + 1)(2a + 1)(3a + 1)
Test the series for convergence. + + ···.
nn (b + 1)(2b + 1)(3b + 1)
[JNTU 1998S] Ans: b > a > 0, convergent a ≥ b > 0, divergent
%∞ *
Solution Here n+1 n
3. 3 +1
·x (x > 0).
n!2n n
un = > 0 for all n; n=0
/
nn convergent if x < 1 and
(n + 1)!2n+1 Ans:
un+1 = divergent if x ≥ 1
(n + 1)n+1
n %
∞
xn
un+1 (n + 1)! nn n 4. .
= ·2= ·2 1 + n2 x2n
un (n + 1)n+1 n! n+1 n=0
/
2 2 convergent if x ≤ 1 and
= → <1 Ans:
1 n divergent if x > 1
1+ n e
By ratio test, the series is convergent (∵ e = 2.72). %

∞
2n n!
5. . [JNTU 1998S].
n=1
nn
Example 9.67
Ans: convergent
Discuss the convergence of the series
2 2 · 5 2 · 5 · 8 2 · 5 · 8 · 11 % 1 · 3 · 5 · · · (2n − 1)
+ + + + ··· . 6. xn−1 (x > 0).
1 1 · 5 1 · 5 · 9 1 · 5 · 9 · 13 2 · 4 · 6 · · · 2n
Solution Here [JNTU2001S]
⎧
⎨ convergent if x < 1 and
2 2·5 2·5·8
u0 = , u1 = , u2 = ,··· Ans: divergent if x ≥ 1
1 1·5 1·5·9 ⎩
(zero test)
2 · 5 · 8 · · · (3n + 2)
un = , 1 x2 x4 x6
1 · 5 · 9 · · · (4n + 1) 7. √ + √ + √ + √ +· · · .[JNTU 1995]
2 · 5 · 8 · · · (3n + 2) (3n + 8) 2 ⎧ 1 3 2 4 3 5 4
un+1 = if x2 ≤ 1 and
1 · 5 · 9 · · · (4n + 1) (4n + 5) ⎨ convergent
Ans: divergent if x2 > 1
The factors in the numerator and denominator of ⎩
(comparison test)
un are in A.P. with common differences 3 and 4,
respectively. 2 3
1 2 3 4
8. + x+ x2 + x3 + · · · .
un+1 3n + 8 3 + 8 · 1n 3 2 3 4 5
= = → <1 [JNTU 2002]
un 4n + 5 4+5· n 1
4
% 1 n+1 n n
Hence by ratio test un converges. u0 = , un = x
2 n+2
⎧
⎨ convergent if x ≤ 1 and Putting r = 1, 2, 3 · · · (n − 1) and adding the
Ans: divergent if x > 1 corresponding terms of the inequalities
⎩
(zero test)
%
n %
n−1
%
∞
f (r) − f (1) ≤ In ≤ f (r) ∵ I1 = 0 (9.8)
9. xn /n. r=1 r=1
⎧
n=1
⎨ convergent if x < 1 and If we write sn = f (1) + f (2) + · · · + f (n) and

Ans: divergent if x ≥ 1 ur = f (r) then
⎩
(zero test)
sn − u1 ≤ In ≤ sn − un
%
∞
10. (n!) / (2n)!
2
[JNTU 1997] ⇒ un − sn ≤ −In ≤ u1 − sn
n=0 ⇒ un ≤ (sn − In ) ≤ u1 (9.9)
Ans: convergent
Let vn = sn − In for all n
9.22 Cauchy—Maclaurin’s Now vn+1 − vn = (sn+1 − In+1 ) − (sn − In )
Integral Test = (sn+1 − sn ) − (In+1 − In )
n+1
9.22.1 Improper Integral
= f (n + 1) − f (x)dx ≤ 0
∞
An integral of the form f (x)dx where the interval n
a [by Eq. (9.7)]
of integration or the integrand f (x) is not bounded
is called an improper integral. It is said to converge
X ∞ ∴ vn is monotonically decreasing.
if lim f (x)dx = l exists and we write f (x)dx;
X →∞ a a By Eq. (9.9), we have
otherwise it is said to diverge.
vn = (sn − In ) ≥ un = f (n) ≥ 0
Theorem 9.21 If f is a non-negative monotoni- i.e., vn ≥ f (n) ≥ 0 for all n
cally decreasing integrable function on [1, ∞), then
$
∞ ∞ Thus vn is a monotonically decreasing sequence
the series f (n) and the improper integral f (x)dx
n=1 1 which is bounded below, and hence it converges.
both converge or diverge together. Since sn = vn + In and vn is convergent it fol-
lows that the sequences sn and In both converge
Proof: Since f is non-negative on [1, ∞) we have
or diverge together.
f (x) ≥ 0 for all x ≥ 1. f is decreasing on [1, ∞) $ ∞
∴ The series f (n) and the integral f (x)dx
⇒ f (r + 1) ≤ f (x) ≤ f (r) for all x ∈ [r, r + 1] 1
converge or diverge together.
Since f is monotonic it is integrable on [r, r + 1] Note
$ ∞
r+1 r+1 r+1 The series f (n) and the integral f (x)dx converge
a
f (r + 1)dx ≤ f (x)dx ≤ f (r)dx or diverge together for a ≥ 1.
r r r
⇒ f (r + 1) ≤ (Ir+1 − Ir ) ≤ f (r) (9.7) Example 9.68
%∞
1
r Prove that the series converges if p > 1 and
where Ir = f (x)dx n=1
np
1
diverges if p ≤ 1.
1 $ ∞
Solution Let f be defined by f (x) = on By integral test, f (n) and 1 F(x)dx converge
xp or diverge together.
[1, ∞). If p > 0, f is integrable, non-negative and ∞ ∞
monotonically decreasing on [1, ∞). Since F(x)dx = log a ax F(ax )dx
1 1 1 a
Since x1 , x2 [1, ∞) and x1 > x2 ⇒ p > p for (a − 1) ∈ N
x1 x2 ∞
$ n
∞ ∴ F(x)dx and a F(an ) converge or
$ 1 1
By integral test, and dx converge or 1
np xp diverge together.
1 $ $
diverge together since Hence f (n) and an f (an ) converge or diverge
together for (a − 1) ∈ N .
∞ X
1 1
dx = lim dx Example 9.70
xp X →∞ xp %
∞
1 1 1
⎧ Show by integral test, that converges.
⎪
⎨ log X if p=1 n=1
n2 + 1
= lim X 1−p − 1
X →∞ ⎪ if p = 1
⎩ 1 1
1−p Solution Let f (n) = , then f (x) =
⎧ n2 +1 x2 +1
⎪
⎪ ∞, p=1 for all x ∈ [1, ∞)
⎪
⎨
∞, (0 0 and decreasing on [1, ∞).
=
⎪
⎪ 1
⎪
⎩ − (p > 1) X
1−p 1
F(X ) = dx = tan−1 X − tan−1 1
$ x2 +1
1
np
converges if p > 1 and diverges if 0 < p < 1. 1
$ 1 = tan−1 X − π/4
When p ≤ 0, lim n1p = 0 and is a series
np $ 1
of natural numbers raised to power p and so
np lim F(X ) = lim tan−1 X − π/4
diverges to +∞. X →∞ X →∞
$ 1 = π/2 − π/4 = π/4
Hence np
converges if p > 1 and diverges if
p ≤ 1.
∞
since f (x)dx converges to π/4.
9.23 Cauchy’s Condensation Test
1
Example 9.69 $ $ 1
If f is a non-negative, monotonically decreasing ∴ By integral test, f (n) = con-
$ $ n n2 + 1
function, then f (n) and a f (an ) converge or verges.
diverge together for (a − 1) ∈ N .
Example 9.71
Solution Let F be defined on [1, ∞) by %
∞
1
By integral test, show that the series p
F(x) = (x − n)f (n + 1) + (n + 1 − x)f (n) n=2
n log n
converges if p > 1 and diverges if 0 < p ≤ 1.
for n ≤ x < (n + 1), n ∈ N
1
F is a non-negative monotonically decreasing Solution Let f (n) = p , then f (x) =
function which is integrable on [1, ∞) and n log n
1
p for all x ∈ [2, ∞). Then f (x) > 0 and
F(n) = f (n) for all n ∈ N x log x
decreasing in [2, ∞). For x ∈ [1, ∞), there exists n ∈ N such that
X X n ≤ x ≤ n + 1 ⇒ f (n) ≥ f (x) > f (n + 1)
1
F(X ) = f (x)dx = p dx n+1 n+1
x log x
2 2 ⇒ f (n)dx ≥ f (x)dx
1−p X
log x n n
= (p = 1) > f (n + 1) (1)
1−p
2
⎧ 1−p 1−p Replacing n by 1, 2, · · · (n − 1) in Eq. (1) and

⎪
⎪ log X − log 2 adding the corresponding members of the inequality
⎪
⎪ (p = 1)
⎪
⎪ 1−p
⎪
⎪ ∞
⎪
⎪ log 2
1−p
⎨ sn − f (n) ≥ In = f (x)dx > sn − f (1)
→−
= 1−p
⎪
⎪
⎪
⎪
1
⎪
⎪ log log X −/log log 2 (p = 1) −f (n) ≥ −sn + In > −f (1)
⎪
⎪
⎪
⎪ ∞ p<1 f (n) ≤ (sn − In ) < f (1)
⎩ →
∞ p=1 1
1−p 1−p 0 < ≤ (sn − In ) < 1 for n∈N
n
log X − log 2
Now, lim =∞
X →∞ 1−p If we write vn = sn −In , then vn is bounded below.
(0 1) stant and is denoted by γ whose value is 0.577
1−p
approximately.
lim log log X − log log 2 = ∞ (p = 1)
X →∞
EXERCISE 9.6
%
∞
1 %
∞
By integral test, f (n) = p Using condensation test/integral test, test for conver-
n=2 n=2
n log n gence:
diverges if 0 1. % 1
1. √ .
Example 9.72 n
Ans: divergent
Show by the integral test that
% 1
1 1 1 2. .
lim 1 + + + · · · + − log n = γ. n2
n→∞ 2 3 n Ans: convergent
1 % 1
Solution Let f (x) = for all x ∈ [1, ∞). 3. .
x n log n
Then f (n) > 0 and decreasing on [1, ∞). Ans: divergent
% 1
Let sn = f (1) + f (2) + · · · + f (n) 4. .
1 1 1 n(n + 1)
= 1 + + + ··· + Ans: convergent
2 3 n
n n % 1
1 5. √ .
In = f (x) dx = dx = log n n n2 − 1
x Ans: convergent
1 1
9.24 Raabe’S Test (Higher Ratio Example 9.73

Test) Test the series for convergence if its nth term is
$
Theorem 9.22 If un is a series of positive
1 · 3 · 5 · · · (2n − 1) x2n
un $ · (x > 0).
terms such that lim n − 1 = l then un 2 · 4 · 6 · · · 2n 2n
un+1
Solution
(i) converges if l > 1
1 · 3 · 5 · · · (2n − 1) x2n
(ii) diverges to +∞ if l < 1 un = · ;
2 · 4 · 6 · · · 2n 2n
and (iii) the test fails if l = 1. 1 · 3 · 5 · · · (2n − 1) (2n + 1) x2n+2
un+1 = ·
$ 2 · 4 · 6 · · · 2n(2n + 2) 2n + 2
Proof: Let vn be an auxiliary series with
1 un+1 2n + 1 2n
vn = p .Then = · · x2
n un 2n + 2 2n + 2
2 + 1n 2
vn n+1 p 1 p = · · x2 → x2
= = 1+ 2 + 2 · 1n 2 + 2n
vn+1 n n
1 p (p − 1) 1 $
= 1+p· + + ··· By ratio test, un converges if x2 < 1, i.e., x < 1
n 2! n2
(by binomial theorem) diverges +∞ if x2 > 1, i.e., x > 1, and the test fails if
x = 1.
vn p (p − 1) 1
⇒n −1 =p+ + ··· When x = 1, we apply Raabe’s test
vn+1 2 n

$ un (2n + 2)2
By comparison
$ test, un converges or diverges n −1 = n −1
according as vn converges or diverges and un+1 2n (2n + 1)
2

4n+ 8n + 4 − 4n−2
2n
un vn =n
≷ for n ≥ m 2n (2n + 1)
un+1 vn+1
3n + 2 3 + 2n 3
i.e., according as = = → >1
2n + 1 2+ n 1
2
/ , / ,
un vn $
lim n −1 lim n −1 =p By Raabe’s test, un converges.
un+1 vn+1
Example 9.74
Discuss the nature of the series
Case (i) l > 1 Choose $ real $
number p such that
12 12 · 52 12 · 52 · 92 12 · 52 · 92 · 132
l ≥ p > 1. In this case,
$ vn = 1
np
converges and, + 2 2+ 2 2 + 2 2 +· · · .
by comparison test un converges. 4 4 · 8 4 · 8 · 12 4 · 8 · 122 · 162
2 2
Solution Here,
Case (ii) l < 1 Choose $ real number
$ 1 p such that
l ≤ p < 1. In this case,
$ vn = np
diverges and, 12 · 52 · 92 · · · (4n − 3)2
by comparison test, un diverges. un = ;
42 · 82 · 122 · · · (4n)2
Case (iii) l = 1 Clearly, the test is inconclusive; 12 · 52 · 92 · · · (4n − 3)2 (4n + 1)2
un+1 =
further investigation is required. 42 · 82 · 122 · · · (4n)2 (4n + 4)2
2
un+1 (4n + 1)2 4 + 1n
Note = = 2 → 1
Raabe’s test is to be applied when ratio test fails and un (4n + 4)2 4+4· 1 n
un /un+1 does not involve ‘e’. The ratio test fails.
We apply Raabe’s test, un+1 4 · 7 · 10 · · · (3n + 1) (3n + 4)

=
un 1 · 2 · 3 · · · n (n + 1)

un (4n + 4)2 1 · 2 · 3···n
n −1 =n − 1 × x
un+1 (4n + 1)2 4 · 7 · 10 · · · (3n + 1)
3 (8n + 5) 3n + 4
=n = x
(4n + 1)2 n+1

3 8 + 5 · 1n 3 3 + 4 · 1n
= → >1 = x → 3x
4+ n1 2 2 1 + 1n
$ $ 1
By Raabe’s test, un converges. By ratio test, un converges if x <and
3
1 1 1
diverges if x > . The test fails if x = . If x = ,
Example 9.75 3 3 3
Test for convergence the series un n+1
then = ·3
3 3·6 2 3·6·9 3 un+1 3n + 4
1+ x+ x + x + · · · (x > 0).
7 7 · 10 7 · 10 · 13 un n+1 −n
Solution Here, n −1 = n 3 −1 =
un+1 3n + 4 3n + 4
3 · 6 · 9 · · · (3n) −1 1
un = xn , = →− <1
7 · 10 · 13 · · · (3n + 4) 3 + 4 1n 3
un+1 3n + 3
= x→x By Raabe’s test, the series diverges to +∞ if
un 3n + 7
1
$ x= .
By ratio test, un converges if x < 1 diverges to 3
∞ if x > 1 and the test fails if x = 1. Example 9.77
We apply Raabe’s test Discuss the convergence of the series

un 3n + 7 4 (1!)2 (2!)2 2 (3!)2 3
n −1 =n −1 =n 1+ x+ x + x + ··· (x > 0).
un+1 3n + 3 3n + 3 2! 4! 6!
4 4 Solution Here,
= → >1
3 + 3 · 1n 3
$ (1!)2
u0 = 1, u1 = x, · · ·
By Raabe’s test, un converges. 2!
(n!)2 n [(n + 1)!]2 n+1
un = x , un+1 = x
Example 9.76 (2n)! (2n + 2)!
Test the series for convergence un+1 [(n + 1)!]2 (2n)!
= · x
% 4 · 7 · 10 · · · (3n + 1) un (2n + 2)! (n!)2
xn (x > 0) [JNTU 1996] (n + 1)2
1 · 2 · 3···n = x
(2n + 1) (2n + 2)
Solution Here, 2
1 + 1n 1
= x → x
4 · 7 · 10 · · · (3n + 1) n 2+ n 2+2· n
1 1
4
un = x,
1 · 2 · 3···n
4 · 7 · 10 · · · (3n + 1) (3n + 4) n+1 By ratio test, the series converges if x < 4 and
un+1 = x diverges to +∞ if x > 4. The test fails if x = 4.
1 · 2 · 3 · · · n (n + 1)
When x = 4, the terms of the series are Case (i) l > 1 Choose number p such that l ≥
$ real$
In this case, vn =
p > 1. $ 1
np
is convergent, and
(1!)2 hence un is convergent, by comparison test.
u0 = 1, u1 = · 4,
2!
(2!)2 2 (n!)2 n Case (ii) l < 1 Choose $real number
$ 1 p such that
u2 = · 4 , · · · , un = 4 l ≤ p < 1. $
In this case, vn = is divergent,
4! (2n!) np
and hence un is divergent to +∞, by comparison
un (2n + 1) (2n + 2) 1
n −1 = n · −1 test.
un+1 (n + 1)2 4

4n+ 6n + 2 − 4 n + 2n + 1
2 2
Case (iii) l = 1 Clearly, the test fails and a fur-
=n
4 (n + 1)2 ther investigation is required to settle the question of
1 1 1 convergency.
=− →− <1
2 1 + 1n 2 Note
un
Log test is to be applied when ratio test fails and
By Raabe’s test, the series diverges to +∞. un+1
involves e.
9.25 Logarithmic Test Example 9.78

Test the series for convergence
$
Theorem 9.23 If un is a series of positive
un $ x 2! 3! 4!
terms such that lim n log = l. Then un 1+ + 2 x2 + 3 x3 + 4 x4 + · · · (x > 0).
un+1 2 3 4 5
(i) converges if l > 1 Solution Here

(ii) diverges to +∞ if l < 1 1
uo = 1, u1 = x,
2
and (iii) the test fails if l = 1. 2! 2 n!
$ u2 = x , · · · , un = xn
Proof: Let vn be an auxiliary series with 3 2 (n + 1)n
1 un+1 (n + 1)! (n + 1)n
vn = p . Then = · x
n un (n + 2)n+1 n!
(n + 1)n+1
vn n+1 p 1 = x
n log = n log = np log 1 + (n + 2)n+1
vn+1

n

n n
1 1 1 1 + 1n 1 + 1n e 1 x
= np − + 3 − ··· =
2 n
→ 2· x=
2 n
n 2n2 3n 1+ n 1+ n e 1 e

1 1 $
= p 1− + − ··· By ratio test, un converges if x < e and diverges
2n 3n2
if x > e.
$
By comparison The test fails if x = e. In this case, the series is
$ test, un converges or diverges
according as vn converges or diverges and 1 2! 3! 4!
1 + e + 2 e2 + 3 e3 + 4 e4 + · · ·
un vn 2 3 4 5
≷ for n ≥ m
un+1 vn+1 un 2
log = (n + 1) log 1 +
un vn un+1 n
or lim n log lim n log =p
un+1 vn+1 1
or l p −(n + 1) log 1 + − log e
n
/ ,
2 1 4 1 8 1 16 1 1 1 1
= (n + 1) − · + · − · +· · · = n 1−n − + · −· · ·
n 2 n2 3 n3 4 n4 n 2n2 3 n3

1 1 1 1 1 11 1
− − · 2 + · 3 + ··· −1 =n − + ···
n 2 n 3 n 2 n 3n2

1 3 7 1 1 1 1
= (n + 1) − 2 + 3 − ··· − 1 = − · + ··· → < 1
n 2n 3n 2 3 n 2
3 1 7 1 1 1
= 1 − · + · 2 − ··· + By log test, the series diverges to +∞ if x = .
2 n 3 n n e
3 7 1 1
− 2 + · 2 − ··· − 1 Hence, the given series converges if x < and
2n 3 n e
1
1 1 5 1 diverges to +∞ if x ≥ .
= − · + · 2 − ··· e
2 n 6 n
un 1 51 1 Example 9.80
n log =− + − ··· = − < 1 Discuss the convergence of the series
un+1 2 6n 2
2x 32 x2 43 x3 54 x 4
By log test, the series diverges to +∞. 1+ + + + + ··· (x > 0).
2! 3! 4! 5!
Example 9.79 Solution Here,
Test the series for convergence
2x nn−1 n−1
22 x 2 33 x3 44 x 4 u1 = 1, u2 = , ··· , un = x
x+ + + + · · · (x > 0). 2! n!
2! 3! 4!
un+1 (n + 1)n−1 1 n 1
[JNTU 1998, 2002] = x = 1 + · ·x → ex
un nn−1 n 1+ 1n
Solution Here $ 1
By ratio test, un converges if x < and
22 x2 33 x3 e
u1 = x, u2 = , u3 = ,··· 1 1
2! 3! diverges to + ∞ if x > . The test fails if x = . In
nn n (n + 1)n+1 xn+1 e e
un = x , un+1 = this case, the series becomes
n! (n + 1)!
2 1 32 1 43 1
un+1 (n + 1)n+1 n! 1+ + + + ··· ;
= · ·x 2! e 3! e 2 4! e3
un (n + 1)! nn
n nn−1 1 (n + 1)n 1
n+1 1 n un = , un+1 =
= x = 1+ x → ex n! e n−1 (n + 1)! en
.
n n
Applying log test,
$ 1
By ratio test, un converges if x < and un 1
e log = log n−1 e
1 1 un+1
diverges to +∞ if x > . The test fails if x = . In 1 + 1n
e e
this case, the series becomes 1
= log e − (n − 1) log 1 +
n
1 1 2 2 1 3 3 nn 1 n
+ + + · · · ; un = 1 1 1
e 2! e 3! e n! e = 1 − (n − 1) − 2 + 3 − ···
n 2n 3n
un nn (n + 1)! e
= · ·e = n 1 1
n! (n + 1)n+1 = 1− 1− + 2 − ···
un+1 1 + 1n 2n 3n

un 1 1 1
⇒ n log = n log e − n log 1 + + − 2 + ···
un+1 n n 2n
3 5 1 $ $
= − + ··· Proof: Let vn be an auxiliary series with vn
2n 6 n2 1
un 3 5 1 3 = . Then,
n log = − · 2 + ··· → > 1 n(log n)p
un+1 2 6 n 2
vn (n + 1) [log (n + 1)]p
1 = p
By log test, the series converges if x = . Hence, vn+1 n log n
e p
1 1 log n 1 + 1n
the series converges if x ≤ and diverges to +∞ if = 1+
e n log n
1 p
x> .
e 1 log n + log 1 + 1n
= 1+
Example 9.81
n log n
p
Examine for convergence the series whose nth term 1 log n + 1n − 2n12 + · · ·
2n n! = 1+
is un = n xn n log n
n p
1 1 1
Solution = 1+ 1+ − + ···
un+1 2x 2 $ n n log n 2n2 log n
= n → x. By ratio test, un con-
un 1+ n 1 e 1 p
e e = 1+ 1+ − ···
verges if x < and diverges if x > . The test fails n n log n
2 2 n by multinomial theorem
e un 1 + 1n
if x = . In this case, we have = . = 1+ +
1 p
+higher powers of or
1 1
2 un+1 e n n log n n log n
We apply log test
$ $
By comparison test, un and vn converge or
un 1
n log = n n log 1 + − log e diverge together according as
un+1 n
un vn
1 1 1 ≷ for n≥m
=n n − 2 + 3 − ··· − 1 un+1 vn+1
n 2n 3n
i.e., according as
1 1
= n 1− + 2 − ···− 1
2n 3n un
lim n − 1 − 1 log n
un+1

1 1 un
= − + − ··· lim n − 1 − 1 log n
2 3n un+1
1 i.e., lp
→− <1
2
Case (i) l > 1 Choose $ real$
number p such that
e
By log test, the series diverges to +∞ if x = . l ≥ p > 1. In this case, vn = 1
converges,
2 $ n(log n)p
and by comparison test, un converges.
9.26 De Morgan–Bertrand Test Case (ii) l < 1 Choose $ real $
number p such that
$ l ≤ p < 1. In this case, vn = 1
diverges,
$ n(log n)p
If un is a series
Theorem 9.24 / of positive
, and by comparison test, un diverges.
un
terms such that lim n − 1 − 1 log n =
$ un+1 Note
l, then un (i) converges if l > 1 and (ii) diverges This test is to be applied when Raabe’s and log tests
if l < 1. fail.
Example 9.82 By Bertrand’s test, we see that the series diverges

Test the following series for convergence: if x = 1.
22 2 2 · 42 22 · 42 · 62 EXERCISE 9.7
(i) 1 + + + + ···
32 32 · 5 2 32 · 5 2 · 7 2 Test for convergence the following series:
22 2 2 · 42 2 2 · 42 · 6 2
(ii) 1 + 2 x + 2 2 x2 + 2 2 2 x3 + · · · 1 x3 1 · 3 x5 1 · 3 · 5 x7
3 3 ·5 3 ·5 ·7 1. x+ · + + +· · · (x > 0).
(x > 0) [JNTU 1995S] 2 3 2·4 5 2·4·6 7
(Ratio/Raabe’s).
Solution Ans: converges if x2 ≤ 1 and diverges to +∞ if
22 22 · 42 x2 > 1
(i) uo = 1, u1 = , u2 = 2 2 , · · · , x 2! 3! 4!
3 2 3 ·5 2. 1 + + 2 x2 + 2 x3 + 4 x4 + · · · (x > 0).
22 · 42 · 62 · · · (2n)2 2 3 4 5
un = 2 2 2
3 · 5 · 7 · · · (2n + 1)2 (Ratio/log ) [JNTU 1993]
2 Ans: converges if x < e and diverges to +∞ if
un 2n + 3 2 2 + 3n x≥e
= =
un+1 2n + 2 2 + 2n
22 22 · 42 22 · 4 2 · 62
→ 1; So, the ratio test fails. 3. 1 + + 2 2 + 2 2 2 + ···.
3 2 3 ·5 3 ·5 ·7
un 4n + 5 4 + 5 n12 Ans: divergent
n −1 = n =
un+1 (2n + 2)2 (2 + 2 1n )2 3 3·6 2 3·6·9 3
4. 1 + x + x + x · · · (x > 0).
→ 1; So, Raabe’s test also fails. 7 7 · 10 7 · 10 · 13
Ans: convergent if x ≤ 1 and divergent +∞, if
Now, to apply Bertrand’s Test, we find
x > 1.

un x 22 x 2 33 · x3 44 x 4
n − 1 − 1 log n 5. + + + + · · · (x > 0)
un+1 1! 2! 3! 4!

4n2 + 5n (Ratio/log). [JNTU 1998]
= − 1 log n 1 1
4n2 + 8n + 4 Ans: convergent if x < and divergent if x ≥
−3n − 4 e e
= 2
4n + 8n + 4 x x2 x3 x4
6. + + + + · · · (x > 0).
−1 3 + 4 · 1n 1·2 3·4 5·6 7·8
= →0<1 Ans: convergent if x ≤ 1 and divergent if x > 1
n 4 + 8 · 1n + 4 · n12
(1!)2 2 (2!)2 4 (3!)2 6
By Bertrand’s test, the series diverges. 7. 1 + x + x + x + · · · (x > 0).
2! 4! 6!
Ans: convergent if x < 4 and divergent if x2 ≥ 4
2
22 22 · 42
(ii) u0 = 1, u1 = x, u2 = 2 2 x2 , · · · % 4 · 7 · · · (3n + 1)
3 2 3 ·5
8. xn (x > 0)
22 · 42 · 62 · · · (2n)2 1 · 2···n
un = 2 2 2 xn (Ratio/Raabe’s). [JNTU 1996]
3 · 5 · 7 · · · (2n + 1)2
1 1
un+1 (2n + 2)2 (2 + 2 · 1n )2 Ans: convergent if x < and divergent if x ≥
= x = x→x 3 3
un (2n + 3)2 (2 + 3 · 1n )2
1 1·3 2 1·3·5 3
$ 9. 1 + x + x + x + · · · (x > 0)
By ratio test, un converges if x < 1 and diverges 2 2·4 2·4·6
if x > 1. The Ratio test fails if x = 1. Similarly, (Ratio/Raabe’s).
Raabe’s test also fails as shown in Solution (i). Ans: convergent if x < 1 and divergent if x ≥ 1
a + x (a + 2x)2 (a + 3x)3 = u1 − k where k > 0

10. + + +· · · (x > 0).
1! 2! 3! ⇒ s2n ≤ u1 for all n.
(Ratio/log).
1 1 Thus, s2n is monotonically increasing sequence
Ans: convergent if x < and divergent if x ≥
e e which is bounded above.
2x 32 x2 43 x3 ∴ lim s2n = l exists.
11. 1 + + + + · · · (Ratio/log).
2! 3! 4!
1 1 Now s2n+1 = s2n + u2n+1 → l
Ans: convergent if x ≤ and divergent if x >
e e ∵ u2n+1 → 0 (by hypothesis)
Thus, sn tends to the same limit whether n is even
12 12 · 32 1 2 · 32 · 52
12. + 2 2 x + 2 2 2 x2 + · · · (x > 0) or odd. Hence, the series is convergent.
22 2 ·4 2 ·4 ·6
(Bertrand’s). Note
Ans: convergent if x < 1 and divergent if x ≥ 1 If lim un = 0, then s2n = s2n+1 and the alternating
series will be oscillatory.
1 1+2 1+2+3
13. + 2 + 2 + ···.
12 1 + 22 1 + 22 + 32
Ans: divergent 9.27.2 Absolute Convergence
$
Definition $A series un is said to be absolutely
9.27 Alternating Series convergent if |un | is convergent.
Definition A series whose terms are alternatively
positive and negative is called an alternating series. 9.27.3 Conditional Convergence
An alternating series is of the form $
%∞
Definition A convergent series un is said to
n−1 n−1 $
(−1) un = u1 −u2 +u3 −u4 +· · ·+(−1) un +· · · be conditionally convergent if |un | is divergent.
n=1
Remark
$ $
where un > 0 for each n.
(1)
$ If un is a series of positive
$ terms, then un and
Example n | are same so that
|u$ un is convergent
$ implies
% 1 1 1 1 that |un | is convergent. Hence, un is absolutely
(1) (−1)n−1
= 1 − + − + ··· convergent.
n 2 3 4 $
1 1 1 (2) If un is a series comprising an infinite number
(2) 1 − + − + · · · of positive terms$and an infinite number of nega-
2! 3! 4!
tive
$terms, then un is absolutely convergent only
if |un | is convergent.
9.27.1 Leibnitz’s Test
Theorem 9.26 Every absolutely convergent
Theorem 9.25 If un is a monotonically decreas-
series is convergent.
ing sequence of positive terms that lim un = 0,
$ suchn−1
then the alternating series (−1) un is conver- $
Proof: Let$ un be absolutely convergent. This
gent. implies that |un | is convergent.
Proof: Let sn = u1 −u2 +u3 −u4 +· · ·+(−1)n−1 un . Let > 0. Then there exists an m ∈ N such that
Then s2n+2 − s2n = u2n+1 − u2n+2 ≥ 0
⇒ s2n+2 ≥ s2n |un+1 | + |un+2 | + · · · + un+p <
Therefore, the sequence s2n is an increasing for all n ≥ m, p ≥ 0
sequence. ⇒ un+1 + un+2 + · · · + un+p
Also, s2n = u1 − [(u2 − u3 ) + (u4 − u5 ) + · · · ≤ |un+1 | + |un+2 | + · · · + un+p <
+ (u2n−2 − u2n+1 ) + u2n ] for all n ≥ m, p ≥ 0
> ?
By Cauchy’s general principle of convergence, the 1
$ So, p is a monotonically decreasing sequence.
series un is convergent. n
1
Note Also, lim p = 0 for p > 0
n
The converse of the theorem is not true as seen from
$By Leibnitz’s test, the alternating series
the following counter-example. (−1)n−1 n1p (0 < p ≤ 1) is convergent, but
$ 1
Example 9.83 np
is divergent. Hence, the series is conditionally
%
∞
1 1 1 1 convergent if 0 < p ≤ 1.
Show that (−1)n−1
= 1 − + − + · · · is
n=1
n 2 3 4 Particular cases
convergent but not absolutely. 1 1 1
(1) Prove that the series 1− √ + √ − √ +· · · is
Solution Here, 2 3 4
conditionally convergent. [JNTU 1996S]
1 1 1
un = , un+1 = and un > un+1 for all n. (Take p = in Example 9.87)
n n+1 2
1 1 1 1 1 1
Also, lim un = lim = 0 (2) Test whether 1 + 2 − 2 − 2 + 2 + 2 −
n 2 3 4 5 6
1 1
− + · · · is convergent. [JNTU 1998]
By Leibnitz’s theorem, the alternating series 72 82
% 1
(−1)n−1 is convergent.
n
% % %1 Solution This is not an alternating series. It is a
1
But |un | = (−1)n−1 = is mixed series.
n n 1 1
p-series with p = 1 and is divergent. Writing u1 = 1, u2 = 2 , u3 = − 2 , · · ·
2 3
% % 1
% 1 we see that the absolute series |un | =
∴ The series (−1)n−1 is conditionally con- n2
vergent.
n is a p-series with p = 2, and it is convergent. The
given series is absolutely convergent. Since every
Example 9.84 absolutely convergent series is convergent, the given
%
∞
1 series is convergent.
Show that the series (−1)n−1
(p > 0) is abso-
n=1
np Example 9.85
lutely convergent if p > 1 and conditionally conver- Show that the exponential series
gent if p ≤ 1.
x x2 x3
1+ + + + ··· (1)
Solution The given series is an alternating series 1! 2! 3!
$ 1 converges absolutely for all real x.
(−1) un where un = p .
n−1
$ $ n $ 1
The series |un | = (−1)n−1 un = np
is Solution Let
p-series which is convergent if p > 1 and divergent if x x2 xn
p ≤ 1. u0 = 1, u1 = , u2 = ,··· , un =
1! 2! n!
Case (i) p > 1 The given series is absolutely un+1 1
convergent ∴ = |x| → 0 as n → ∞
un n+1
1 $
Case (ii) 0 By ratio test, |un | is convergent for all real x.
np
1 Hence, the given series (1) converges absolutely for
for all n. all x.
(n + 1)p
%
Example 9.86 3. (−1)n−1 (n + 1) / (2n).
2 3 4
x x x Ans: oscillatory
Show that the series x − √ + √ − √ + · · ·
2 3 4
converges if −1 < x ≤ 1. 1 1 1 1
4. − + − + ···.
1·2 3·4 5·6 7·8
Solution Let Ans: absolutely convergent
√ %
∞
xn un+1 n
un = √ ⇒ = |x| 5. (−1)n−1 (n + 2) / (2n + 5).
n un (n + 1)1/2 n=1
1 Ans: absolutely convergent
= |x| → |x|
1 1/2
1+ 2 3 4 5
n
6. − 2 + 2 − 2 + ···.
$ 12 2 3 4
By ratio test, un converges absolutely if |x| < 1. Ans: conditionally convergent
If |x| = 1 the test fails.
1 1 1 1 1 1
In this case, the series is 1− √ + √ − √ +· · · , 7. p − p + p − · · ·
2 3 4 2 log 2 3 log 3 4 log 4
which (p > 0).
$ is conditionally convergent as shown above. /
∴ un is convergent for x such that −1 < x ≤ 1. absolutely convergent p > 1
Ans:
conditionally convergent 0 < p ≤ 1
EXERCISE 9.8 % √ √
8. (−1)n−1 n+1− n .
Test the following series for convergence: Ans: conditionally convergent
1 1 1 %
∞
1. 1 − + − + ···. 1
1! 2! 3! 9. (−1)n−1 sin .
n=1
n
Ans: absolutely convergent
Ans: conditionally convergent
1 1 1 1 % cos nα
2. − + − + ···. 10. (−1)n−1 √ (α ∈ R).
log 2 log 3 log 4 log 5 n n
Ans: conditionally convergent Ans: absolutely convergent
Vector Differential Calculus 10
10.1 Vector Differential Calculus The following results hold:
We consider continuity and derivability of vector If ū(t) → l̄, v̄(t) → m̄ and λ is a scalar, then
functions of one or more variables and the new
concepts of gradient, divergence and curl, whose (1) lim [ū(t) + v̄(t)] = l̄ + m̄
applications lie in topic such as potential theory. (2) lim[λū(t)] = λl̄
Vector function of a scalar variable (3) lim[ū(t) · v̄(t)], = l̄ · m̄ = lm
Let S ⊂ R. If to each scalar t in S there corresponds (4) lim[ū(t) × v̄(t)] = l̄ × m̄.

a unique vector ū(t) then we say that a vector-valued
function or a vector function ū : S → V is defined 10.1.1 Continuity of a Vector
on domain S. Function
If ī, j̄, k̄ be three mutually perpendicular unit vec- Let ū be a vector function defined for each t in domain
tors along the x-, y- and z-axes of a cartesian S and c ∈ S. Then it is said to be continuous at ‘c’ if
coordinate system then
lim ū(t) = ū(c)
t→c
u(t) = u1 (t)i + u2 (t)j + u3 (t)k
ū is continuous on S if it is continuous at every point
where ui (t) (i = 1, 2, 3) are real-valued functions of S. This implies that if ū = u1 ī + u2 j̄ + u3 k̄ is
and are called the components of ū(t). continuous then u1 , u2 , u3 are continuous on S and
conversely, if ū and v̄ are continuous on s then
Limit of a vector function ū ± v̄, ū · v̄, λū and ū × v̄
are also continuous on S.
Let u(t) be a vector function defined in a deleted
neighbourhood of c of S and l be a given vector. If
to each > 0, there corresponds a δ > 0 such that 10.1.2 Derivability
Let ū be a vector function on S and c ∈ S. Then if
0 < |t − c| < δ ⇒ |u(t) − l| < the limit
ū(t) − ū(c)
then u(t) is said to tend to the vector l as t → c. lim
t→c t−c
Symbolically, we write
exists then we say that ū is derivable (or differen-
lim ū(t) = l̄ tiable) at ‘c’ and the limit value is called the derivative
t→c (differential coefficient) of ū with respect to t at ‘c’.
Symbolically, we write
In practice, if u(t) = u1 (t)l + u2 (t)j + u3 (t)k and
l = l1 i + l2 j + l3 k, then lim u(t) = l ⇔ lim ui = li d ū(t)
(i = 1, 2, 3). or ū (c)
dt t=c
If ū is derivable at each point of S, then we say that (7) If ū and φ are differentiable vector and scalar
d ū(t) functions respectively on a domain S, then
ū is derivable on S and is called the derivative
dt (ūφ) is differentiable on S and
or derived function of ū(t) on S. Usually an interval
I = [a, b] is taken for S.
Derivative of ū at ‘t’ can also be written as (ūφ) = ū φ + ūφ
d ū ū(t + t) − ū(t) Note that in (4) and (6) the order of the functions
(t) = lim , must be preserved.
dt t→0 t
Let P(x, y, z) be a point whose position in space is
if the limit exists. given by the vector
ū(t) = īu1 (t) + j̄u2 (t) + k̄u3 (t) is derivable at ‘t’
⇔ each ui (t) is derivable and r̄ (t) = x (t) ī + y (t) j̄ + z (t) k̄
d ū where x, y, z are differentiable functions of t. As t

= īu1 (t) + j̄u2 (t) + k̄u3 (t) varies continuously, P traces a curve in space.
dt
If ū(t) is derivable at t = c, then ū(t) is continuous Example 10.1
at t = c. The converse of this is not true. The equation
r̄ = a cos ωt ī + a sin ωt j + bt k̄
10.1.3 Higher Order Derivatives
where a, b and ω are positive constants, represents a
d ū
Let ū be differentiable on S and ū = be the first
dt z
order derivative of ū on S. If
ū (t + t) − ū (t)
lim
t→0 t
exists, then we say that ū(t) is twice differentiable on
d 2 ū
S and write the limit as ū (t) or 2 (t) and call it
dt
the second order derivative of ū(t) and so on.
10.1.4 Algebra of Derivatives

P(t)
Let ū and v̄ be two differentiable vector functions of O b y
ω a
t on S. Then
x
(1) (ū ± v̄) (t) = ū (t) ± v̄ (t)
Figure 10.1
(2) (λ u) (t) = λ ū (t)
(3) (ū · v̄) (t) = ū (t) · v̄ (t) + ū(t) · v̄ (t) circular helix (Fig. 10.1). The helix traced by
r̄ = a cos ωt ī + a sin ωt j̄ + bt k̄ (b > o)
(4) (ū × v̄) (t) = ū (t) × v̄ (t) + ū(t) × v̄ (t)
% spirals up from the xy-plane as t increases from zero.
(5) [ūv̄w̄] = ū v̄w̄ The projection of the point P(x, y, z) on to the xy-
plane moves around the circle x2 + y2 = a2 , z = 0

(6) [ū ×(v̄ × w̄)]
= ū × (v̄ × w̄) + ū × v̄ × w̄ + as t varies while the distance |bt| between P and the
ū × v̄ × w̄ xy-plane, changes steadily with t.
Vector Differential Calculus 10-3
Arc length Example 10.2

Find t̄ for the helix r̄ = a cos ωt ī + a sin ωt j̄ + bt k̄.
The formula for the length of the arc from t = a to
t = b of space curve Solution
r̄ (t) = x (t) ī + y (t) j̄ + z (t) k̄ Velocity vector
b + v̄ = ī(−aω sin ωt) + j̄(aω cos ωt) + k̄b
is L = [x (t)]2 + [y (t)]2 + [z (t)]2 dt
whose length is
a
b b |v̄| = a2 ω2 sin2 ωt + a2 ω2 cos2 ωt + b2
d r̄ ¯
= dt = |v|dt. = a2 ω 2 + b 2
dt
a a v̄ (aω(−ī sin ωt + j̄ cos ωt) + bk̄)
t̄ = = √ .
|v̄| a2 ω 2 + b 2
Unit tangent vector t
Theorem 10.1 The necessary and sufficient con-
Suppose r̄(t) is a curve in space whose component dition for ū(t) to be a constant function on a domain
functions x(t), y(t), z(t) are differentiable. Let s(t) be S is that ū (t) = 0̄.
the arc length along the curve measured from some
Proof:
point P0 (t). Then
(1) The condition is necessary
t
Let ū(t) be constant on S. Then
s(t) = (x (t))2 + (y (t))2 + (z (t))2 dt
ū(t) − ū(c)
t0 lim = ō
t→c t−c
ds ∵ ū(t) = ū(c) for all t ∈ S
s(t) is a differentiable function of t. If = |v̄| is
dt
never zero, then s(t) is one-to-one and has an inverse ⇒ ū (t) = ō
that gives t as a differentiable function of s whose
dt 1 (2) The condition is sufficient
derivative is = Let ū (t) = 0̄ on S and let ū(t) = īu1 (t) + j̄u2 (t) +
ds |v|
k̄u3 (t)
d r̄ d r̄ dt 1 d r̄
Now = · = ⇒ īu1 (t) + j̄u2 (t) + k̄u3 (t) = 0̄ for all t ∈ S
ds dt ds |v̄| dt
d r̄ 1 d r̄ 1 ⇒ ui (t) = 0 (i = 1, 2, 3) for all t ∈ S
⇒ = = |v̄| = 1
ds |v̄| dt |v̄| ⇒ ui (t) are constant scalar functions on S
dr ⇒ ū(t) = īu1 (t) + j̄u2 (t) + k̄u3 (t)
This implies that is a unit vector in the direction
ds is a constant vector function on S.
of v̄.
d r̄
We call the unit tangent vector of the curve and Theorem 10.2 If ū is a differentiable vector
ds
denote it by t̄. function of a scalar t on a domain S, then
If r̄(t) = xī + yj̄ + z k̄ where x, y, z are functions of d 2 d ū

(ū ) = 2ū · .
t that are twice differentiable, then the velocity and dt dt
acceleration of P(x, y, z) are defined by Proof:
d r̄ dx dy dz d 2 d d ū d ū
v̄ = = ī + j̄ + k̄ (ū ) = (ū · ū) = · ū + ū ·
dt dt dt dt dt dt dt dt
2
d x d 2y d 2z du du du
a = ī 2 + j̄ 2 + k̄ 2 , respectively. = 2u · ∵ ·u=u·
dt dt dt dt dt dt
Corollary Taking cross-product of members of Eqs. (10.1)

d 2 du du and (10.2).
(u ) = 0 ⇒ u · = 0 ⇒ ū and
dt dt dt d ū d v̄ du
are mutually perpendicular vectors. ū × = uv̄ × u + uv̄ × v̄
dt dt dt
(If ū is a non-zero, non-constant vector). d v̄
= u v̄ ×
2
+ 0̄ (∵ v̄ × v̄ = 0̄) (10.4)
dt
Example 10.3
(1) The condition is necessary
d v̄
ū = ī cos t + j̄ sin t Let the direction of ū be constant ⇒ = 0̄
dt
d ū
⇒ = −ī sin t + j̄ cos t d ū
dt ⇒ ū × = 0̄ by (10.4) (10.5)
d ū dt
⇒ ū · = 0. (2) The condition is sufficient
dt
d ū d v̄
10.1.5 Vector of Constant Let ū × = 0̄ ⇒ u2 v̄ × = 0̄
dt dt
Magnitude d v̄
Theorem 10.3 The necessary and sufficient con- ⇒ v̄ × = 0̄ (∵ u = 0) (10.6)
dt
dition for ū(t) to be of constant magnitude is that
d ū Since v̄ is of constant (unit) magnitude (∵ u = 0)
ū · = 0.
dt d v̄
∴ v̄ · =0 (10.7)
Proof: dt
(1) The condition is necessary d v̄
Let ū(t) be a vector of constant magnitude From Eq. (10.6), it follows that v̄ and must be
dt
⇒ ū2 (t) = ū(t) · ū(t) = |ū(t)|2 = constant collinear. From Eq. (10.7), it follows that v̄ must be
zero or constant vector. ∵ They are collinear and
d ū d ū they are non-perpendicular. v̄ being unit vector, its
⇒ 2ū · = 0 ⇒ ū · =0
dt dt direction must be constant.
(2) The condition is sufficient
d ū 1d 10.1.7 Derivative of composite
Let ū · =0⇒ (u · u) = 0
dt 2 dt vector function (chain rule)
⇒ ū · ū = |ū|2 = constant Theorem 10.5 Let s be a real-valued function on
⇒ ū(t) is a vector function of constant magnitude. S differentiable at t ∈ S and suppose ū is a vector-
valued function on S differentiable at s(t). Then the
10.1.6 Vector of Constant Direction composite vector function ū is differentiable at t and
[ū(s)] (t) = ū [s(t)] · s (t)
Theorem 10.4 The necessary and sufficient con-
dition for ū(t) to be of constant direction is that where denotes differentiation with respect to the
d ū argument (proof omitted).
ū × = 0̄.
dt
Proof: Let Partial derivatives
ū(t) = u(t) · v̄(t) (10.1) Let ū be a function of more than one variable
where |ū(t)| = u(t) and |v̄(t)| = 1 (10.2) p, q, t, · · · say. If we consider ‘t’ as a variable and
ū = uv̄ all others constant and if
d d v̄ du ū(p, q, t + t) − ū(p, q, t)
⇒ (ū) = u + v̄ (10.3) lim
dt dt dt t→0 t
exists then ū is said to have the partial derivative with (3) ū + v̄ = (2t, t, 3t) + (3t, −t, −2t)
∂ū
respect to the variable t and it is denoted by . = (5t, 0, t)
∂t + √
∂u |ū + v̄| = (5t)2 + 0 + t 2 = 26t
Also, as t varies in domain S, defines a vector
∂t d d√ √
function whose derivative exists if |ū + v̄| = 26t = 26.
∂u ∂u dt dt
(p, q, t + t) − (p, q, t)
lim ∂t ∂t Example 10.5
t→0 t
f¯ (t) = t 3 ī − t j̄ + (2t 2 + 1)k̄
If
exists. It is called the second order partial deriva-
∂2 u ḡ (t) = (2t + 3) ī + j̄ − t 2 k̄,
tive of ū with respect to ‘t’ and is denoted by 2
∂t find (1) f¯ × ḡ (t) (2) f¯ + ḡ (t) at t = 1.
and so on.
Solution
Example 10.4
If ū = 2t ī + t j̄ + 3t k̄ and v̄ = 3t ī − t j̄ − 2t k̄, find (1) (f¯ × ḡ) (t) t=1
= f¯ (t) × ḡ(t)+ f¯ (t) × g (t) t=1
(1)
d
(ū · v̄) , (2)
d
(ū × s̄v) and (3)
d
|ū + v̄| = f¯ (1) × ḡ(1) + f¯ (1) × ḡ (1)
dt dt dt
f¯ (t) = t 3 ī − t j̄ + (2t 2 + 1)k̄,
at t = −1.
ḡ (t) = (2t + 3) ī + j̄ − t 2 k̄
Solution f¯ (t) = 3t 2 ī − 1 · j̄ + 4t k̄,
(1) ū · v̄ = 6t 2 − t 2 − 6t 2 = −t 2 ḡ (t) = 2ī − 0j − 2t k̄
f¯ (1) = ī − j̄ + 3k̄,
d
(u · v) = −2t|t=−1 = 2. ḡ(1) = 5ī + j̄ − k̄
dt t=−1
¯f (1) = 3ī − j̄ + 4k̄,
ī j̄
k̄ ḡ (1) = 2ī + 0.j̄ − 2k̄
(2) ū × v̄ = 2t t
3t
3t −t
−2t f¯ (1) × ḡ(1)+ f¯ (1) × ḡ (1)
2
= t , 13t , −5t 2
2
ī j̄ k̄ ī j̄ k̄
d = 3 −1 4 + 1 −1 3
(ū × v̄) = (−2t, 26t, −10t)
dt t=−1 t=−1
5 1 −1 2 0 −2
= (−2, −26, 10).
= (−3, 23, 8) + (2, 8, 2) = (−1, 31, 10).
d ī j̄ k̄
Otherwise (ū × v̄) = 2 1 3 (2) f + g (t) = (t 3 + 2t + 3, −t + 1, t 2 + 1)
dt
3t −t −2t
f¯ + ḡ (t)
+
2
ī j̄ k̄ = t + 2t + 3 + (1 − t) + (t + 1)
2 2 2 2
+ 2t t 3t

3 −1 −2 = 2t 4 + 4t 3 + 13t 2 + 10t + 11
(differential of the determinant ū × v̄ is the sum
1 4 1
of the two determinants where R2 and R3 are = (2t + 4t 3 + 13t 2 + 10t + 11)− 2
2
respectively replaced by their derivatives)
(8t 3 + 12t 2 + 26t + 10)
= (2t, 26t, −10t)t=−1 = (−2, −26, 10) .

f¯ + ḡ (t) t=1 5. If r̄ = et (c̄1 cos 3t + c̄2 sin 3t) where c̄1 , c̄2 are
1 1 constant vectors then r̄ − 2r̄ + 10r̄ = 0̄.
= (2 + 4 + 13 + 10 + 11)− 2
2 6. If Ā = x2 y − x3 ī + exy + y cos x j̄ +
(8 + 12 + 26 + 10) 2
x cos y k̄ then find
28 14
= √ =√ . ∂2 Ā ∂2 Ā ∂2 A ∂2 Ā
2 10 10 (a) (b) (c) (d) and
∂x2 ∂y 2 ∂x∂y ∂y∂x
Example 10.6
∂2 Ā ∂2 Ā
If r̄(t) = a cos t ī + a sin t j̄ + ta sin t k̄, find show that = .
(1) |r̄ × r̄ |, (2) [r̄ r̄ r̄ ] . ∂x∂y ∂y∂x
Ans: (a) (2y − 6x)ī + (y2 exy − y cos x)j̄ + 2 cos yk̄
Solution (b) 0 · ī + x2 exy j̄ − x2 cos yk̄
(c) = (d) = 2xī + (xyexy − sin x)j̄ − 2x sin yk̄
(1) r̄ = a cos t ī + a sin t j̄ + at k̄
r̄ = −a sin t ī + a cos t j̄ + a k̄ 7. If φ x, y, z = x2 y2 z 2 and Ā = zx2 ī − xy2 j̄ +
∂3
r̄ = −a cos t ī − sin t j̄ + 0̄ yz 2 k̄ find 2 φĀ at (1, −1, 1) .
∂x ∂z
r̄ = a sin t ī − a cos t j̄ Ans: 36ī − 12j̄ − 8k̄
ī j̄ k̄ 8. Prove that if r̄ = xī + yj̄ + z k̄ and ā is constant

function then
r̄ × r̄ = −a sin t a cos t a
% ∂
−a cos t −a sin t 0 (a) (ā.r̄) ī = ā
= (a2 sin t, −a2 cos t, a2 ) ∂x
% ∂

√ 2 (b) (ā × r̄) .ī = 0
r̄ × r̄ = 2a . ∂x
% ∂
(2) r̄ r̄ r̄ = r̄ × r̄ · r̄ (c) (ā × r̄) × ī = 2ā.
∂x
= a3 sin2 t + a3 cos2 t + a2 · 0
= a3 . 9. Let C̄1 , C̄2 be constant vector functions, m be a
real constant and
EXERCISE 10.1 ū = emx (C̄1 sin my+ C̄2 cos my). Then ∇ 2 ū = 0̄.
1. If r = (t 2 + 1)i + (4t − 1)j + (2t 2 − t)k show 10. ā, w, c are constant functions
that the unit vector in the direction of r at t = 1 a ∂2 F̄ 2∂F̄ 1 ∂2 F̄
1 2 2 F̄ (r̄, t) = eiw(t−r/c) then 2 + = .
is i + j + k. r ∂r r∂r c2 ∂t 2
3 3 3
2. If ū = 3t 2 ī − t 2 j̄ − t 3 k̄, v̄ = sin t ī + cos t j̄ 10.2 Differential Operators
then, 10.2.1 Sets of Points
2
π π
(a) (ū · v̄) at t = is 3πī + j̄ A collection of points is called a space. Let S be a
2 4 subset of a space. A point ‘a’ is an interior point of S if
there is a neighbourhood of ‘a’ such that every point
π3 −3π2 π2
(b) (ū × v̄) (t) t=π/2 = − , , . of it belongs to S. A point ‘b’ is called a boundary
8 4 4 point of S if every neighbourhood of b contains points
3. Show that [v̄v̄ v̄ ] = [v̄v̄ v̄ ] . in S as well as points not in S.
A set S is called open if all its points are inte-
4. If r̄ = cos ωt ī + sin ωt j̄ then show that rior points of S. A set is called connected if any two
(a) r̄ = −ω2 r̄ (b) r̄ × r̄ = ωk̄ points of S can be joined by means of a broken line
(c) r̄ · r̄ = 0. entirely lying in S. An open connected set is called
a domain. A domain together with none, some or all defining the derivative we need to associate it with a
its boundary points is called a region. specific direction.
Directional derivative at a point

10.2.2 Scalar and Vector Point
Function Let P be a point in space and L be a ray through P
Let S be a domain in space. A rule φ which assigns in the direction of a given unit vector ē. Let a scalar
to each point P in S a unique scalar φ(P) is called a point function φ be defined in a neighbourhood D of
scalar point function φ(SPF φ) over S and a rule f¯ P. Let Q = P and Q ∈ L ∩ D.
which assigns to each point P in S a unique vector
f¯ (P) is called a vector point function f¯ (VPF f¯ ) L
over S.
Q
A common example for a scalar point function is
the function r(P) the distance of point P(x, y, z) from e
the origin O defined by r (p) = OP = x2 + y2 + z 2
and that for a vector point function is the position P
vector r̄ (P) = OP = xī + yj̄ + z k̄.
In representing a point function, we can use P or
(x, y, z) or r̄ as its arguments. Thus, φ(P), φ(x, y, z) or
φ(r̄) may be used to denote a scalar point function and o
similarly f¯ (P), f¯ (x, y, z) or f¯ (r̄) to denote a vector
point function. Figure 10.3
The density ρ(P) at any point P of a material body
occupying a region is an example of a scalar point φ Q − φ (P)
function and the velocity v̄(P) of a particle at any If lim exists, then it is called
Q→P QP
point P of a fluid flowing in a channel occupying a the directional derivative of scalar point function φ at
region is an example of a vector point function. ∂φ ∂φ
P in the direction of ē. It is denoted by or where
The union of the domain S and the range R of a ∂ē ∂s
scalar point function φ is said to form a scalar field s = PQ. Similarly, for a vector point function f¯ (P)
of φ and similarly the domain S and the range R of a ∂f¯
the directional derivative is defined by
vector point function f¯ constitute a vector field of f¯ . ∂ē

∂f¯ ∂f¯ f¯ Q − f¯ (P)
φ=c+2h
or = lim
∂ē ∂s Q→P QP
φ=c+h
φ=c
provided the limit exists.
φ=c–h
Replacing ē by ī, j̄ or k̄ (or by x, y, z respectively)
we get the corresponding directional derivatives of
Scalar field Vector field scalar point function φ or vector point function f¯ .
Figure 10.2 Theorem 10.6 Let P be a point in the domain of
r̄ and L be a ray through P in the direction of ē. If
Temperature or density in a body occupying a Q ∈ L and Q = P then

region are scalar fields and force field F̄(P) or electric ∂r̄ r̄ Q − r̄ (P) PQ
field Ē(P) are vector fields. = lim = lim = ē
∂ē Q→P QP Q→P QP
The concepts of limit and continuity can be defined
for a scalar point function and a vector point function The directional derivatives of r̄ along x, y, z direc-
as for real-valued functions in real calculus. But for tions are ī, j̄, k̄, respectively.
∂r̄ ∂r̄
i.e., or = ī etc.
∂x ∂ī
If r = xi + yj + zk then
∂r̄ ∂r̄ ∂x ∂y ∂z
= = ī + j̄ + k = ē.
∂ē ∂s ∂s ∂s ∂s
Example 10.7
∂r (r̄ · ē)
If r = |r̄| , then = .
∂ē r
Solution
r̄ 2 = r̄ · r̄ = r 2 Figure 10.4
∂ ∂r̄ ∂r
(r̄ · r̄) = 2r̄ · = 2r̄ · ē = 2r
∂ē ∂ē ∂ē ∂φ ∂φ ∂φ ∂x ∂y ∂z
∂r (r̄ · ē) = ī + j̄ + k̄ · ī + j̄ + k̄
⇒ = . ∂x ∂y ∂z ∂s ∂s ∂s
∂ē r
∂r̄ ∂r̄
= ∇φ · = ∇φ · = ∇φ · ē
Gradient of a scalar point function φ ∂s ∂ē
∂r ∂x ∂y ∂z̄
Let φ(r̄) be a scalar point function having directional ∵ = ī + j̄ + k̄ .
∂s ∂s ∂s ∂s
∂φ ∂φ ∂φ
derivatives , , in the directions of ī, j̄, k̄, In terms of differentials, we have
∂x ∂y ∂z
respectively. Then the vector point function called ∂φ ∂φ ∂φ
the gradient of φ is defined by dφ = dx + dy + dz
∂x ∂y ∂z

∂φ ∂φ ∂φ
∂φ ∂φ ∂φ = ī + j̄ + k̄ · īdx + j̄dy + k̄dz
gradφ = ∇φ = ī + j̄ + k̄ ∂x ∂y ∂z
∂x ∂y ∂z = ∇φ · dr.
∂ ∂ ∂
The symbol ∇ ≡ ī + j̄ + k̄ called the
∂x ∂y ∂z
‘del’ or ‘nabla’ is a vector differential operator which
when applied on a scalar point function φ having first
∂φ ∂φ ∂φ
order partial derivations , , (not all of them
∂x ∂y ∂z
vanish simultaneously) produces a vector point func-
∂φ ∂φ ∂φ
tion grad φ having , , as its components. In
∂x ∂y ∂z
fact

∂ ∂ ∂ Figure 10.5
gradφ = ∇φ = ī + j̄ + k̄ φ
∂x ∂y ∂z
∂φ ∂φ ∂φ
= ī + j̄ + k̄ Level surface
∂x ∂y ∂z
Let φ be a non-constant scalar point function and
By chain rule, if the partial derivatives exist, we c be a real constant. The set of points Q such that
have φ(Q) = c is called a level surface at level c.
∂φ ∂φ ∂φ ∂x ∂φ ∂y ∂φ ∂z Equipotential function and isothermal function are
= + + + examples of level surfaces.
∂ē ∂s ∂x ∂s ∂y ∂s ∂z ∂s
% ∂
Potential When we apply ∇ ≡ ī on a scalar point
∂x
The operation of forming the gradient leads from function φ, we obtain the vector point function
a scalar field φ = φ(r̄) to vector field grad φ. In
general, we cannot go in the opposite direction. That % ∂φ
∇φ = ī
is, not every vector field is the gradient of a scalar ∂x
∂φ ∂φ ∂φ
field. Vector fields F̄ = ī + j̄ + k̄ which are
∂x ∂y ∂z Thus, ∇ is a vector differential operator expressed
gradients of some scalar point function φ are called as a linear combination of vectors ī, j̄, k̄ with the par-
conservative and φ is called the potential of F̄. ∂ ∂ ∂
tial differential operators , , as coefficients
the scalar point function φ = φ (r̄) =
Consider ∂x ∂y ∂z
φ x, y, z . If r̄ is increased by d r̄ = dxī + dyj̄ + dz k̄, or components.
then the change in φ is Note that
∂φ ∂φ ∂φ ∇x = ī, ∇y = j̄, ∇z = k̄.

dφ = dx + dy + dz
∂x ∂y ∂z
This can be regarded on the scalar product of the

10.3.2 Scalar Differential Operator
vector d r̄ describing the small movement in space
∂φ ∂φ ∂φ ē.∇ (S.D.O. ē · ∇)
and a vector grad φ = ∇φ = ī + j̄ + k̄, the
∂x ∂y ∂z Consider the identities
gradient of φ. The change in the direction d r̄ is then
% ∂φ
(1) (ē · ∇) φ ≡ ē · ī
dφ = grad φ · d r̄ = ∇φ · d r̄ ∂x
∂φ ∂φ ∂φ
In particular, if we choose d r̄ so that it lies in ≡ ē · ī + ē · j̄ + ē · k̄
∂x ∂y ∂z
the level surface φ = constant, then φ does not % ∂f¯
change so that dφ = 0. Hence, the grad φ = ∇φ (2) (ē · ∇) f¯ ≡ ē · ī
is perpendicular to the level surface φ = constant. ∂x
Now suppose that d r̄ is at an angle θ to grad φ so ∂f¯ ∂f¯ ∂f¯
≡ ē · ī + ē · j̄ + ē · k̄ .
that dφ = grad φ · d r̄ = |∇φ| |d r̄| cos θ and hence ∂x ∂y ∂z
the function increases most rapidly when θ = 0, i.e.,
in the direction of the gradient φ and the maximum % ∂
When we apply (ē · ∇) ≡ ē · ī
rate of growth of φ is |∇φ|. ∂x
∂φ
Since n̄ = ∇φ we see that ∇φ represents both (i) on an scalar point function φ we obtain the scalar
∂n̄ % ∂φ
in magnitude and in direction the maximum rate of point function ē · ī ,
increase of φ. ∂x
(ii) on an a vector point function f¯ we obtain the

10.3 Scalar and Vector % ∂f¯
Differential Operators vector point function ē · ī .
∂x
10.3.1 Vector Differential Operator
∇ (V.D.O. ∇) Thus, (ē · ∇) is a scalar differential operator
Consider the identity expressed as a linear combination of partial
∂ ∂ ∂
differential operators , , with the scalars
% ∂φ ∂φ ∂φ ∂φ ∂x ∂y ∂z
∇φ = ī ≡ ī + j̄ + k̄
∂x ∂x ∂y ∂z e · i , ē · j̄ , ē · k̄ as coefficients.
% ∂
10.3.3 Vector Differential Operator Thus, Identity (2) ē × ī · is a scalar dif-
ē × ∇ (V.D.O. ē × ∇) ∂x
% ∂
ferential operator and Identity (3) ē × ī ×
Consider the identities ∂x
is a vector differential operator.
% ∂φ
(1) (ē × ∇) φ ≡ ē × ī
∂x 10.3.4 Scalar Differential Operator
∂φ ∂φ ∇. (S.D.O. ∇)
≡ ē × ī + ē × j̄
∂x ∂y
∂φ Consider the identity
+ ē × k̄
∂z % ∂φ ∂f¯ ∂f¯ ∂f¯
% ∂f¯ ∇ · f¯ ≡ ī · f¯ ≡ ī · + j·¯ + k̄ ·
(2) (ē × ∇) · f¯ ≡ ē × ī · ∂x ∂x ∂y ∂z
∂x
∂f¯ ∂f¯ % ∂
≡ ē × ī · + ē × j̄ · When we apply ∇· ≡ ī · on a vector point
∂x ∂y ∂x
of function f¯ , we obtain the scalar point function
∂f¯ % ∂ % ∂
+ ē × k̄ · ī. f¯ . Thus, ∇· ≡ ī · is a scalar dif-
∂z ∂x ∂x
% ∂f¯ ferential operator with partial differential operators
(3) (ē × ∇) × f¯ ≡ ē × ī × ∂ ∂ ∂
∂x , , applied on the respective components
∂f¯ ∂f¯ ∂x ∂y ∂z
≡ ē × ī × + ē × j̄ × ∂f1 ∂f2 ∂f3
∂x ∂y f1 , f2 , f3 of f¯ and yielding a scalar + + .
∂x ∂y ∂z
∂f¯
+ ē × k̄ × .
∂z Note
f¯ = f1 ī + f2 j̄ + f3 k̄ then
When we apply Identity (1) ē × ∇ ≡
% ∂ ∂ ∂ ∂
ē × ī on a scalar point function φ, we ¯
∇ · f = ī · + j̄ · + k̄ ·
∂x ∂x ∂y ∂z
% ∂φ
obtain the vector point function (ē × ī) . Thus, · f1 ī + f2 j̄ + f3 k̄
∂x
% ∂ ∂f1 ∂f2 ∂f3
ē × ī is the vector differential operator = + + ,
∂x ∂x ∂y ∂z
expressed
as a linear combination of vectors
ē × ī , ē × j̄ , ē × k̄ with the partial differential is called the divergence of f¯ and denoted by div f¯ .
∂ ∂ ∂
operators , , as coefficients.
∂x ∂y ∂z Physical interpretation of divergence of a
When we apply Identity (2) (ē × ∇) · ≡ vector field f¯
% ∂
ē × ī · on a vector point function f¯ , we obtain Let f¯ = f1 ī + f2 j̄ + f3 k̄ be a vector field e.g. velocity
∂x
% ∂f¯ field at a point P(x, y, z) in a region occupied by a
the scalar point function (ē × ī) ·
∂x
and when we fluid, then ∇ · f¯ ≡ div f¯ is the rate of change of
% ∂ volume per unit of an infinitesimal portion of the
apply (3) (ē × ∇) × ≡ ē × ī × on a vec- fluid containing P.
∂x
tor point function, we obtain vector point function
% ∂f¯ Solenoidal A vector point function f¯ is said to be
ē × ī × .
∂x solenoidal if ∇ · f¯ = 0.
10.3.5 Vector Differential Operator ∂r x ∂r y ∂r z

⇒ = , = , =
∇× (V.D.O.∇×) ∂x r ∂y r ∂z r
% ∂r % x
Consider the identity ∇r = ī = ī
∂x r
% ∂ ¯

∂ ∂ ∂ xī + yj̄ + z k̄
∇×f = ī × f ≡ ī× f¯ +j̄× f¯ +k̄× f¯ = = r̄/r.
∂x ∂x ∂y ∂z r
% ∂
% (ii) ∇φ (r) = ī φ (r)
∂ ∂x
When we apply ∇× ≡ ī × on a vector % ∂
∂x ∂r
¯ = ī φ (r)
point function f , we obtain a vector point function ∂r ∂x
% ∂ ¯ % x r̄
ī × f. = φ (r) ī = φ (r) .
∂x r r
% ∂ %
Thus, ∇× ≡ (ī × ) is a vector differential 1 ∂ 1
∂x
(iii) ∇ = ī
r ∂x r
operator as it is expressed as a linear combination of
∂ ∂ ∂ 1 % ∂r
vectors ī, j̄, k̄, in which partial derivatives , , =− 2 ī
∂x ∂y ∂z r ∂x
are present as coefficients. 1 1 r̄ r̄
= − 2 ∇r = − 2 = − 3 .
∇ × f¯ is also called the curl of f¯ . In fact, if f¯ = r r r r
f1 ī + f2 j̄ + f3 k̄, then we have
Example 10.9
% ∂f3 ∂f2

If r 2 = x2 + y2 + z 2 prove that
∂r
=
(r̄ · ē)
where ē
¯
∇ × f ≡ curlf ≡ ¯ ī − ∂ē r
∂y ∂z is a given unit vector.
ī j̄ k̄
∂ ∂ ∂ ∂φ
= Solution We know that = ∇φ · ē, where φ is
∂x ∂y ∂z ∂ē
f1 f2 f3 a scalar point function

In expanding the determinant, the functions ∂r (r̄ · ē) r
⇒ = ∇r · ē = ∵ ∇r = .
f1 , f2 , f3 are always written to the right of the operators ∂ē r r
∂ ∂ ∂
, , . Example 10.10
∂x ∂y ∂z If ā is a constant vector point function then
Irrotational A vector field F̄ is called irrotational ∇(ā · r̄) = ā.
if curl F̄ ≡ ∇ × F̄ = 0̄.
Solution
Example 10.8 % ∂
r̄ ∇ (ā · r̄) = ī (ā · r̄)
Prove that (i) ∇r = r̄/r, (ii) ∇φ(r) = φ (r) , ∂x
r % ∂r̄
1 r̄ = ī ā ·
(iii) ∇ =− 3 ∂x
r r
= ī ā · ī + j̄ ā · j̄ + k̄ ā · k̄ = ā
Solution ∂r̄ ∂r̄ ∂r̄
(i) We have r 2 = x2 + y2 + z 2 and r̄ = xī + yj̄ + z k̄. r̄ = xī + yj̄ + z k̄, = ī, = j̄, = k̄,
∂x ∂y ∂z
Differentiating r 2 partially with respect to x, y, z ā = a1 ī + a2 j̄ + a3 k̄
∂r ∂r ∂r ⇒ ā · ī = a1 , ā · j̄ = a2 , ā · k̄ = a3
2r = 2x, 2r = 2y, 2r = 2z
∂x ∂y ∂z so that ā = ī ā · ī + j̄ ā · j̄ + k̄ ā · k̄ .
Example 10.11 (∵ Scalar triple product with two equal vectors=0).

Prove that % ∂r̄
6. F̄ × ∇ × r̄ = F̄ × ī ×
∂x
1. ∇ · r̄ = 3 %
= F̄ × ī × ī
2. ∇ × r̄ = 0̄ %
= F̄ · ī ī − F̄
∂φ ∂φ ∂φ = F̄ − 3F̄ = −2F̄.
3. ∇ · (φ(r)r̄) = x +y +z + 3φ
∂x ∂y ∂z
7. ∇ · (r n r̄) = (∇r n ) · r̄ + r 3 (∇ · r̄)
4. ∇ × (φ(r) × r̄) = 0̄
r̄
= nr n−1
· r̄ + r n ∇ · r̄
r
5. (F̄ × ∇) · r̄ = 0̄
= nr n−2 r 2 + r n · 3 = (n+3) r n
6. (F̄ × ∇) × r̄ = −2F̄ ∴ ∇ · (r n r̄) = 0 ⇒ n = −3.

1 1
7. ∇ · (r n r̄) = 0 ⇔ n = −3 8. ∇ 2
= ∇ ·∇
r r

1 1 r̄
8. ∇ 2
= 0. =∇· − 2
r r r

r̄
Solution = −∇ · 3 = 0 [by (7)].
% r
∂
1. ∇ · r̄ = ī · xī + yj̄ + z k̄ Example 10.12
∂x
If φ(u, v, w) is a differentiable scalar point function
= ī · ī + j̄ · j̄ + k̄ · k̄ of u, v, w which are differentiable functions of x, y, z,
= 1 + 1 + 1 = 3. ∂φ ∂φ ∂φ
then prove that ∇φ = ∇u + ∇v + ∇w.
% ∂ ∂u ∂v ∂w
2. ∇ × r̄ = ī × xī + yj̄ + z k̄ Solution
∂x
= ī × ī + j̄ × j̄ + k̄ × k̄ = 0̄. ∂φ ∂φ ∂φ
∇φ = ī + j̄ + k̄
∂x ∂y ∂z
3. ∇ · (φ (r) r̄) = (∇φ) · r̄ + φ (∇ · r̄)
∂φ ∂u ∂φ ∂v ∂φ ∂w
∂φ ∂φ ∂φ = ī + +
= ī + j̄ + k̄ ∂u ∂x ∂v ∂x ∂w ∂x
∂x ∂y ∂z
∂φ ∂u ∂φ ∂v ∂φ ∂w
· xī + yj̄ + z k̄ + 3φ +j̄ + +
∂u ∂y ∂v ∂y ∂w ∂y

[∵ ∇ · r̄ = 3, by (i)] ∂φ ∂u ∂φ ∂v ∂φ ∂w
∂φ ∂φ ∂φ +k̄ + + ,
=x +y +z + 3φ. ∂u ∂z ∂v ∂z ∂w ∂z
∂x ∂y ∂z (by chain rule of differentiation)

4. ∇ × (φ (r) × r̄) = (∇φ) × r̄ + φ (∇ × r̄) ∂φ ∂u ∂u ∂u
= ī + j + k̄
r̄ ∂u ∂x ∂y ∂z
=φ
× r̄ + φ(0̄) = 0̄
r ∂φ ∂v ∂v ∂v
+ ī + j̄ + k̄
∵ r̄ × r̄ = 0̄ ∇ × r̄ = 0̄, by (2) . ∂v ∂x ∂y ∂z

% ∂φ ∂w ∂w ∂w
∂r̄ + ī + j̄ + k̄
5. F̄ × ∇ · r̄ = F̄ × ī · ∂w ∂x ∂y ∂z
∂x
% ∂φ ∂φ ∂φ
= F̄ × ī · ī = 0 = ∇u + ∇v + ∇w.
∂u ∂v ∂w

Find the directional derivative of φ(x, y, z) = xy + Find the greatest value of the directional derivative
yz + zx in the direction of −2ī + j̄ + 2k̄ at the point of φ(x, y, z) = 2x2 − y − z 4 at (2, −1, 1).
(1, 2, 0).
Solution The greatest value of a scalar point
Solution
function φ is |∇φ| in the direction of the unit vector
φ(x, y, z) = xy + yz + zx ∇φ
ē = at P.
∂φ ∂φ ∂φ |∇φ|
∇φ = ī + j̄ + k̄
∂x ∂y ∂z
∂φ ∂φ ∂φ
= ī(y + z) + j̄(x + z) + k̄(x + y) ∇φ = ī + j̄ + k̄
∂x ∂y ∂z
Let ē be unit vector along the given direction, then = ī4x + j̄(−1) + k̄(−4z 3 )

−2ī + j̄ + 2k̄ 2 1 2 ∇φ|(2,−1,1) = 4x, −1, −4z 3 (2,−1,1)
ē = = − ī + j̄ + k̄
(−2)2 + +
12 22 3 3 3 = (8, −1, −4)
+
∂φ 2 1 2
= ē · ∇φ = − , , · y + z, z + x, x + y |∇φ|P = 82 + (−1)2 + (−4)2
∂ē 3 3 3 √
= 64 + 1 + 16 = 9
1
= [−2(y + z) + (z + x) + 2(x + y)] 1 8 1 4
3 Direction = ē = (8, −1, −4) = ,− ,− .
= (3x − z)/3 9 9 9 9
∂φ 3x − z EXERCISE 10.2
= = 1.
∂ē P 3 (1,2,0)
1. Prove that ∇r n = nr n−2 r̄.

Example 10.14
Find the directional derivative of φ(x, y, z) = x2 yz + r̄
2. Prove that ∇ log r = 2 .
4xz 2 at (1, −2, −1) in the direction of 2ī − j̄ − 2k̄. r
r̄ 2
Solution 3. If F̄ = , then show that div F̄ = and F̄ = 0̄.
∂φ ∂φ ∂φ r r
∇φ = ī + j̄ + k̄
∂x ∂y ∂z 4. If φ and ψ are differentiable scalar point func-

= 2xyz + 4z , x z, x2 y + 8xz
2 2
tions, then show that div (∇φ × ∇ψ) = 0̄.

2 1 2 2 1 2
ē = ī − j̄ − k̄ = ,− ,− 5. If ē is a unit vector point function, then prove that
3 3 3 3 3 3
(a) ∇ · [(ē · r̄) ē] = 1,
2 1 2
ē · ∇φ = 2xyz + 4z 2 − xz (b) ∇ · [(e × r) × e] = 2,
3 3 (c) ∇ × [(ē · r̄) ē] = 0̄,
2 2
− x y + 8x (d) ∇ × [(ē × r̄) × ē] = 0̄.
3
2 1 2 6. If F̄ is any vector point function, prove that
ē · ∇φ|P = 2xyz + 4z 2 − xz 1
3 3 ∇ F̄ · F̄ = (F · ∇) F̄ + F̄ × ∇ × F̄ .
2
2 2
− x y + 8x
3 (1,−2−1)
7. If ā is a constant vector point function, prove that
2 1 2 ∇ · (ā × r̄) = 0 and ∇ × (ā × r̄) = 2ā.
= (4 + 4) − (−1) − (−2 − 8)
3 3 3 8. If φ(r) is a scalar point function in scalar ‘r’, then
16 + 1 + 20 37 2
= = . prove that ∇ 2 φ (r) = φ (r) + φ (r )
3 3 r
A 19. Find the unit normal to the surface x2 +y2 +2z 2 =

Also, ∇ 2 φ (r) = 0 ⇔ φ (r) = +B
r 26 at the point (2,
√1, 3). [JNTU 2001]
(A and B are scalar constants).
Ans: ī + j̄ + 3k̄ / 11
9. If f , g, h are any differentiable scalar point func- 20. Find the angle between the surfaces x2 +y2 +z 2 =
tions, then prove that x2 + y − z = 12 at (2, 2, 2).
2
12 and
√
(a) ∇ 2 (fg) = g∇ 2 f + 2∇f · ∇g + f ∇ 2 g, Ans: 7/ 3 π
$ $
(b) ∇ 2 (fgh) = gh∇ 2 f + 2 f ∇g · ∇h .
10.4 Identities
10. Let F̄ = r n r̄ find scalarpoint function φ such that 1. Prove that for any scalar point functions φ and ψ
∇φ = F̄. Prove that ∇ ∇ · F̄ = n (n + 3) r n−2 r̄.
(a) ∇ (φ + ψ) = ∇φ + ∇ψ
11. Prove that curl F̄ = 0̄ if
(b) ∇ (φψ) = φ (∇ψ) + (∇φ) ψ.
(a) F̄ = yz ī + zxj̄ + xyz̄

(b) F̄ = 2xyī + x2 + log z j̄ + (y/z) k̄. Proof:
% ∂
12. Find a scalar point function φ such that F̄ = ∇φ (a) ∇ (φ + ψ) = ī (φ + ψ)
∂x
in each of the cases (a) and (b) in Problem 11 % ∂φ ∂ψ
above. = ī +
∂x ∂x
13. If ā is a constant vector point function and F̄ = % ∂φ % ∂ψ
r n (ā × r̄), then = ī + ī
∂x ∂x
(a) divF̄ = 0 = ∇φ + ∇ψ.
(b) curl F̄ = (n + 2) r n ā − nr n−2 (ā·r)
¯ r̄.
% ∂
14. If F̄ = (r̄ − r̄1 ) × (r̄ − r¯2 ) then prove that (b) ∇ (φψ) = ī (φψ)
∂x
(a) div F̄ = 0̄ % ∂ψ ∂φ
(b) curl F̄ = 2 (r̄2 − r¯1 ) = ī φ + ψ
dx ∂x
(c) grad [(r̄ − r¯1 ) · (r̄ − r¯2 )] = 2r − r¯1 − r¯2 . % ∂ψ % ∂φ
=φ ī + ī ψ
15. If ē is a unit vector point function in a fixed direc- ∂x ∂x
tion and F̄ is any vector = φ (∇ψ) + (∇φ) ψ.
point function,
thenprove
that ∇ · F̄ = ē · ∇ F̄ · ē − ∇ × F̄ × ē
2. Prove that for any vector point functions f¯ and ḡ
16. If F̄ = F1 ī + F2 j̄ + F3 k̄ is any vector point
(a) ∇ · f¯ + ḡ = ∇ · f¯ + ∇ · ḡ
function, then prove that
(b) ∇ × f¯ + ḡ = ∇ × f¯ + ∇ × ḡ.
(a) ∇ · F̄ = ī · (∇F1 ) + j̄ · (∇F2 ) + k̄ · (∇F3 )
(b) ∇ × F̄ = (∇F1 ) × ī + (∇F2 ) × j̄ + (∇F3 ) × k̄. Proof:
% ∂
17. Find the unit normal vector to the surface x + 3 (a) ∇ · f¯ + ḡ) = i · (f¯ + ḡ)
∂x

3xyz at (1, −1, 2) .
y3 + z 3 = 14+√
% ∂f¯ ∂ḡ
Ans: 3ī − j̄ + k̄ / 35 = ī · +
∂x ∂x
18. Find the directional derivative of xyz 2 + xz at
% ∂f¯ % ∂ḡ
(1,1,1) in the direction of the normal to the sur- = ī · + ī ·
face√3xy2 + y + z at (0,1,1). [JNTU 2000] ∂x ∂x
Ans: 4/ 11 = ∇ · f¯ + ∇ · ḡ.
%
∂ % ∂f¯ ∂
(b) ∇ × (f¯ + ḡ) = ī × (f¯ + ḡ) = ī · × ḡ + f¯ ×
ḡ
∂x
∂x ∂x
% ∂f¯
= ī × +
∂ḡ % ∂ % ∂ḡ

∂x ∂x = ī. f¯ × ḡ + ī. f¯ ×
∂x ∂x
% ∂f % ∂ḡ
= ī × + ī × % ∂f¯ % ∂ḡ
∂x ∂x = ī × · ḡ + f¯ · × ī ,
¯ ∂x ∂x
= ∇ × f + ∇ × ḡ.
3. Prove that for any scalar point function φ and any [Interchange dot and cross in box product]
vector point function f¯
% ∂f¯ % ∂ḡ

= ī × · ḡ − ī × · f¯ ,
(a) ∇ · φf¯ = (∇φ) · f¯ + φ ∇ · f¯ ∂x ∂x

(b) ∇ × φf¯ = (∇φ) × f¯ + φ ∇ × f¯ .
[Interchange members in cross product]
Proof:
% ∂ ¯ = ∇ × f¯ · ḡ − ∇ × ḡ · f¯ .
(a) ∇ · φf¯ = ī · φf
∂x

%
∂φ ¯ ∂f¯ (b) ∇ × f¯ × ḡ
= ī · f +φ
∂x ∂x % ∂ ¯
% = ī × f × ḡ
∂φ % ∂ ∂x
= ī · f¯ + φ ī. f¯ %
∂x ∂x ∂f¯ ∂ ḡ
= ī× × ḡ + f¯ ×
= ∇φ · f¯ + φ ∇ · f¯ . ∂x ∂x

% ∂ ¯ % ∂f¯ % ∂ḡ
(b) ∇ × φf¯ = ī × φf = ī × × ḡ + ī × f × ¯
∂x ∂x ∂x
% ∂φ ¯ ∂f¯
= ī × f +φ % ∂f¯ % ∂f¯
∂x ∂x = ḡ · ī − ī · ḡ
% ∂x ∂x
% ∂f¯
= ī
∂φ
× f¯ + φ ī × % ∂ḡ % ∂ḡ
∂x ∂x + ī. f¯ − f¯ · ī
∂x ∂x

= ∇φ × f¯ + φ ∇ × f¯ . = ḡ · ∇ f − ∇ · f ḡ + ∇ · ḡ f¯ − f¯ · ∇ ḡ.
¯ ¯
4. Prove that for any vector point functions f¯ and ḡ

% ∂
(a) ∇ · f¯ × ḡ = ḡ · ∇ × f¯ (c) ∇ f¯ · ḡ = ī f¯ · ḡ
∂x

−f¯ · (∇ × ḡ) [JNTU 1995] % ∂f¯ ∂ ḡ
= ī · ḡ + f¯ ·
(b) ∇ × f¯ × ḡ = ḡ · ∇ f¯ − f¯ · ∇ ḡ ∂x ∂x

+ ∇ · ḡ f¯ − ḡ(∇ · f¯ ) % ∂f¯ % ∂ḡ
= ī · ḡ + ī f ·
(c) ∇ f¯ · ḡ = f¯ · ∇ ḡ + ḡ · ∇ f¯ ∂x ∂x

+f¯ × ∇ × ḡ + ḡ × ∇ × f¯ .
∂f¯ ¯ ∂ḡ
= ḡ × ī× + f × ī×
Proof: ∂x ∂x
% ∂ ¯ ∂f¯ ∂ḡ
(a) ∇ · f¯ × ḡ = ī · f × ḡ + ḡ · ī + f¯ · ī
∂x ∂x ∂x

∂f¯ ∂f¯ ∂f¯ Irrotational vector point function
∵ ḡ × ī × = ī ḡ · − ḡ · ī ;
∂x ∂x ∂x A vector point function f¯ whose curl vanishes is
called an irrotational vector point function.
∂ḡ ∂ḡ ∂ḡ
f¯ × ī × = ī f¯ . − f¯ · ī If φ is any scalar point function having continu-
∂x ∂x ∂x
ous second order partial derivatives, then vector point
¯ ¯
= ḡ × ∇ × f + f × ∇ × ḡ function ∇φ is an irrotational vector point function.

+ ḡ.∇ f¯ + f¯ .∇ ḡ. 7. If ρ and p are non-zero scalar point functions such
5. If φ is a scalar point function that has second par- that ρĒ = ∇p where Ē is a vector point function
tial derivatives, then prove that then prove that Ē · curl Ē = 0.
∂2 φ ∂2 φ ∂2 φ Taking ∇× on both sides of
∇. (∇φ) = 2 + 2 + 2 · Proof:
∂x ∂y ∂z
∇ρ = ρĒ
Proof: we have ∇ × ∇ρ = 0̄ = ρ∇ × Ē + (∇ρ) × Ē
%
∂ ∂φ ∂φ ∂φ 1
∇ · (∇φ) = i· ī + j̄ + k̄ ⇒ ∇ × Ē = − [(∇ρ) × Ē]
∂x ∂x ∂y ∂z ρ
∂φ ∂φ ∂φ
2 2 2
= 2 + 2 + 2. Now taking dot product with Ē
∂x ∂y ∂z
Ē · ∇ × Ē = 0 ⇒ Ē · curl Ē = 0
From the above identity, we see that
∂2 ∂2 ∂2 ∵ E · [ ∇ρ × Ē ] = [Ē, ∇ρ, Ē] = 0.
∇2 = ∇ · ∇ ≡ 2 + 2 + 2
∂x ∂y ∂z 8. If F̄ is non-zero, non-constant vector point func-
is a second order differential operator. It is called the
tion in a constant direction, then F̄ and curl F̄ are
Laplacian. When applied on a scalar point function
perpendicular.
having second partial derivatives it yields a scalar
point function. Proof: Let F̄ = F ē, where ē is a unit vector in
the constant direction, and |F̄| = F is a scalar point
6. If φ is a scalar point function which has continuous
function.
second order partial derivatives, then prove that
Now, ∇ × F̄ = ∇ × (F ē)
∇ × (∇φ) ≡ curl grad φ = 0̄
= (∇F) × ē + F (∇ × ē)
Proof: = (∇F) × ē

∂ ∂φ ∂φ ∂φ ∵ ∇ × ē = 0̄ (∵ ē is constant)
∇ × ∇φ = ī × ī + j̄ + k̄
∂x ∂x ∂y ∂z
Taking dot product on both sides with ē
∂ ∂φ ∂φ ∂φ
+ j̄ × ī + j̄ + k̄ ē. ∇ × F̄ = ē · (∇F × ē) = [ē, ∇F, ē] = 0
∂y ∂x ∂y ∂z
⇒ F ē · ∇ × F̄ = 0 ⇒ F̄ · ∇ × F̄ = 0
∂ ∂φ ∂φ ∂φ
+ k̄ × ī + j̄ + k̄
∂z ∂x ∂y ∂z ⇒ F̄ and curl F̄ are perpendicular.
2
∂φ ∂2 φ ∂2 φ ∂2 φ
= ī − + j̄ − Solenoidal A vector point function having contin-
∂y∂z ∂z∂y ∂z∂x ∂x∂z uous second partial derivatives is called a solenoidal
2
∂φ ∂2 φ if its divergence is zero.
+ k̄ − =0
∂x∂y ∂y∂x Let f¯ be a vector point function. Then the function
F̄ = ∇ × f¯ isalso a vector point function such that
∂2 φ ∂2 φ ∇ · F̄ = ∇. ∇ × f¯ = 0. So, F̄ = ∇ × f¯ is a
∵ = .
∂y∂z ∂z∂y solenoidal vector point function.
9. If F̄ is a vector point function which has continu- By Identity 3(b), we have

ous second partial derivatives, then
∇ × φf¯ = (∇φ) × f¯ + φ(∇ × f¯ )
∇ × ∇ × F̄ = ∇ ∇.F̄ − ∇ 2 F̄.
Proof:
Taking f¯ = r̄, we get
∇ × ∇ × F̄
% ∂ ∇ × (φ(r)r̄) = (∇φ(r)) × r̄ + φ(r)(∇ × r̄)
= ī × ∇ × F̄
∂x r̄
% = φ (r) × r̄ + 0̄ ∵ ∇ × r̄ = 0̄
∂ ∂F̄ ∂F̄ ∂F̄ r
= ī × ī × + j̄ × + k̄ ×
∂x ∂x ∂y ∂z = 0̄ ∵ r̄ × r̄ = 0.
%
∂ F̄
2
∂ F̄
2
∂2 F̄
= ī × ī × 2 + j̄ × + k̄ × Example 10.17
∂x ∂x∂y ∂x∂z
% ∂2 F̄ ∂2 F̄
2
∂ F̄
Prove that ∇φ × ∇ψ is solenoidal.
= ī · 2 ī − 2 + ī. j̄
∂x ∂x ∂x∂y Solution A vector is solenoidal if its divergence
2
∂ F̄ is zero. So, we have to show that ∇ · (∇φ × ∇ψ) = 0.
+ ī. k̄
∂x∂z By Identity (4), we have
% ∂ ∂F̄ ∂

∂F̄

= ī · ī + ī · j̄ ∇ · f¯ × ḡ = ḡ · ∇ × f¯ − f¯ · ∇ × ḡ
∂x ∂x ∂y ∂x

∂ ∂F̄ ∂2 F̄
+ ī · − 2 Taking f¯ = ∇φ and ḡ = ∇ψ, we have
∂z ∂x ∂x
∵ F̄ has continuous second partial derivatives, ∇ · (∇φ × ∇ψ) = ∇ψ · (∇ × ∇φ)
∂2 F̄ ∂2 F̄ ∂2 F̄ ∂2 F̄ −∇φ · (∇ × ∇ψ) = 0
= , =
∂x∂y ∂y∂x ∂x∂z ∂z∂x ∵ ∇ × ∇φ = curl (grad φ) = 0
% ∂ ∂F̄ ∂

∂F̄

∇ × ∇ψ = curl (grad ψ) = 0.
= ī ī · + j̄ ī.
∂x ∂x ∂y ∂x
Example 10.18
∂ ∂F̄ ∂2 F̄
+ k̄ ī. − 2 For asolenoidal
vector
f¯
, prove that
∂z ∂x ∂x
∇ × ∇ × ∇ × ∇ × f¯ = curl curl curl curl f¯
% ∂ ∂ ∂

∂F̄

∂2 F̄

= ī + j̄ + k̄ ī · − 2 = ∇ 4 f¯ .
∂x ∂y ∂z ∂x ∂x
% ∂F̄ ∂2 F̄ Solution By Identity (9), we have
= ∇ ī · − 2
∂x ∂x
∇ × ∇ × f¯ = ∇ ∇ · f¯ − ∇ 2 f¯
% ∂F̄ % ∂2 F̄
=∇ ī. − = −∇ 2 f¯ (1)
∂x ∂x2

= ∇ ∇ · F̄ − ∇ F̄. 2
∵ f¯ is solenoidal and ∇ · f¯ = 0
Example 10.16 Applying ∇ × ∇× on Eq. (1), we get
Prove that the vector φ(r)r̄, where φ is a scalar, is
irrotational. ∇ × ∇ × ∇ × ∇ × f¯

Solution The vector φ(r)r̄ is irrotational if ∇ × = ∇ ∇ · −∇ 2 f¯ + ∇ 2 ∇ 2 f¯

(φ(r)r̄) = 0̄. That is, if its curl is a zero vector. = −∇ ∇ · ∇ 2 f¯ + ∇ 4 f¯ (2)

But ∇ 2 f¯ = ∇ 2 f1 ī + f2 j̄ + f3 k̄ Example 10.20
= ī∇ f1 + j̄∇ f2 + k̄∇ f3
2 2 2 If f¯ = ψ∇φ, then f¯ · curl f¯ = 0.

⇒ ∇ · ∇ 2 f¯ Solution
% ∂ 2 curl f¯ = ∇ × (ψ∇φ)
= ī · ī∇ f1 + j̄∇ 2 f2 + k̄∇ 2 f3
= ∇ψ × ∇φ + ψ (∇ × ∇φ)
∂x
∂f1 = ∇ψ × ∇φ
= ∇ 2
∂x ∵ ∇ × ∇φ = curl grad φ = 0

= ∇ ∇ · f¯ = ∇ 2 (0) = 0
2
(3) ¯f · curl f¯ = ψ∇φ · (∇ψ × ∇φ)
= [ψ∇φ, ∇ψ, ∇φ]
From Eqs.(1) and (2), we get
= ψ [∇φ, ∇ψ, ∇φ]

∇ × ∇ × ∇ × ∇ × f¯ = ∇ 4 f¯ . (4) = ψ · 0 = 0.

r̄ Prove that ∇ · (∇φ × ∇ψ) = 0.
Prove that 1. ∇f (r) = f (r)
r Solution By Identity (4), we have
2
2. ∇ f (r) = f (r)+ f (r) [JNTU 1995S, 2003S(4)]
2
∇ · f¯ × ḡ = ḡ · ∇ × f¯ − f¯ · ∇ × ḡ
r
Solution Take f¯ = ∇φ, ḡ = ∇ψ
∇ · (∇φ × ∇ψ) = ∇ψ · (∇ × ∇φ)−∇φ · (∇ × ∇ψ)
% ∂
1. ∇f (r) = ī f (r) = ∇ψ · 0̄ − ∇φ · 0̄ = 0.
∂x
% ∂ %
∂r x Example 10.22
= ī f (r) = if (r)
∂r ∂x r Show that 1. ∇ · (r n r̄) = (n + 3)r n . Hence show that

f (r) % 2. ∇ 2 (r n r̄) = n(n + 3)r n−2 r̄
= xīr 2 = x2 + y2 + z 2
r Solution By Identity 3(i), we have
f (r) ∂r
= xī + yī + zr̄ 2r
r ∂x 1. ∇ · φf¯ = (∇φ) · f¯ + φ ∇ · f¯
∂r x f (r)
= 2x ⇒ = = r̄. Take φ(r) = r n , f¯ (r) = r̄
∂x r r
∇ · (r n r̄) = (∇r n ) · r̄ + r n (∇ · r̄)
f (r)
2. ∇ 2 f (r) = ∇ · ∇f (r) = ∇ · r̄ ∵ ∇ · r̄ = 3, ∇f (r) = f (r) ∇r
r
= r̄ · nr n−1 (∇r) + 3r n
f (r) f (r)
= ∇ · r̄ + (∇ · r̄)
r r n−1 1 r̄
= r̄· nr r̄ +3r n ∵ ∇r =
1 1 r r
= ∇f (r) + f (r) ∇ · r̄ = (n + 3)r n .
r r
r̄
3f (r) 2. ∇ 2 r n = ∇ · (∇r n ) = ∇ · nr n−1
+ , ∵ ∇ · r̄ = 3 r
r n−2
1 f (r) 1 r̄ = n∇ · r r̄
= r̄ +f (r) − 2 · r̄
r r r r = n(n − 2 + 3)r n−2
3f (r) = n(n + 1)r n−2 by (1).
+
r
2 Note 1
= f (r) + f (r). n = −1, ∇ 2 =0
r r

1 ā · r̄
Prove that ā · ∇ = − 3 , where ā is a
r r 1. Prove that ∇ × (ā · r̄) = 0 where ā is a constant
constant vector.
vector.
Solution
1
1 1 1 r̄ 1 2. Prove that ∇ f¯ 2 = f¯ · ∇ f¯ + f¯ × curl f¯ .
∇ = − 2 ∇r = − 2 = − 3 r̄ 2
r r r r r
1 1 ā · r̄ 3. If f¯ (x, y, z) = xyī + yzi + zxk̄, show that ∇ 2 ī = 0.
⇒ ā · ∇ = ā · − 3 r̄ = − 3 .
r r r
1 n(n − 2)
4. Prove that ∇ · r∇ n = .
r r n+1
Example 10.24

Prove that 5. Prove that ∇ 2 r 2 log r = 5 + 6 log r.

1 3 (ā · r̄) b̄ · r̄ ā · b̄ 1
b̄· ā · ∇ = − 3 6. Prove that ∇ 2 log r = 2 .
r r5 r r
2
where ā, b̄ are constant vectors. 7. Prove that if f¯ = ∇ x + y2 + z 2 − 3xyz then
∇ · f¯ = 6 and ∇ × f¯ = 0̄.
Solution We have proved in Example 10.23 that
1 ā · r̄
ā · ∇ = − 8. If φ = x2 + y2 + z 2 and r̄ = xi + yj̄ + zw, then
r r3
% ∂ ā · r̄ show that div (φr̄) = 5φ.
1
b̄ · ∇ ā · ∇ = b̄ · ī − 3
r ∂x r 9. Prove that
%
1 ∂ ∇ · (φ∇ψ) − ∇ · (ψ∇φ) = φ∇ 2 ψ − ψ∇ 2 φ.
= b̄ · ī − 3 (ā · r̄)
r ∂x
f (r) 1
∂ 1 10. Prove that ∇ · r̄ = 2 r 2 f (r) .
+ (ā · r̄) − 3 r r
∂x r
% 11. Prove that curl (ψ∇φ) = ∇ψ×∇φ=−curl (φ∇ψ).
1 ∂r̄
= b̄ · ī − 3 ā ·
r ∂x
12. If ā is a constant vector and ū is any vector function,
1 ∂r then prove that
+ 3 (ā · r̄) 4 ,
r ∂x
(a) ∇ (ā · ū) = (ā · ∇) ū + ā× curl ū
∵ ā is constant
% ā · ī 3x (b) ∇ · (ā × ū) = −ā· curl ū
= b̄ · ī − 3 + 5 (ā · r̄) ,
r r (c) ∇ × (ā × ū) = ā (∇ · ū) − (a · ∇)ū
∂r̄ ∂r x
∵ = ī; =
∂x ∂x r
10.5 Curvilinear Coordinates
% ā · ī i % (ā · r̄) xī
= b̄ · − +3
r3 r5 10.5.1 Introduction

ā 3 % Many problems in engineering and physics are solved
= b̄ · − 3 + 5 (ā · r̄) xī , by choosing a suitable coordinate system. So, there is
r r
% % a need for the study of transformation of coordinates.
∵ ā · ī i = ā, xī = r̄ Further, we need to develop a general coordinate

ā · b̄ 3 (ā · r̄) b̄ · r̄ system from which we can derive all the other
=− 3 + . coordinate systems.
r r5
10.5.2 Transformation of 10.5.4 Unit Vectors in Curvilinear

Coordinates System
Let the rectangular coordinates (x, y, z) of a point P Let r̄ = xī + yj̄ + z k̄ be position vectors of a point
in space be expressed as functions of (q1 , q2 , q3 ) P, then r̄ = r̄ (q1 , q2 , q3 ). A tangent vector to the q1 -
⎫ ∂r̄
curve at P (for which q2 and q3 are constants) is
x = x q1 , q2, q3 , ⎬ ∂q1
y = y (q1 , q2 , q3 ) , (10.8) so that the unit tangent vector in this direction is
⎭
z = z (q1 , q2 , q3 )
∂r̄ ∂r̄ ∂r̄
ē1 =/ ⇒ = h1 e1
Suppose Eq. (10.1) can be solved for (q1 , q2 , q3 ) ∂q1 ∂q1 ∂q1
in terms of (x, y, z): ∂r̄
where h1 =
⎫ ∂q1
q1 = q1 x, y, z , ⎬
q2 = q2 x, y, z , (10.9) Similarly if ē2 , ē3 are unit tangent vectors to the
⎭ q2 - and q3 -curves at P, respectively.
q3 = q3 x, y, z
∂r̄ ∂r̄
Assume that functions (10.8) and (10.9) are single- = h2 ē2 and = h3 ē3
∂q2 ∂q3
valued and have continuous derivatives so that the
correspondence between (x, y, z) and (q1 , q2 , q3 ) is ∂r̄ ∂r̄
where h2 = , h3 =
unique. ∂q2 ∂q3
The coordinates q1 , q2 , q3 are called the curvilinear
The quantities h1 , h2 , h3 are called scale factors.
coordinates of point P and the set of Eqs. (10.8) or
The unit vectors ē1 , ē2 , ē3 are in the direction of
(10.9) defines a transformation of coordinates.
increasing q1 , q2 , q3 , respectively.
Since ∇u1 is a vector at P normal to the surface
10.5.3 Orthogonal Curvilinear u1 = e1 a unit vector in this direction is given by
Coordinates Ē1 = ∇q1 / |∇q1 |. Similarly, the unit vectors Ē2 =
The surfaces q1 = c1 , q2 = c2 , q3 = c3 where ∇q2 / |∇q2 | , Ē3 = ∇q3 / |∇q3 | at P are normal to the
c1 , c2 , c3 are constants are called coordinate sur- surfaces q2 = c2 and q3 = c3 , respectively.
faces and each pair of these surfaces intersects in Thus at each point of a curvilinear system there
curves called coordinate curves or lines (Fig. 10.6). exist, in general, two sets of unit vectors: ē1 , ē2 , ē3
If the coordinate surfaces intersect at right angles, tangent to the coordinate curves and Ē1 , Ē2 , Ē3
the curvilinear coordinate system is called orthogo- normal to the coordinate surfaces. The sets become
nal. The coordinates q1 , q2 , q3 are coordinate curves identical iff the curvilinear system is orthogonal.
of the curvilinear system and are analogous to the
x, y, z coordinate axes of a rectangular system. 10.5.5 Arc Length and Volume
Elements
e3 q3 Curve From r̄ = r̄(q1 , q2 , q3 ), we have
∂r̄ ∂r̄ ∂r̄
z q2 = c 2 d r̄ = dq1 + dq2 + dq3
q1 = c1 ∂q1 ∂q2 ∂q3
e3 = h1 dq1 ē1 + h2 dq2 ē2 + h3 dq3 ē3
P y
q1 Curve
Differential of arc length ds is determined from
e1 q3 = c3 (d s̄)2 = d r̄ · d r̄.
x q2 Curve For orthogonal system
ds2 = h21 dq12 + h22 dq22 + h23 dq32
Figure 10.6
q3 h1 ē1 h2 ē2 h3 ē3

1 ∂ ∂ ∂
= ∂q1 ∂q2 ∂q3
h3dq3e3 q2 h1 h2 h3
h1 A1 h2 A2 h3 A3
(4) ∇ = Laplacian of
2

h1dq1e1 P 1 ∂ h2 h3 ∂ ∂
h2dq2e2 = +
h1 h2 h3 ∂q1 h1 ∂q1 ∂q2

h3 h1 ∂ ∂ h1 h2 ∂
+
q1 h2 ∂q2 ∂q3 h3 ∂q3
Figure 10.7
If h1 = h2 = h3 = 1; ē1 , ē2 , ē3 are replaced by
Along a q1 -curve, q2 and q3 are constants so that ī, j̄, k̄ and q1 , q2 , q3 are taken as x, y, z we get the
d r̄ = h1 dq1 ē1 . Then the differential of arc length results of the rectangular cartesian coordinates.
ds1 along q1 at P is h1 dq1 . Similarly, the differential
arc lengths along q2 and q3 at P are ds2 = h2 dq2 , 10.6 Special Orthogonal
ds3 = h3 dq3 . Coordinate Systems
From Fig. 10.7 the volume element for an orthog- (1) Cylindrical coordinates (ρ, φ, z)
onal curvilinear coordinate system is given by
x = ρ cos φ, y = ρ sin φ, z=z
ρ ≥ 0, 0 ≤ φ ≤ 2π, −∞ < z < ∞
dV = |(h1 dq1 ē1 ) . (h2 dq2 ē2 ) × (h3 dq3 ē3 )|
hρ = 1, hφ = ρ, hz = 1
= h1 h2 h3 dq1 dq2 dq3
z
since |ē1 (ē2 × ē3 )| = 1. This formula, in terms of
ez
∂ x, y, z
Jacobian, is dV = dq1 dq2 dq3 .
∂ (q1 , q2 , q3 ) P eφ
eρ
z
10.5.6 Expressions for the o

ρ
y
Gradient, Divergence and
Curl in terms of Curvilinear x
Coordinates
Figure 10.8
If is a scalar function and Ā = A1 ē1 + A2 ē2 +
A3 ē3 is a vector function of orthogonal curvilinear (2) Spherical coordinates (r, θ, φ)
coordinates, q1 , q2 , q3 , then we have x = r sin θ cos φ, y = r sin θ sin φ, z = r cos θ
r ≥ 0, 0 ≤ φ ≤ 2π, 0≤θ≤π
(1) ∇ = grad hr = 1, hθ = r, hφ = r sin θ
1 ∂ 1 ∂ 1 ∂
= ē1 + ē2 + ē3
h1 ∂q1 h2 ∂q2 h3 ∂q3
z ez
(2) ∇.Ā = divĀ

1 ∂ ∂ eφ
= (h2 h3 A1 ) + (p, φ, z)
h1 h2 h3 ∂q1 ∂q2 P
z eρ
∂
(h3 h1 A2 ) + (h1 h2 A3 ) xφ ρ y
∂q3 y
(3) ∇ × Ā = curl Ā x
y
r= x2 + y2 , θ = arc tan , z = z
x
10.6.2 Spherical Coordinates

Coordinate surfaces:
r = C1 spheres with centre at the origin.
θ = C2 circular half-angle cones with the z-axis as
their axis with the vertex at the origin.
φ = C3 half-planes adjoining the z-axis through the
Figure 10.9 z-axis.
Coordinate axes:
r: rays emanating from the origin.
10.6.1 Cylindrical Coordinates θ: meridian on a sphere (vertical lines on a sphere).
Coordinate surfaces φ: parallels on a sphere (horizontal curves on a
ρ = C1 circular cylinders coaxial with the z-axis. sphere). Cartesian coordinates are related to spheri-
φ = C2 half-plane adjoining the z-axis. cal coordinate as follows:
z = C3 plane perpendicular to the z-axis. x = r cos φ sin θ
Coordinate (axes) lines
ρ: rays with the z-axis and perpendicular to the z-axis. y = r sin φ sin θ
φ: circles with centre on the z-axis and lying in the z = r cos θ.
plane perpendicular to the z-axis.
z: straight lines parallel to the z-axis. 10.6.3 Gradient in Orthogonal
Curvilinear Coordinates
Transformation formulae relating cartesian Let F (q1 , q2 , q3 ) be a scalar point function
coordinates to cylindrical coordinates
∇F = f1 ē1 + f2 ē2 + f3 ē3 (10.10)
where f1 , f2 , f3 are to be determined
r̄ = r̄ (q1 , q2 , q3 )
∂r̄ ∂r̄ ∂r̄
d r̄ = dq1 + dq2 + dq3
∂q1 ∂q2 ∂q3
= h1 ē1 dq1 + h2 ē2 dq2 + h3 ē3 dq3
∇F · d r̄ = h1 f1 dq1 + h2 f2 dq2 + h3 f3 dq3
= dF (10.11)
∂F ∂F ∂F
But dF = dq1 + dq2 + dq3 (10.12)
∂q1 ∂q2 ∂q3
From Eqs. (10.11) and (10.12), we have
∂F 1 ∂F
h1 f1 = ⇒ f1 =
∂q1 h1 ∂q1
∂F 1 ∂F
Figure 10.10 h 2 f2 = ⇒ f2 = (10.13)
∂q2 h2 ∂q2
∂F 1 ∂F
x = r cos θ, y = r sin θ, z = z h3 f3 = ⇒ f3 =
∂q3 h3 ∂q3
Substituting for f1 , f2 , f3 in Eq. (10.3), we get 1 ∂

= f1 h2 h3 + 0 + 0
h1 h2 h3 ∂q1
1 ∂f1 1 ∂f2 1 ∂f3
∇F = ē1 + ē2 + ē3 (10.14) 1 ∂
h1 ∂q1 h2 ∂q2 h3 ∂q3 = f1 h2 h3 (10.17)
h1 h2 h3 ∂q1
Hence, the general expression for the vector dif- Similarly
ferential operator ∇ in terms of ortho curivilinear
coordinates is 1 ∂
∇ · f2 ē2 = f2 h3 h1 (10.18)
ē1 ∂ ē2 ∂ ē3 ∂ h1 h2 h3 ∂q2
∇= + + . (10.15) 1 ∂
h1 ∂q1 h2 ∂q2 h3 ∂q3 ∇ · f3 ē3 = f3 h1 h2 (10.19)
h1 h2 h3 ∂q3
10.6.4 Divergence in Curvilinear Adding Eqs. (10.17), (10.18), (10.19) and using
Coordinates Eq. (10.19), we get
(1) Show that ē1 = h2 h3 ∇q2 × ∇q3 .
1 ∂ ∂
∇ · F̄ = f1 h2 h3 + f2 h3 h1
Solution With F = q1 in the above result, we h1 h2 h3 ∂q1 ∂q2
have ∂
+ f3 h1 h2
ē1 ∂q1 ē1 ∂q3
∇F = ∇q1 = +0+0=
h1 ∂q1 h1 Corollary Put h1 = h2 = h3 = 1,
ē2 ē3
Similarly, ∇q2 = , ∇q3 = q1 = x, q2 = y, q3 = z we get,
h2 h3
ē2 ē3 ē1 ∂f1 ∂f2 ∂f3
∇q2 × ∇q3 = × = ∇.F̄ = + + .
h2 h3 h2 h3 ∂x ∂y ∂z
⇒ ē1 = h2 h3 ∇q2 × ∇q3
Similarly, ē2 = h3 h1 ∇q3 × ∇q1 10.6.5 Curl in Curvilinear
Coordinates
ē3 = h1 h2 ∇q1 × ∇q2
(2) Let F̄ = F1 ē1 + F2 ē2 + F3 ē3 , so that F̄ = f1 ē1 + f2 ē2 + f3 ē3 ⇒ ∇ × F̄ (10.20)

= ∇ × f1 ē1 + ∇ × f2 ē2 + ∇ × f3 ē3
∇ · F̄ = ∇ · f1 ē1 + f2 ē2 + f3 ē3 (10.16)
Consider ∇ × f1 ē1 = ∇ × f1 h1 ∇q1
= ∇ · f1 ē1 + ∇ · f2 ē2 + ∇ · f3 ē3
since ē1 = h1 ∇q1
Consider
= ∇ f1 h1 × ∇q1 + f1 h1 ∇ × ∇q1
∇ · f1 ē1 = ∇ · f1 h2 h3 ∇q2 × ∇q3 ē1
= ∇ f1 h 1 × +0
= ∇ f1 h2 h3 · (∇q2 × ∇q3 ) h1
e2 ē3
= ∇ f1 h2 h3 × +0 Substituting the value of gradient
h2 h3
ē1 ē1 ∂ ē2 ∂
= ∇ f1 h2 h3 · ∇ × f1 ē1 = f1 h1 + f1 h1
h2 h3 h1 ∂q1 h2 ∂q2

ē3 ∂ ē1
Using ∇F from the above + f1 h 1 × (10.21)
h3 ∂q3 h1
ē1 ∂ ē2 ∂ ē3 ∂ ē2
∇ · f1 ē1 = f1 h2 h3 + f1 h2 h3 = 0− f1 h1 +
h1 ∂q1 h2 ∂q2 h3 h1 ∂q2 h3 h1

ē3 ∂ ē1 ∂
+ f1 h2 h3 · f1 h1
h3 ∂q3 h2 h3 ∂q3
and two similar results. Adding these, we obtain Example 10.26

Determine the transformation from cylindrical to
h1 ē1 h2 ē2 h3 ē3 rectangular cartesian coordinates.
1 ∂ ∂ ∂
∇ × F̄ =
h1 h2 h3 ∂q1 ∂q2 ∂q3 Solution The equations of transformation from
f1 h1 f2 h2 f3 h 3 rectangular cartesian to cylindrical coordinates are
x = ρ cos φ (1)
ī j̄ k̄ y = ρ sin φ (2)
∂ ∂ ∂
In cartesians, ∇ × F̄ = . z=z (3)
∂x ∂y ∂z
f1 f2 f3
Squaring Eqs. (1) and (2) and adding, we get
Example 10.25
ρ cos2 φ + sin2 φ = x2 + y2 or ρ2 = x2 + y2
2
Describe the coordinate surfaces and coordinate
curves for 1. cylindrical coordinates and 2. spherical ∵ cos2 φ + sin2 φ = 1, ρ ≥ 0
coordinates
From Eqs. (1) and (2), we get
Solution y ρ sin φ y
= = tan φ or φ = tan−1
1. Coordinate surfaces (level surfaces) x ρ cos φ x
ρ = c1 cylinders coaxial with the z-axis (or the
z-axis if c1 = 0). ∴ The required transformation is
φ = c2 half-planes through the z-axis. ρ= x2 + y2 (4)
z = c3 planes perpendicular to the z-axis. y
φ = tan−1 (5)
x
Coordinate curves z=z (6)
z curve: straight line which is the intersection of the
surfaces ρ = c1 and φ = c2 . For points on the z-axis, (i.e., x = 0, y = 0)
φ curve: circle (or point) which is the intersection of note that φ is indeterminate. These are called singular
ρ = c1 and z = c3 . points of the transformation.
ρ curve: straight line which is the intersection of
φ = c2 and z = c3 . Example 10.27
Prove that a cylindrical coordinate system is orthog-
2. Coordinate surfaces onal.
r = c1 spheres with centre at the origin (or origin
when c1 = 0). Solution The position vector of any point is
θ = c2 cones with vertex at the origin or lines if r̄ = xī + yj̄ + z k̄ = ρ cos φī + ρ sin φj̄ + z k̄,
c2 = 0 or π and the xy-plane if c2 = π/2. in cylindrical coordinates
φ = c3 half-planes through the z-axis.
∂r̄ ∂r̄ ∂r̄
Now, , , are tangent vectors to the ρ, φ, z
Coordinate curves ∂ρ ∂φ ∂z
φ-curve: circle (or point) which is the intersection curves, respectively.
of r = c1 and θ = c2 . ∂r̄
= cos φ ī + sin φ j̄;
θ-curve: semicircle (c1 = 0) which is the intersec- ∂ρ
∂r̄
tion of r = c1 and φ = c3 . = −ρ sin φ ī + ρ cos φ j̄;
∂φ
r-curve: straight line which is the intersection of θ = ∂r̄
c2 and φ = c3 . = k̄
∂z
The unit vectors in ρ, φ, z directions are, respec- ∂r̄ ∂r̄ ∂r̄

d r̄ = dρ + dφ + dz
tively, ∂ρ ∂φ ∂z
B
∂r̄ ∂r̄ cos φī + sin φj̄ = cos φ ī + sin φ j̄ dρ
ē1 = ēρ = =
∂ρ ∂ρ cos2 φ + sin2 φ + −ρ sin φ ī + ρ cos φ j̄ dφ + k̄ dz
= cos φī + sin φj̄ = (cos φdρ − ρ sin φdφ) ī
B
∂r̄ ∂r̄ −ρ sin φī + ρ cos j̄ + (sin φ dρ + ρ cos φ dφ) j̄ + k̄ dz
ē2 = ēρ = =
∂φ ∂φ sin2 φ + cos2 φ ∴ (ds)2 = d r̄ · d r̄ = (cos φ dρ − ρ sin φ dφ)2
= − sin φī + cos φj̄ + (sin φ dρ + ρ cos φ dφ)2 + (dz)2
B
∂r̄ ∂r̄ = (dρ)2 + ρ2 (dφ)2 + (dz)2 .
ē3 = ēz = = k̄
∂z ∂z
Now, Example 10.29

ē1 · ē2 = cos φī + sin φj̄ · − sin φī + cos φj̄ = 0 Find the square of the element arc length in spherical
coordinates and determine the corresponding scale
ē1 · ē3 = cos φj + sin φj̄ · k̄ = 0
factors.
ē2 · ē3 = − sin φī + cos φk̄ · k̄ = 0
So ē1 , ē2 , ē3 are mutually perpendicular and the Solution We have
coordinate system is orthogonal. x = r sin θ cos φ, y = r sin θ sin φ, z = r cos θ
Example 10.28 dx = −r sin θ sin φ dφ + r cos θ cos φ dθ

Find the square of the element of arc length in cylin-
+ sin θ cos φ dr
drical coordinates and determine the corresponding
scale factors. dy = r sin θ cos φ dφ + r cos θ sin φ dθ
+ sin θ sin φ dr
Solution dz = −r sin θ dθ + cos θ dr
Method 1
2
We have ∴ (ds)2 = (dx)2 + dy + (dz)2
x = ρ cos φ, y = ρ sin φ, z = z = (−r sin θ sin φdφ + r cos θ cos φdθ
dx = −ρ sin φdφ + cos φdρ, + sin θ cos φdr)2 + (r sin θ cos φ dφ
dy = ρ cos φdφ + sin φdρ, dz = dz +r cos θ sin φ dθ + sin θ sin φ dr)2
2
(ds)2 = (dx)2 + dy + (dz)2 + (−r sin θ dθ + cos θ dr)2

= (−ρ sin φdφ + cos φdρ)2 = (dr)2 sin2 θ sin2 φ + cos2 φ + cos2 θ

+ (ρ cos φdφ + sin φdρ)2 + (dz)2 + (rdθ)2 cos2 θ cos2 φ + sin2 φ

= (dρ)2 + ρ2 (dφ)2 + (dz)2 + sin2 θ +(r sin θdφ)2 sin2 φ + cos2 φ
= h21 (dρ)2 + h22 (dφ)2 + h23 (dz)2 +2r 2 dθdφ [− sin θ sin φ cos θ cos φ
h1 = hρ = 1, h2 = hφ = ρ, h3 = hz = 1 + sin θ cos φ · cos θ sin φ]
+2rdθdr [− cos θ sin φ sin θ sin φ
are the scale factors.
+ cos θ sin φ · sin θ sin φ]
Method 2 +2rdφdr [− sin θ sin φ · sin θ cos φ
The position vector of any point is + sin θ cos φ · sin θ sin φ]
r̄ = ρ cos φ ī + ρ sin φ j̄ + z k̄ = (dr)2 + r 2 (dθ)2 + r 2 sin2 θ (dφ)2 .
Example 10.30 Solution The area elements are given by

Sketch a volume element in 1. cylindrical and 2.
dA1 = |h2 dq2 ē2 × h3 dq3 ē3 |
spherical coordinates giving magnitudes of its edges.
= h2 h3 |ē2 × ē3 | dq2 dq3 = h2 h3 dq2 dq3
since |ē2 × ē3 | = |ē1 | = 1
z ez Similarly
eφ dA2 = |h1 dq1 ē1 × h3 dq3 ē3 | = h1 h3 dq1 dq3
(p, φ, z)
P dA3 = |h1 dq1 ē1 × h2 dq2 ē2 | = h1 h2 dq1 dq2 .
z eρ
Example 10.33
x φ ρ y
y
If q1 , q2 , q3 are orthogonal curvilinear coordinates
find the Jacobian of x, y, z with respect to q1 , q2 , q3 .
x
Solution

x, y, z ∂ x, y, z
z
er J =
q1 , q2 , q3 ∂ (q1 , q2 , q3 )
eφ ∂x ∂y ∂z
(p, φ, z)
r ∂q1 ∂q1 ∂q1
θ z
xφ r y ∂x ∂y ∂z
y' eθ
= ∂q2 ∂q2 ∂q2
x ∂x ∂y ∂z
∂q3 ∂q3 ∂q3
Figure 10.11

∂x ∂y ∂z
Example 10.31 = ī+ j̄+ k̄
Find the volume element dV in 1. cylindrical and 2. ∂q1 ∂q1 ∂q1
∂x ∂y ∂z
spherical coordinates. × ī+ j̄+ k̄
∂q
2 ∂q 2 ∂q 2
Solution The volume element in orthogonal ∂x ∂y ∂z
× ī + j̄ + k̄
curvilinear coordinates q1 , q2 , q3 ∂q3 ∂q3 ∂q3
dV = h1 h2 h3 dq1 dq2 dq3
1. In cylindrical coordinates, q1 = ρ, q2 = φ, q3 = ∂r̄ ∂r̄ ∂r̄
= · ×
z; scale factors are ∂q1 ∂q2 ∂q3
h1 = 1, h2 = ρ, h3 = h1 ē1 · h2 ē2 × h3 ē3
dV = 1 · ρ · 1 · dρdφdz = ρdρdφdz. = h1 h2 h3
2. In spherical coordinates q1 = r, q2 = θ, q3 = φ; [∵ ē1 · ē2 × ē3 = 1]
scale factors are Note
h1 = 1, h2 = r, h3 = r sin θ ∂r̄ ∂r̄ ∂r̄
Jacobian = 0 ⇒ the vectors , , are
dV = 1 · r · r sin θdrdθdφ = r 2 sin θdrdθdφ. ∂q1 ∂q2 ∂q3
coplanar.
Example 10.32 The curvilinear coordinates transformation fails
Find expressions for the elements of area in orthog- since x, y, z are now related in the form F(x, y, z) = 0.
onal curvilinear coordinates. So, we require J = 0.
EXERCISE 10.4 to the√surface ψ(x, y, z) = 3xy2 + y + z at (0, 1, 1, ).

4 11
Ans:
1. Find the directional derivative of φ(x, y, z) = 11
xy2 + yz 2 at (2, −1, 1) in the direction of the vector 4. Prove that ∇φ · d r̄ = dφ.
ī + 2j̄ + 2k̄.
Ans: −11/3 1 ā · r̄
5. Prove that ā · ∇ =− 3 .
r r
2. Find the angle between the surfaces x2 + y2 + z 2 =
3 = x2 + y2 at P = (−2, 1, 2).
9, and z + √ 6. If φ1 and φ2 are scalar point functions, then show
8 21 that the components of φ1 along and normal to the
Ans: cos−1 ∇φ2 × (φ1 × ∇φ2 )
63 level surface φ2 = 0 are and
(∇φ2 )2
3. Find the directional derivative of φ(x, y, z) = (φ1 · ∇φ2 ) ∇φ2
xyz 2 + xz at (1, 1, 1) in the direction normal .
(∇φ2 )2
Vector Integral Calculus
11
11.1 Vector Integral Calculus interior of a sphere or cube in space, and the
domain between two concentric spheres are simply-
11.1.1 Curves in a connected, while the domain between concentric cir-
Simply-Connected Domain cles in a plane and the interior of a torus in space are
Let C be a curve in space represented by multiply-connected.

r̄ (t) = x (t) , y (t) , z (t)
11.1.2 Line Integral
= x (t) ī+y (t) j̄+z (t) k̄ (a ≤ t ≤ b) (11.1)
The concept of a line integral is a simple and natural
The point A with position vector r̄(a) is called the generalisation of
initial point and the point B with position vector r̄(b) b
is called the terminal point. The direction from A to f (x) dx (11.2)
B in which t increases is called the positive direction
a
on C. If A and B coincide, then C is called a closed
path. which is known from real calculus. In Eq. (11.2) we
An open connected set is called a domain. A integrate f (x) along the x-axis from x = a and x = b.
domain D is called simply-connected if every closed In a line integral, we integrate a given function along
curve in D can be continuously shrunk to a point a curve on a plane or in space. Hence, it is actually a
in D without leaving D. Otherwise, D is called curve integral but it has become a standard practice
multiply-connected. to call it a line integral. Also, C is called the path of
integration.
Example
C is called a smooth curve if it has a unique
The interior of a circle or ellipse on a plane, the tangent at each of its points whose direction varies
continuously as we move along C.
This means that C has a representation such that
r̄(t) is differentiable and the derivative
d r̄
r̄ (t) =
dt
is continuous and different from the zero vector at
every point of C.
Let F̄(P) be a continuous vector point function
defined at every point P of a simply connected
domain D and let C be a curve in D. Divide the curve
C into n parts by means of the points
Figure 11.1 A = P0 , P1 , . . ., Pk , Pk+1 , . . ., Pn = B
Pn = B Example
11.2
Evaluate xdy−ydx around the circle C : x2 +y2 = 1.
C
Pk+1
Qk Solution The given curve C is the circle
Pk
x2 + y2 = 1 whose parametric equations are
x = cos t, y = sin t and t varies from t = 0 to
c t = 2π
2π
dy dx
(xdy − ydx) = x −y dt
P1 dt dt
C 0
A = P0
2π
Figure 11.2 = [cos t ·cos t −sin t(−sin t)] dt
0
whose position vectors are = 2π.
r̄0 , r̄1 , · · · , r̄k , r̄k+1 , · · · , r̄n Example 11.3
Evaluate

respectively. Let Qk be any point on the arc Pk Pk+1 3
2xy − y2 cos x dx+ 1 − 2y sin x + 3x2 y2 dy
and δr̄k = r̄k − r̄k−1 .
C
If the sum %
n
where C is the arc of the parabola 2x = πy2 from
F̄ Qk · δr̄k π
(0, 0) to ,1 .
k=0 2
tends to a finite limit as n → ∞ and δr̄k → 0̄ then Solution We know that Mdx + Ndy is an exact
we say that F̄ is integrable along the curve C and ∂M ∂N
the limit is called a line integral of F̄ along C. It is differential if = .
∂y ∂x
denoted by
Here M = 2xy3 − y2 cos x,

d r̄ ∂M
F̄ · d r̄ or F̄ · dt. = 6xy2 − 2y cos x
dt ∂y
C C
N = 1 − 2y sin x + 3x2 y2 ,
Example 11.1 ∂N ∂M
= −2y cos x + 6xy2 =
Evaluate F̄ · d r̄ where F̄ = cos yī − x sin yj̄ and C ∂x ∂y
√ ∴ The integral is an exact differential dφ so that
is the curve: y = 1 − x2 in the xy-plane from (1,0)
to (0,1). ∂φ ∂φ
dφ = dx + dy
∂x ∂y
Solution 3
F̄ = cos y, −x sin y = 2xy − y2 cos x dx

r̄ = (x, y), dt = (dx, dy) + 1 + 2y sin x + 3xy2 dy
F̄ · d t̄ = x cos ydx − x sin ydy = d (x cos y) ∂φ
⇒ = 2xy3 − y2 cos x,
(0,1) ∂x
F̄ · d t̄ = d(x cos y) = x cos y (1,0) ⇒ φ(x, y) = x2 y3 − y2 sin x + f1 (y) (1)
c c ∂φ
= 0 · cos 1 − 1 · cos 0 = −1. = 1 − 2y sin x + 3x2 y2 ,
∂y
Vector Integral Calculus 11-3
⇒ φ(x, y) = y − y2 sin x + x2 y3 + f2 (x) (2) 11.1.4 Applications

The values of φ given by Eqs. (1) and (2) agree if (1) Work done by a variable force
f1 (y) = y and f1 (x) = 0. Consider a body on which a constant force F̄ acts
∴ The given integral = dφ and moves the body through a displacement d̄. Then
C
the work done is

W = F̄ d̄ cos α = F̄ · d̄
= d y − y2 sin x + x2 y3
C If the force F̄ is variable and the displacement is

3 (π/2,1) from A to B along a curve C, then the work done is
= y − y2 sin x + x2 y (0,0)
B
π π2 d r̄
= 1−1·sin + · 1−0 W = F̄ · d r̄ = F̄ (r̄ (t)) . dt
2 4 dt
C A
π2
= . t2
4 dx dy dz
= F1 + F 2 + F 3 dt
dt dt dt
11.1.3 Properties t1
Consider the curve C in Fig. 11.3. The following t2

properties are satisfied by the line integral: = F1 dx + F2 dy + F3 dz
t1
b
where F̄ = F1 ī + F2 j̄ + F3 k̄ and
c2
r̄ = xī + yī + z k̄
c
c1
Figure 11.4
a Circulation of F̄ When F̄ denotes the velocity of

a fluid the circulation of F̄ around a simple closed
Figure 11.3 curve C in a simply connected domain D is defined

by
(1) k F̄ · d r̄ = k F̄ · d r̄, k = constant
Circulation = F̄.d r̄ = F1 dx + F2 dy
C C

C
(2) F̄ ± Ḡ · d r̄ = F̄ · d r̄ ± Ḡ · d r̄ If the circulation of F̄ around every simple closed
C C C curve in a domain D vanishes, then F̄ is said to be
irrotational in D.
(3) F̄ · d r̄ = F̄ · d r̄ + F̄ · d r̄
(2) Independence of path: Conservative field and
C C1 C2
scalar potential
where C = c1 + c2
A line integral
b a

(4) F̄ · d r̄ = − F̄ · d r̄. F̄ (r̄) · d r̄ = F1 dx + F2 dy + F3 dz (11.3)
a b C C
C
path, then the integral symbol is sometimes used.
C
We assume that every path of integration we consider
here will be a piece-wise smooth, i.e., it consists
of many finitely smooth curves. Such line integrals
arise naturally in mechanics e.g. in calculating the
work done by a force F̄ in a displacement along
C, in determining the conservative field, in finding
scalar potential, in determining the exact differential
and in calculating the area of a region.

Find the circulation of F̄ round the curve C where
with continuous F1 , F2 , F3 in a domain D in space F̄ = yī + z j̄ + xk̄ and C is the circle x2 + y2 = 1,
is independent of path in D if F̄ = (F1 , F2 , F3 ) is the z=0
gradient of a scalar function φ in D such that F̄ = ∇φ
Solution By C definition, the circulation of F̄ along
or, in components,
the curve C is d r̄ where r̄ = xī + yj̄ + 2k̄
∂φ ∂φ ∂φ D C
F1 = , F2 = , F3 = = (y, z, x) · (dx, dy, dz)
∂x ∂y ∂z
C
The above integral, in which the integrand is a D D
scalar function is called the tangential line integral = (ydx + zdy + xdz) = ydx
of F̄ along C. C C
If the scalar product in Eq. (11.3) is replaced by the ∵ on C, z = 0 and dz = 0
vector
product F̄ × d r̄, then we have the line integral 2π
F̄ × d r̄, which is a vector.
C
= sint(− sin t)dt x = cos t, y = sin t, on C
Also, we can replace the vector F̄ by a scalar 0
function φ so that we have φd r̄ which is again a 2π 2π
C 1 1 1
vector. =− (1−cos 2t)dt = − t − sin 2t = −π.
2 2 2 0
Let F̄ (r̄) = F̄ x, y, z = F1 ī + F2 j̄ + F3 k̄ and 0
r̄ (t) = x (t) ī + y (t) j̄ + z (t) k̄ be the parametric

Example 11.5
representation of the curve C. We have d r̄ = dxī +
dyj̄ + dz k̄. Now the integral (11.3) is Evaluate F̄.dr where F̄ = 1 + x2 ī − y3 j̄ and C
C
F̄ · d r̄ = F1 dx + F2 dy + F3 dz is the arc of the parabola y = x2 in the xy-plane from

(0,0) to (1,1).
C C
t2 Solution
dx dy dz
= F1 + F2 + F3 dt Method 1
dt dz dt
t1 C : y = x2 (0,0) to (1,1)
Let x = t then y = t 2 . Let r̄ be the position vector
where t = t1 at the initial point A and t = t2 at the of a point P on C.
terminal point B.
d r̄ r̄ (t) = x (t) ī + y (t) j̄ + z (t) k̄ = t ī + t 2 j̄
If t is replaced by s (the arc length of C) F̄ ·
ds d r̄
gives the tangent component of F̄. If C is a closed = ī + 2t j̄
dt
and (ii) along the curve C defined by x2 = 4y, 3x2 =

8z from x = 0 to x = 2.
Solution
(i) Force field is F̄ = 3x2 ī + 2xz − y j̄ + z k̄
Equations of the straight line joining (0, 0, 0) to
(2, 1, 3) are
x y z
= = = t say ⇒ x, y, z = (2t, t, 3t)
2 1 3
Figure 11.6
The points (0, 0, 0) and (2,1,3) correspond to t = 0

Also F̄ = 1 + t ī − t 6 j̄
2 and 1, respectively.
At (0, 0), t = x = 0. At the point (1, 1), t = 1 Work done by F̄ is
1
d r̄
F̄.d r̄ = F̄ · dt F̄ ·d r̄= 3 (2t)2 2dt +(2·2t ·3t − t) dt +3t ·3dt
dt
C C C 0
1 1
2 1
= 1 + t ī − t j̄ · ī + 2t j̄ dt
2 6
= 36t +8t dt = 12t 3 +4t 2 0 = 16.
0 0
1 1
1 2 (ii) Let x = t. Then the equations of the curve C are
= 1+t 2 −2t 7 dt = t + t 3 − t 8
3 8 0
0 1 2 1 3 3 3
y= x = t2, z= x = t3
1 1 13 4 4 8 8
= 1+ − = .
3 4 12 where t varies from 0 to 2.
Method 2
Work done by F̄ is
In the xy-plane we have 2 2
8 3 1 2 t
F̄ · d r̄ = 3t dt + 2t · t − t d
2
r̄ = xī + yj̄ ⇒ d r̄ = dxī + dyj̄ 3 4 4
C 0
F̄ = (1 − x2 )ī − y3 j̄
3t 3 3 3
∴ F̄ · d r̄ = 1 + x2 dx − y3 dy + d t
8 8
(1,1) 2

F̄ · d r̄ = (1 + x2 )dx − y3 dy 3 1 27
= 3t 2 + · t 5 − t 3 + t 5 dt
C (0,0)
8 8 64
0
2
1 3 1 4 (1,1) 1 4 51 t 6
= x+ x − y = t3 − t + ·
3 4 (0,0) 32 64 6 0
1 1 13 1 17
= 1+ − = . = 23 − · 24 + · 26
3 4 12 32 128
1 17
Example 11.6 = 8− + = 16.
Find the work done in moving a particle in the force 2 2
field Example 11.7
F̄ = 3x2 ī + 2xz − y j̄ + z k̄
If F̄ = 2x + y ī + 3y − x j, evaluate F̄ · d r̄
(i) along the straight line from (0, 0, 0) to (2, 1, 3). C
where C is the curve in the xy-plane consisting of r̄ (t) = x (t) ī + y (t) j̄ + z (t) k̄, (a ≤ t ≤ b) then
straight line segments from (0, 0) to (2, 0) and then b
from (2,0) to (3, 2). F̄ · d r̄ is independent of the paths and its value
a
Solution The path of integration C is the line depends on the end points only.
segments OA and AB shown in the figure.
Proof:
F̄ = 2x + y ī + 3y − x j̄
∂φ ∂φ ∂φ
r̄ = xī + yj̄, d r̄ = dxī + dyj̄ F̄ = gradφ = ī + j̄ + k̄
∂x ∂y ∂z
F̄ · d r̄ = 2x + y dx + 3y − x dy ∂φ ∂φ ∂φ
⇒ F1 = , F2 = , F3 =
∂x ∂y ∂z
B

Now, F1 dx + F2 dy + F3 dz
A
B
∂φ ∂φ ∂φ
= dx + dy + dz
∂x ∂y ∂z
A
b
∂φ dx ∂φ dy ∂φ dz
= + + dt
∂x dt ∂y dt ∂z dt
Figure 11.7 a
Along OA b
dφ b
y = 0, dy = 0 x varies from 0 to 2 = dt = φ x (t) , y (t) , z (t)
dt t=a
F̄ · d r̄ = 2xdx + 0 a
2 = φ (B) − φ (A)
2
F̄ · d r̄ = 2xdx = x2 0 = 4
OA 0 This shows that the value of the integral is simply
the difference of the values of φ at the two end points
Along AB of C and is, therefore, independent of the path C.
Equation of AB is (Proof of the converse is omitted.)
2−0
y−0 = (x−2) ⇒ y = 2x−4; dy = 2dx
3−2 11.1.5 Integration Around Closed
F̄ · d r̄ = (2x+2x−4) dx+[3 (2x−4)−x] 2dx Curves and Independence of
3 Path
3
F̄ · d r̄ = (14x − 28) dx = 7x2 − 28x x=2 Theorem 11.2 The integral

AB x=2
F̄.d r̄ = F1 dx + F2 dy + F3 dy (11.4)
= 7 32 − 22 − 28 (3 − 2) = 7
C C
F̄ · d r̄ = F̄ · d r̄ + F̄ · d r̄ = 4 + 7 = 11. is independent of path in a domain D if its value

C OA AB around every closed path in D is zero.
Theorem 11.1 Let F̄ = grad φ hold for Proof: Assume that Eq. (11.4) is independent of
some function φ in D. Let C be any path in path in D. Let C1 and C2 be any two paths, joining
D joining a point A to a point B given by any two points A and B in D
11.1.7 Criterion for Exactness and

Independence of Path
Let F̄ = F1 ī + F2 j̄ + F3 k̄ where the components
F1 , F2 , F3 are continuous and have continuous first
partial derivatives in D. Then

(i) if F̄ · d r̄ = F1 dx + F2 dy + F3 dz (11.6)
C C
Figure 11.8
is independent of path in D and thus the differential
form
F̄ · d r̄ = F̄ · d r̄ F1 dx + F2 dy + F3 dz
ABP AQB
under the integral sign is exact, then curl F̄= 0̄ in D.
⇒ F̄ · d r̄ − F̄ · d r̄ = 0 In component form, this means
ABP

AQB
∂F3 ∂F2 ∂F1 ∂F3 ∂F2 ∂F1
= , = , = (11.7)
⇒ F̄ · d r̄ + F̄ · d r̄ = 0 ∂y ∂z ∂z ∂x ∂x ∂y
ABP BQA
(ii) If Eq. (11.7) holds in D and D is simply-
connected then Eq. (11.6) is independent of path in
⇒ F̄ · d r̄ = 0 ⇒ F̄ · d r̄ = 0
D.
ABPQA C
Retracing the steps, we get the proof of the Proof: If Eq. (11.6) is independent of path in D
converse of the theorem. then
F̄ = grad φ
11.1.6 Exactness and Independence for some scalar point function φ. This implies that
of Path
The differential form curlF̄ = curl (grad φ) = 0̄
F1 dx + F2 dy + F3 dz (11.5)
The converse of the theorem requires Stokes’
is called exact in a domain D in space if it is the theorem given in Section 11.5.
differential
∂φ ∂φ ∂φ Example 11.8
dφ = dx + dy + dz
∂x ∂y ∂z If F̄ = 5xy − 6x2 ī+ 2y − 4x j̄, evaluate F̄ · d r̄
of a differentiable function φ(x, y, z) everywhere in C
D. Thus Eq. (11.5) is exact if and only if there is a along the curve C in the xy-plane: y = x3 from the
differentiable function φ(x, y, z) such that point (1, 1) to (2, 8).
∂φ ∂φ ∂φ
F1 = , F2 = , F3 = Solution
∂x ∂y ∂z

or in vectorial form F̄ = grad
φ everywhere in D. F̄ = 5xy − 6x2 ī + 2y − 4x j̄
Therefore, the integral F̄ · d r̄ is independent of d r̄ = dxī + dyj̄
path in D iff the differential form F1 dx + F2 dy +
F3 dz has continuous components F1 , F2 and F2 , and F̄ · d r̄ = 5xy − 6x2 dx + 2y − 4x dy
is exact in D. C : y = x3 ⇒ dy = 3x2 dx
2

F̄ · d r̄ = 5x x3 − 6x2 dx + 2x3 − 4x · 3x2 dx
C x=1
2

= 5x4 − 6x2 + 6x5 − 12x3 dx
x=1
2
= x5 − 2x3 + x6 − 3x4 1

= 25 − 1 − 2 23 − 13 + 26 − 16

−3 24 − 1 Figure 11.9
= 31 − 2 (7) + 63 − 3 (15) = 35. 0

Example 11.9 F̄ · d r̄ = e0 · sin π/2 · 0 + e0 · cos y dy
Find the circulation of F̄ round the curve C where CO π/2
F̄ = ex sin y ī + ex cos yj̄ and C is the rectangle with
= [sin y]0π/2 = 0 − 1 = −1
vertices (0, 0) , (1, 0) , (1, π/2) , (0, π/2) . ⎛ ⎞

Solution ∴ F̄ · d r̄ = ⎝ + + + ⎠ F̄ · d r̄

F̄d r̄ = ex sin y, ex cos y · dx, dy C OA AB BC CO
= ex sin ydx + ex cos ydy = 0 + e + (1 − e) + (−1) = 0.

Example 11.10
We have to evaluate F̄.d r̄ round the rectangle C (a,o)
C Evaluate I = xdy + ydx along the upper half of
which comprises the four line segments OA, AB, BC
(−a,o)
and CO.
the circle: x + y = a2 .
2 2
Along OA
y = 0, dy = 0; x varies from 0 to 1 Solution Changing to polar coordinates by
putting
F̄ · d r̄ = (ex · 0 + ex · 1.0) = 0. x = a cos θ, y = a sin θ,
OA OA
dx = −a sin θdθ, dy = a cos θdθ
Along AB
Limits for θ are 0 to π corresponding to x = a and
x = 1, dx = 0; y varies from 0 to π/2
−a, respectively.
1 0
F̄ · d r̄ = e · sin y · 0 + e1 · cos ydy 2
AB AB
I = a cos2 θ − a2 sin2 θ dθ
π/2 π
= [e · sin y]0 = e.
0 0
Along BC sin 2θ
=a 2
cos 2θdθ = a 2
= 0.
y = π/2, dy = 0; x varies from 1 to 0 2 π
π
0
Example 11.11
F̄ · d r̄ = (ex · 1 · dx + ex · 0)
BC 1
Evaluate y2 dx − x2 dy about a triangle whose ver-
= [ex ]01 = 1 − e. tices are (1, 0), (0, 1), (−1, 0). [JNTU 2003]
Along CO Solution Let A = (1, 0), B = (0, 1), C = (−1, 0)
x = 0, dx = 0; y varies from π/2 to 0 be the vertices of the triangle ABC. Then

2 Example
y dx−x dy =
2 2
+ + y dx−x2 dy 11.12

ABC AB BC CA If F̄ = y − 2x ī+ 3x + 2y j̄, calculate the circula-
tion of F̄ about the circle C in the plane: x2 + y2 = 4
oriented in the anticlockwise direction.
Solution The curve is C: x2 + y2 = 4 in the

xy-plane.
Put x = 2 cos t, y = 2 sin t, t varies from 0 to 2π
dx = −2 sin tdt, dy = 2 cos tdt

F̄ = y − 2x, 3x + 2y ; d r̄ = dx, dy
Figure 11.10
F̄ · d r̄ = y − 2x dx + 3x + 2y dy
Along AB = (2 sin t − 4 cos t) (−2 sin tdt)
Equation of AB: y = 1 − x, dy = −dx; x varies
from 1 to 0 + (6 cos t + 4 sin t) (2 cos tdt)

0 = −4 sin 2 t + 8 sin t cos t + 12 cos2 t

y dx − x dy =
2 2
(1 − x)2 dx − x2 (−dx) +8 sin t cos t) dt

AB 1 = −4 + 16 cos2 t + 16 sin t cos t dt
0

= 1 − 2x + 2x2 dx 2π 2π 2π
1 F̄ ·d r̄= −4dt +16 cos tdt +16 sin tcos tdt
2
0
2 C 0 0 0
= x − x2 + x3 π/2 2π
3 16
1
= (−4)·2π+16·4 cos tdt + · sin 2tdt
2
2 2 2
= − 1−1 + ·1 =− .
2
0 0
3 3
1 π cos 2t 2π
Along BC = −8π + 16 · 4 · · + 8 −
Equation of BC: y = 1 + x, dy = dx; x varies from 2 2 2 0
0 to –1 = 8π + 0 = 8π.
−1 Example 11.13

y dx − x dy =
2 2
(1 + x)2 dx − x2 dx Evaluate

3
BC 0 2xy − y2 cos x dx + 1 − 2y sin x + 3x2 y2 dy
−1
−1 C
= (1 + 2x) dx = x + x2 0 where C is the arc of the parabola 2x = πy2 from
0 (0, 0) to (π/2, 1).
= (−1 + 1) − 0 = 0.
Solution Here
Along CA
Equation of CA: y = 0 x varies form –1 to 1 F̄ · d r̄ = 2xy3 − y2 cos x dx

1 + 1 − 2y sin x + 3x2 y2 dy

y dx − x dy =
2 2
0 · dx − x2 · 0 = 0 = 2xy3 dx + 3x2 y2 dy + −y2 cos xdx

CA 0 + 1 − 2y sin x dy

2 2 = d x2 y3 + d y − y2 sin x
∴ y2 dx − x2 dy = − + 0 + 0 = − .
3 3 = d x2 y3 + y − y2 sin x = dφ say,
ABC

which is an exact differential. ∂ 2 ∂
= ī 3xz + 2 − 2y sin x − 4
(π/2,1) (π/2,1) ∂y ∂z
F̄ · d r̄ = φ (0,0) = x2 y3 + y − y2 sin x (0,0)
∂ 2 ∂ 2
+ j̄ y cos x + z −
3
3xz + 2
C
∂z ∂x
π2 π π2
= · 1 + 1 − 12 · sin − 0 = . ∂ ∂ 2
4 2 4 + k̄ 2y sin x−4 − y cos x+z 3
∂x ∂y
Example 11.14 2
Find the circulation of F̄ around the curve C where = ī (0 − 0) + j̄ 3z − 3z 2

F̄ = yī+z j̄+xk̄ and C is the circle x2 +y2 = 1, z = 0. +k̄ 2y cos x − 2y cos x = 0̄
∴ F̄ is conservative.
Solution
(ii) Let φ be a scalar potential such that F̄ = ∇φ
F̄ = (y, z, x), d r̄ = (dx, dy, 0)
∂φ ∂φ ∂φ
C is the unit circle x2 + y2 = 1, z = 0 so that ⇒ F1 ī + F2 j̄ + F3 k̄ = ī + j̄ + k̄
∂x ∂y ∂z
x = cos θ, y = sin θ, z = 0, dz = 0 ∂φ
⇒ = F1 = y2 cos x + z 3 (1)
D θ are 0 and
Limits for D 2π ∂x
∂φ
F̄ · d r̄ = ydx + zdy = F2 = 2y sin x − 4 (2)
∂y
C C
∂φ
2π = 3xz 2 + 2 (3)
∂z
= sin θ (− sin θdθ) + 0
Integrating Eq. (1) partially with respect to x
0
φ = y2 sin x + z 3 x + f y, z (4)
2π
1
=− (1 − cos 2θ)dθ where f is an arbitrary function of y and z.
2 Differentiating Eq. (4) partially with respect to y
0
2π and equating with Eq. (2)
1 1
=− θ − sin 2θ ∂φ ∂f
2 2 0 = 2y sin x + 0 +
∂y ∂y
1
= − (2π − o) = −π. = 2y sin x − 4
2
∂f
Example 11.15 ⇒ = −4 (5)
Test if F̄ = y2 cos x + z 3 ī + 2y sin x − 4 j̄ ∂y
2
+ 3xz + 2 k̄ is Integrating Eq. (5) with respect to y
(i) a conservative field, f = −4y + g (z) (6)
(ii) find the scalar potential of F̄ and
where g is a function of z.
(iii) compute the work done in moving an object in
Substituting Eq. (6) in Eq. (4)
this field from (0, 1, −1) to (π/2, −1, 2).
φ = y2 sin x + z 3 x − 4y + g (z) (7)
Solution (i) A field F̄ is conservative if
∇ × F̄ = 0̄. Here Differentiating Eq. (7) partially with respect to z
and equating with Eq. (3)
F̄ = y2 cos x + z 3 ī+ 2y sin x − 4 j̄+ 3xz 2 + 2 k̄
∂φ dg
ī j̄ k̄ = 0 + 3z 2 x − 0 + +z
∂z dz
∇ × F̄ = ∂/∂x ∂/∂y ∂/∂z
= 3xz 2 + 2
y2 cos x + z 3 2y sin x − 4 3xz 2 + 2 dg
⇒ =2
dz
On integration, we get 11.2.2 Definition

Any integral which is to be evaluated over a surface is
g (z) = 2z + c (8) called a surface integral. It is a natural generalisation
of the double integral over plane areas.
where c is an arbitrary constant. Let F̄(P) = F̄(r̄) be a continuous vector point
function and suppose S is a two-sided surface.
∴ φ x, y, z = y2 sin x + xz 3 − 4y + 2z + c. (9) Divide S into a finite number of sub-surfaces
δS1 , δS2 , · · · , δSn . Let Pk be any point in δSk and
(iii) Work done: n̄k be the unit vector at Pk in the direction of the
Let P = (0, 1, −1) and Q = (π/2, −1, 2) outward-drawn normal to the surface at Pk .
Work done in moving the object from P to Q is

F̄ · d r̄ = φ Q − φ (P)
C
π π
= (−1)2 sin + · 23 − 4 (−1)
2 2
+2 (2) + c] − 12 sin 0 + 0 (−1)3
−4 · 1 + 2 (−1) + c]
= 15 + 4π.
Figure 11.11
11.2 Surface Integrals Let the surface element surrounding the point Pk
be δS̄k which can be regarded as a vector, its magni-
11.2.1 Introduction tude being the area and its direction being outward
to the element.
A surface S is a geometric figure consisting of those $
points whose coordinates satisfy any equation such Consider the sum F̄(r̄k )δS̄k where the summa-
as z = f (x, y) or F(x, y, z) = 0 or parametric tion extends over all the sub-surfaces. The limit of
equations this sum if it exists, as the number n of the sub-
x = x(u, v), y = y(u, v), z = z(u, v) surfaces tends to infinity and at the same time max
with condition of continuity and non-vanishing of a |δs̄k | tends to 0, is called the normal surface integral
Jacobian imposed. of F̄(r̄) over
S and is denoted by
A smooth surface is one which has a tangent plane F̄ · d S̄ or F̄ · n̄dS
at each point P and the direction of the normal is a S S
continuous function of the point of tangency. where n̄ is the outward normal
at P to S. Other types
A surface S is two-sided or orientable if the posi- of surface integrals are F̄ × d S̄ and φd S̄, both of
tive direction of the normal at any point P of S can S
which are vectors.
be continued in a unique and continuous way to the
entire surface.
We consider here, surfaces which are orientable
(unlike a Mobius strip which is non-orientable) and 11.2.3 Flux Across a Surface
smooth or piece-wise smooth, i.e., it can be divided If F̄ represents the velocity of a fluid particle in a
into finitely many sub-surfaces each of which is domain D, then the total outward flux of F̄ across a
smooth in D. closed surface S inD is given by the surface integral
A sphere or ellipsoid is a smooth surface whereas F̄ · d S̄.
a cube or a parallelopiped is a piece-wise smooth
s
surface.
11.2.4 Solenoidal Vector Point n̄ at P(r̄) on S makes with the z-axis, then
Function
cos γdS = dxdy
When the flux of F̄ across every closed surface S in dxdy dxdy
a domain D vanishes then F̄ is said to be a solenoidal ⇒ dS = =
cos γ |n̄ · k̄|
vector point function in D.
dxdy
It may be noted that F̄ could be taken as any other Hence, F̄ · n̄dS = F̄ · n̄ .
physical quantity e.g. gravitational force or electric R |n̄ · k̄|
S
force or magnetic force.
Example 11.16
Note
It is convenient to express surface integrals as double Evaluate F̄ · n̄dS where F̄ = z ī+x j̄−3y2 z k̄ and
S
integrals taken over the orthogonal projection of the S is the surface of the circular cylinder x2 + y2 = 16
surface S on one of the coordinate planes. But this included between the planes z = 0 and z = 5 in the
is possible only if any line perpendicular to the coor- first octant.
dinate plane chosen meets the surface S at not more Solution Equation of the cylinder is
than one point. If S does not satisfy this condition, we
assume that it can be sub-divided into sub-surfaces x2 + y2 = 16 (1)
which satisfy this condition.
The outward-drawn normal to the curved surface
of the cylinder

2 ∂ ∂ 2
∇ x + y = ī + j̄
2
x + y2 = 2xī + 2yj̄
∂x ∂y
ds Unit outward normal to the curved surface of the

cylinder
2xī + 2yj̄ xī + yj̄ xī + yj̄

= =
(2x)2 + (2y)2 x2 + y2 4
∵ x2 + y2 = 16
Figure 11.12
Suppose that a line perpendicular to the xy-plane

meets S at not more than one point. Then S can be
represented by S : z = h(x, y)
Let R be the orthogonal projection of S on the xy-
plane. If γ is the acute angle which the unit normal
Figure 11.13
If R be the projection of S on the yz-plane (x = 0) π/2 1

=2 (r cos θ)(r sin θ)rdθdr
dydz
F̄ · n̄dS = F̄ · n̄ a=0 r=0
S R |n̄ · ī| (changing to polar coordinates)
where R is OBB O enclosed by y = 0 and y = 4; π/4 4 1
r
z = 0 and z = 5. = 3·2 sin θ cos θdθ
4 0
0
x y x(y + z)
F̄ · n̄ = z, x, −3y2 z · , , 0 = 3 2 π/4 3 π 3
4 4 4 = sin θ 0 = · sin2 = .
x y x 4 4 4 8
n̄ · ī = , , 0 · (1, 0, 0) =
4 4 4 Example 11.18
Evaluate F̄ ·n̄dS where F̄ = (x+y2 )ī−2xj̄+2yz k̄

dydz S
Hence F̄ · n̄dS = F̄ · n̄ and S is the surface of the plane 2x + y + 2z = 6 in
|n̄ · ī|
S
R the first octant.
1 x(y + z)
= dydz
R 4 (x/4) Solution
The
normal to the surface S is given by
5 4 ∇ 2x + y + 2z = 2ī + j̄ + k̄. The unit normal n̄ is
= (y + z) dydz given by
z=0 y=0
2ī + j̄ + 2k̄
5 4 n̄ =
1 2 |2ī + j̄ + 2k̄|
= y + yz dz 1
2 0 = (2ī + j̄ + 2k̄) ⇒ n̄ · k̄ = 2/3
0
3
5 2 1 2
= (8 + 4z) dz Also, F̄ · n̄ = x + y , −2x, 2yz ·
2
, ,
3 3 3
0 2 2 4 2 4
5 = (x + y2 ) − x + yz = y2 + yz
= 8z + 2z 2 0 = 40 + 50 = 90. 3 3 3 3 3
2 2 4 6 − 2x − y 4y
= y + y = (3 − x)
Example 11.17 3 3 2 3
Evaluate F̄ · n̄dS where F̄ = yz ī+zxj̄+xyk̄ and S Since on the plane S, z = 12 (6 − 2x − y)
S
is that part of the surface of the sphere x2 +y2 +z 2 = 1
dxdy
which lies in the first octant. Hence F̄ · n̄ds = F̄ · n̄
S S |n̄k̄|

Solution Unit vector normal to S is 4 3
= y(3 − x) dxdy
r̄ = xī + yj̄ + zr̄ R 3 2
3 6−2x

dxdy = 2y(3 − x)dydx
I= F̄ · n̄dS F̄ · n̄ 0 0
S R |n̄ · k̄| 3 y2 6−2x
where R is the projection of S on the xy-plane = 2(3 − x) dx
0 2 0
3
F̄ · n̄ = (yz, zx, xy) · (x, y, z) = 3xyz = (3 − x)3 dx
n̄ · k̄ = (x, y, z) · (0, 0, 1) = z 0
(3 − x)4 3
I = 3xyz
dxdy
= 3xydydx =4 = 81.
z 4(−1) 0
R R
Example 11.19 Solution Let

Find the angle between the normal at points P =
(1, 4, 2) and Q = (−3, −3, 3) of the surface xy = z 2 . φ1 (x, y, z) = 3x2 − 2y2 − 3z 2 + 8 (1)
φ2 (x, y, z) = ax2 + 9y2 − bz (2)
Solution Let
P lies on Eq. (2)
φ(x, y, z) = xy − z 2

∂ ∂ ∂ ⇒ a(−1)2 + 9(2)2 − b(1) = 0
∇φ = ī + j̄ + k̄ xy − z 2
∂x ∂y ∂z ⇒ b = a + 36 (3)
= (y, x, −2z) |
n̄1 = ∇φ1 P = (6x, −4y, −6z)P
Let n̄1 = ∇φ|P = (y, x, −2z)(1,4,2) = (6(−1), −4(2), −6(1)) = (−6, −8, −6)
= (4, 1, −4) n̄2 = ∇φ2 |P = (2ax, 18y, −b)P (4)
n̄2 = ∇φ|Q = (y, x, −2z)(−3,−3,3) = (2a(−1), 18(2), −b) = (−2a, 36, −6)
= (−3, −3, −6)
n̄1 · n̄2
cos θ = n̄1 · n̄2 = 0 ⇒ (−6)(−2a) + (−8)36 + (−6)(−6)
|n̄1 | |n̄2 |
4(−3)+1(−3)−4(−6) = 0 → 2a + b = 48
=
4 +1 +(−4)2 (−3)2 +(−3)2 +(−6)2
2 2 ⇒ 2a + (a + 36) = 48
1 ⇒ a = 4, b = 40.
= √ .
22 EXERCISE 11.1
Example 11.20
In what direction from the point (1, 1, −1) is the
If F̄ = 3x2 + 6y ī − 14yz j̄ + 20xz 2 k̄, evaluate
1.
directional derivative of φ(x, y, z) = x2 − 2y2 +
4z 2 a maximum? Find the value of the maximum F̄ · d r̄ along the curve C : x = t, y = t 2 ,
directional derivative. C
z = t 3 from (0, 0, 0) to (1, 1, 1).
Solution Let Ans: 5

φ(x, y, z) = x2 − 2y2 + 4z 2 , P = (1, 1, −1) 2. Evaluate F̄ · d r̄ where F̄ = x2 ī − xyj̄ from (0,
∂φ ∂φ ∂φ
∇φ = ī + j̄ + k̄ C
∂x ∂y ∂z 0) to (1, 1) along C: y2 = x.
= 2xi − 4yj̄ + 8z k̄ = (2x, −4y, 8z) Ans: −
1
12
The maximum directional derivative at (1, 1, −1)
is in the direction of 3. Evaluate F̄ · d r̄ where F̄ = z ī + xj̄ + yk̄ and
∇φ|P = (2, −4, 8) C
C is the arc of the curve r̄ = cos t ī + sin t j̄ + t k̄
Its magnitude is √ from 0 to 2a.
|∇φ|P = 22 + (−4)2 + 82 = 2 21. Ans: 3π
Example 11.21 4. Find the work done in moving a particle in a

Find the constants a and b such that the surface 3x2 − force field F̄ = 3x2 ī + (2xz − y) j̄ + z k̄ along
2y2 − 3z 2 + 8 = 0 is orthogonal to ax2 + 9y2 = bz the line joining (0, 0, 0) to (2, 1, 3).
at P = (−1, 2, 1). Ans: 16

5. Evaluate F̄ · n̄dS where F̄ = 18z ī − 12j̄ + ∴ f (x, y, z)dxdydz
S V
3yk̄ and S is the surface of the plane 2x + 3y + ⎡ 4−2x⎛ 8−4x−2y ⎞ ⎤
6z = 12 in the first octant. 2
Ans: 24 = 45 ⎣ ⎝ x2 ydz ⎠ dy⎦ dx
x=0 y=0 z=0
6. Evaluate Ā · n̄dS where Ā = xyī − x2 j̄ + ⎡ 4−2x ⎤
S
2
(x + z)k̄ where S is the portion of the plane = 45 ⎣ x2 y(8 − 4x − 2y)dy⎦ dx
2x + 2y + z = 6 included in the first octant and x=0 y=0
n̄ is the unit normal to S. 2 4−2x
Ans: −48 y2 y3
= 45 x (8 − 4x) − 2x2 ·
2
dx
2 3 y=0
x=0
11.3 Volume Integral 2
2
Consider a continuous vector point function F̄(r̄) = 45 x2 − x2 (4 − 2x)3 dx
defined in a region V bounded by a surface S. Volume 3
x=0
integral is an integral taken over the volume V of 2 2
a region bounded by a surface S. Divide V into x2
= 45 (4 − 2x)3 dx = 15 (2 − x)2 (2x)3 dx
sub-regions V1 , V2 , · · · , Vn . Let δVk be the volume 3
element of the sub-region Vk enclosing a point Pk x=0 x=0
with position vector r̄k . Then the limit of the sum a a

∵ f (x)dx = f (a − x)dx
%
∞
F̄(r̄k )δVk (11.8) 0 0
k=1 2

= 15 × 8 x3 4 − 4x + x2 dx
if it exists, as n tends to infinity in such a way that
δVk → 0, is called thevolume integral of F̄(r̄) over
x=0
2
x4 x5 1 6
V and is denoted by F̄dV or as a triple integral = 15 × 8 4 · −4· + x
4 5 6 0
V
2
4 1
F̄(x, y, z) dxdydz. = 15 × 8 x4 1 − x + x2
V 5 6 0
In components form, it is
8 42
= 15 × 8 × 16 1 − + 3
5 6
F̄ · dV = ī F1 (x, y, z)dxdydz (15 − 24 + 10)
V
V = 15 × 8 × 16 = 128.
15
+ j̄(· · · ) + k̄(· · · ) (11.9)
Example 11.23
Example 11.22
If F̄ = 2xz ī − xj̄ + y2 k̄ evaluate F̄dV where
Evaluate f (x, y, z)dV where f (x, y, z) = 45x y 2
V is the region bounded by the surface x = 0, y =
V
V
and V is the region bounded by the planes x = 0, y = 0, x = 2, y = 6, z = x2 , z = 4
0, z = 0 and 4x + 2y + z = 8.
Solution
Solution We write the limits of variation for x, y

and z. z varies from 0 to 8 − 4x − 2y. y varies from F̄dV = 2xz ī − xj̄ + y2 k̄ dxdydz
0 to 4 − 2x. x varies from 0 to 2. V
V
⎧ ⎡ ⎤ ⎫
⎪ 6 4
2 ⎨ ⎪
⎥ ⎬

11.4 Gauss’s Divergence Theorem
⎢
= ⎣ 2xz, −x, y2 z dz ⎦ dy dx (G.D.T.)
⎪
⎩ ⎪
⎭
x=0 y=0 z=x2
⎡ ⎤ 11.4.1 Transformation Between
2 6 Surface Integrals and
4
= ⎣ xz 2 , −xz, y2 z dy⎦ dx Volume Integrals
z=x2
x=0 y=0 Statement If F̄ is a vector point function having
⎡ ⎤
2 6 continuous first order partial derivatives in the region

= ⎣ 16x−x5 ,−4x+x3 , 4y2 −x2 y2 dy⎦ dx V bounded by a closed surface S, then

x=0 y=0
∇ · F̄dV = F̄ · n̄dS (11.10)
2 6 V S
4 x2 y3
= 16xy − x5 y, −4xy + x3 y, y3 − dx where n̄ is the outward-drawn unit normal vector
3 3 0
x=0 to the surface S. (i.e., the volume integral of the
2 divergence of a vector point function F̄ taken over
4 2 216
= 96x−6x ,−24x+6x , · 216−x ·
5 3
dx the volume V enclosed by a surface S is equal to the
3 3 surface integral of the normal component of F̄ taken
x=0
2 over the closed surface S.)
3 4
= 48x − x , −12x + x , 288x − 24x
2 6 2 3
2 Proof: Let the surface S be such that a line parallel
0

3 to any coordinate axis meets it at a maximum of two
= 48·4−2 ,−12·2 + ·2 , 288·2−24 · 2
6 2 4 3
2 points. Let S1 and S2 denote the lower and upper parts
= (128, −24, 384) = 128ī − 24j̄ + 384k̄. of S with equations
z = F1 (x, y) and z = F2 (x, y), respectively.
EXERCISE 11.2 Let F̄ = F1 ī + F2 j̄ + F3 k̄

∂F1 ∂F2 ∂F3
so that ∇ · F̄ = div F̄ = + +
1. If F̄ = 2x2 − 3z ī − 2xyj̄ − 4xk̄ then evaluate ∂x ∂y ∂z

∇ · F̄dV where V is bounded by the planes
V
Let α, β, γ be the angles which the outward-
x = 0, y = 0, z = 0 and 2x + 2y + z = 4 drawn unit normal vector n̄ makes with the positive
Ans: 8/3 x-, y-, z-axes, respectively. Then cos α, cos β, cos γ
are the direction cosines of n̄ so that
2. Evaluate f dV where f = 90x2 y and V is
V n̄ = cos α ī + cos β j̄ + cos γ k̄
the closed region bounded by the planes x =
Now F̄ · n̄ = F1 ī + F2 j̄ + F3 k̄ ·
0, y = 0, z = 0 and 4x + 2y + z = 8.
Ans: 256 cos α ī + cos β j̄ + cos γ k̄
= F1 cos α + F2 cos β + F3 cos γ
3. If F̄ = 2x2 − 3z ī − 2xyj̄ − 4xk̄, evalu-
Hence, Eq. (11.10) can be expressed in cartesian
ate ∇ × F̄dV where V is the closed region
coordinates, as
V

bounded by the planes x = 0, y = 0, z = 0 ∂F1 ∂F2 ∂F3
and 2x + 2y + z = 4. + + dxdydz
∂x ∂y ∂z
8 V
Ans: j̄ − k̄
3 = (F1 cos α+F2 cos β+F3 cos γ) dS (11.11)
S
z Similarly,
k
r2 n2
dS2 ∂F1
dxdydz = F1 ī · n̄ dS
V ∂x
S2: z = f2 (x, y)
S
= F1 cos α dS (11.16)
S
S1: z = f1 (x, y) ∂F2
dS1 dxdydz = F2 j̄ · n̄ dS
r1 V ∂y
n1
S
= F2 cos β dS (11.17)
S
o y
Adding Eqs. (11.15), (11.16) and (11.17), we get
Eq. (11.11).
x R Note
dxdy
Gauss’s divergence theorem is also known as Green’s
theorem in space.
Figure 11.14 Gauss’s divergence theorem
If F̄ denotes the fluid velocity then Gauss diver-
If R is the projection of S on the xy-plane, then gence theorem has the physical interpretation that the
volume of fluid flowing out of a closed surface S =
⎛ ⎞
f2 (x,y) Volume of fluid supplied from inside V in unit time.

∂F3 ⎜ ∂F3 ⎟
dxdydz = ⎝ dz⎠dxdy
V ∂z R ∂z 11.4.2 Some Deductions from
z=f1 (x,y)
Gauss’s Divergence Theorem

= F3 x, y, f2 x, y dxdy Green’s first identity
R (11.12)
Prove that

− F3 x, y, f1 x, y dxdy
∂g 2
R f dS = f ∇ g + ∇f · ∇g dV .
S ∂n V
Now the outward-drawn normal n̄2 to S2 makes an
acute angle γ2 with k̄ so that Solution By Gauss’s divergence theorem

dxdy = cos γ2 dS2 = k̄ · n̄2 dS2 (11.13) F̄ · n̄ds = ∇ · F̄ dV
S V

while the outward-drawn normal n̄1 to S1 makes an Take F̄ = f ∇g ⇒ ∇ · F̄ = ∇ · f ∇g
obtuse angle γ1 with k̄ so that = f ∇ 2 g + ∇f · ∇g
dxdy = − cos γ1 dS1 = −k̄ · n̄1 dS1 (11.14) ∂g
F̄ · n̄ = f ∇g · n̄ = f
∂n̄
Now Eq. (11.12) becomes ∂g
∴ F̄ · n̄ dS = f dS
∂n̄
∂F3
S
S
dxdydz = F3 k̄ · n̄2 dS1 = ∇ · F̄ dV
V ∂z S2
V
+ F3 k̄ · n̄1 dS1 2
≡ f∇ g +∇f ·∇g dV
S1
V
∂g
= F3 k̄ · n̄ dS = F3 cos γ dS (11.15) ∵ dS = ∇g · n̄ dS = ∇g · d S̄
S R ∂x̄
Green’s first identity can also be written as Solution By Gauss’ divergence theorem

f ∇g · d S̄ ≡ f ∇ 2 g + ∇f · ∇g dV . ∇ · F̄dV = F̄ · n̄ dS (1)
S V V S
Green’s second identity

Let F̄ = xī so that ∇ · F̄ = 1; we get from Eq. (1)

Prove that Volume V = 1 · dV = ī · n̄ x dS
V S
2 ∂g ∂f
f ∇ g −g∇ 2 f dV ≡ f −g dS. = x dydz (2)
V S ∂n ∂n
S

Solution From Green’s first identity, we have Similarly, V = j̄· n̄ y dS = y dzdx (3)
S
∂g S

f dS= f ∇ 2 g +∇f ·∇g dV (11.18)
S ∂x V V= k̄ · n̄ z dS = z dxdy (4)
Interchanging the roles of f and g in Eq. (11.17) S S
Adding Eqs. (2), (3) and (4) and dividing by 3

∂f
≡ f dS = g ∇ 2 f + ∇g · ∇f dV (11.19)
S ∂x 1
V = x dydz + y dzdx + z dxdy .
3 S
Subtracting Eq. (11.19) from Eq. (11.18)
Example 11.25
∂g ∂f
f −g dS For any closed surface S, prove that
∂x ∂x
S

curl F̄ · n̄ dS = 0. (1)
= f ∇ 2 g − g ∇ 2 f dV (11.20)
S
V
Green’s second identity can also be written as Solution By Gauss’s divergence theorem

2 curl F̄ · n̄ dS = div curl F̄ dV = 0
f ∇g −g ∇f ·d S̄ = f ∇ g −g ∇ 2 f dV S V
S V
where V is the volume of the region enclosed by S.

since f ∇g · n̄ − g ∇f · n̄ dS
Example 11.26
S
Evaluate r̄ · n̄ dS. where S is a closed surface
= f ∇g − g ∇f · n̄d S̄
S
S enclosing a region of volume V .

= f ∇g − g ∇f · d S̄
S Solution By Gauss’s divergence theorem

If f and g are harmonic functions, then f , g satisfy
r̄ · n̄ dS = (∇ · r̄) dV
Laplace’s equation so that ∇ 2 f = ∇ 2 g = 0. S
V
Now Eq. (11.20) yields
=3 dV = 3V (∵ ∇ · r̄ = 3) .

∂g ∂f V
f −g dS = 0.
S ∂x ∂x Example 11.27
If F̄ = ax ī+
by j̄ + cz k̄; a, b, c being constants
Example 11.24 4
Find the volume of a region bounded by a surface S. show that F̄ · n̄ dS = π (a + b + c) where S
S 3
6
%
is the surface of a unit sphere.
F̄ · n̄ dS = F̄ · n̄dS (i = 1, 2, 3, · · · 6)
i=1 S
Solution By Gauss’s divergence theorem S i
where (S1 , S2 ); (S3 , S4 ); (S5 , S6 ) are faces normal to

¯f · n̄ dS = ∇ · F̄ dV the z-, y- and x-axes, respectively.

% S V

∂ Now F̄ · n̄ dS = F̄ · −k̄ dS (2)
(ax)dV = (a + b + c) V
V ∂x S1 S1
4π b a
= (a + b + c) .
3 =− 0 − xy dxdy
Example 11.28 0 0
If n̄ is the outward-drawn unit normal to any closed 2 2
ab
surface S, then show that =
4
div n̄ dV = S F̄ · n̄ dS = F̄ · k̄ dS (3)
V
. S2 S2
Solution By Gauss’ divergence theorem b a

= c2 − xy dxdy (4)
div n̄ dV = n̄ · n̄dS
0 0
V
S
a 2 b2
= dS = S. =
4
S

Example 11.29 b2 c 2
F̄ · n̄ dS = (5)
Verify Gauss’s
divergence
theorem for F̄ = 4
x2 − yz ī + y2 − zx j̄ + z 2 − xy k̄ taken over the S3

rectangular parallelopiped 0 x a, 0 y b2 c 2
F̄ · n̄ dS = a2 bc − (6)
b, 0 z c. 4
S4

Solution We have c2 a2
∂ 2 ∂ 2 ∂ 2 F̄ · n̄ dS = (7)
∇ · F̄ = x −yz + y −zx + z − xy 4
∂x ∂y ∂z S5

= 2 x+y+z c2 a2
F̄ · n̄ dS = ab2 c − (8)
4
c b a S6

∇ · F̄ dV = 2 x + y + z dxdydz ∴ From Eqs. (2) and (7), we have
V
0 0 0

⎡ ⎤ F̄ · n̄ dS = abc(a + b + c) (9)
a b 2
a
=2 ⎣ + ay + az dy⎦ dz S
2
0 0 Equality of (1) and (8) provides verification of
c Gauss’s divergence theorem.
a2 ab2
=2 b+ + abz dz
2 2 Example 11.30
0
2
ab ab2 c2 Evaluate x2 dydz + y2 dzdx + 2z(xy − x − y) dxdy
=2 c+ c + ab S
2 2 2 where S is the surface of the cube 0 ≤ x ≤ 1,
= abc(a + b + c) (1) 0 ≤ y ≤ 1, 0 ≤ z ≤ 1.
a
Solution By Gauss’s divergence theorem, the 20 3/2
given surface integral = a2 − y 2 z dy
3
y=0
/ ,
∂ 2 ∂ 2 ∂ a
= x + y + 2z xy−x−y dV 20
V ∂x ∂y ∂z = b(a2 − y2 )3/2 dy
3
1 1 1 y=0

= 2x + 2y + 2 xy − x − y dx dy dz
Put y = a sin t, dy = a cos tdt.
z=0 y=0 x=0
Limits y = 0 ⇒ t = 0; y = a ⇒ t = π/2.
1 1 1 1 1 1
x2
= 2xydxdydz = 2 y dydz 2
π
2
π
2 x=0 20 20 4
z=0 y=0 x=0 z=0 y=0 = ·b a4 cos3 t cos tdt = ab cos4 tdt
1 1 1 1 3 3
0 0
y2 1 1 1
= ydydz = dz = [z]0 = . 20 4 3 1 π 5πa4b
2 0 2 2 = a b· · · = ·
z=0 y=0 z=0 3 4 2 2 4

Evaluate x3 dy dz + x2 y dz dx + x2 z dx dy where Apply
Gauss’s theorem to evaluate
S 3
S is the closed surface bounded by the planes z = 0, x − yz dydz − 2x2 ydzdx + zdxdy
S
z = b and the cylinder x2 + y2 = a2 . over the surface of the cube bounded by the coordi-
Solution By Gauss’s divergence theorem, the nate planes and the planes x = y = z = a.
required surface integral
Solution
By Gauss’s divergence theorem
F1 dydz + F2 dzdx + F3 dxdy
∂ 3 ∂ 2 ∂ 2
= x + xy + x z dV S
V ∂x ∂y ∂z ∂F1 ∂F2 ∂F3
√ = dxdydz
b a a −y
2 2 V ∂x ∂y ∂z
2 Here F1 = x3 − yz, F2 = −2x2 y, F3 = z
= 3x + x2 + x2 dxdydz
z=0 y=−a
√ ∂F1 ∂F2 ∂F3
x=− a2 −y2 ∇ · F̄ = + + = x2 + 1
√ ∂x ∂y ∂z
a −y
2 2
b a a a 3 a 3
x a
= 4×5 x2 dxdydz RHS = + x dydz = + a a2
3 0 3
z=0 y=0 x=0 z=0 y=0
√
b a a2 −y2

x3 a3
= 20 dydz ∴ (x3−yz)dydz−2x3 ydzdx+zdxdy = +a a2 .
3 3
z=0 z=0 x=0 S
b a EXERCISE 11.3
20
= (a2 − y2 )3/2 dydz
3
z=0 y=0 1. Evaluate S F̄ · n̄ dS whereF̄ = 2x2 yī − y2 j̄ +
⎡ ⎤
a b 4xz 2 k̄ where S is the surface x2 + y2 = 9, x = 2
20 ⎣
= (a2 − y2 )3/2 dz ⎦ dy in the first quadrant. [JNTU 2000]
3
y=0 z=0 Ans: 180
2.
Evaluate ∂N
in a region R of the xy-plane bounded by a simple
S
(x + z)dzdx + (x + y)dxdy+(y + z)dydz ∂x
closed curve C then
where S is the surface of the sphere
x2 + y2 + z 2 = 4. ∂N ∂M
64 Mdx + Ndy = − dxdy (11.21)
Ans: π R ∂x ∂y
C
3
where C is traversed in the counter-clockwise direc-
3. Using
divergence theorem evaluate tion.
S
x dydz + y dzdx + z dxdy where S is the
surface x2 + y2 + z 2 = 1. Proof: Assume that the region R is such that the
Ans: 4π vertical or horizontal line meets its boundary in at
most two points (or that R can be sub-divided into
sub-regions satisfying this condition).
4. Evaluate S x dydz + y dzdx + z dxdy where S is
the surface bounded by the coordinate planes and Suppose R is bounded between the vertical lines
the planes x = y = z = a. x = a and x = b (b > a) and the arcs APB and
a3 BQA defined by y = f (x) and y = g(x) respectively
Ans: [f (x) < g(x)]
3
⎡ ⎤
2 b g(x)
5. Evaluate z dxdydz taken over the volume ∂M ∂M
V Now dxdy = ⎣ dy⎦dx
bounded by x2 + y2 = a2 , x2 + y2 = z 2 and z = 0. R ∂y ∂y
[JNTU 1997] a y=f (x)
Ans: πa8 /12 b

g(x)

= M x, y dx
6. Evaluate F̄ · d S̄ where F̄ = xī + yj̄ + z k̄ and S
a f (x)
is the triangle (1,0,0) (0,1,0) and (0,0,1).
Ans: 1/12 b

= M x, g (x) − M x, f (x) dx
e log
y ex a
7. Evaluate log zdz.
a b

y=1 x=1 z=1

Ans: e2 − 8e + 13 /4 =− M x, g (x) dx − M x, f (x) dx
b a
a x x+y
⎡ ⎤

8. Evaluate ex+y+z dzdydx.
= −⎣ M x, f dx + M x, g dx⎦
x=0 y=0 z=0

Ans: e4a − 6e2a + 8ea + 3 /8 APB BQA

=− Mdx (11.22)
9. Find S F̄ · n̄ dS where F̄ = (2x + 3z)ī − (2x +
y)j̄ + (y2 + 2z)k̄ and S is the surface of the sphere C
with centre (3, –1, 2) and radius 3 units. Similarly,we

can show that

Ans: 108π ∂N
dxdy = Ndy (11.23)
R ∂x
C
11.4.3 Green’s Theorem in the From Eqs. (11.22) and (11.23), we obtain Eq.
Plane (11.21).
If M (x, y) and N (x, y) be continuous functions of x This theorem is useful for changing a line integral
∂M around a simple closed curve C into a double integral
and y having continuous partial derivatives and over the region R enclosed by C.
∂y
Example 11.34
Verify Green’s theorem for y2 dx − x2 dy where C
C
is the boundary of the triangle whose vertices are
(1, 0), (0, 1), (−1, 0). [JNTU 1998]
Solution For verification of Green’s theorem, we

have to show that

∂N ∂M
Mdx + Ndy = − dxdy (1)
R ∂x ∂y
Figure 11.15 C
Example 11.33 Here M = y2 , N = −x2 and the closed curve C is

Show that the area of the region bounded
bya simple composed of the line segments AB, BC and CA.
closed curve C is given by 12 xdy − ydx . Hence Along AB
C
y = 0, dy = 0 x varies from –1 to 1
x2 y2
show that the area of the ellipse + = 1 is πab.
a2 b2 y2 dx − x2 dy = 0.
AB
Solution By Green’s theorem, we have

∂N ∂M
Mdx + Ndy = − dxdy
R ∂x ∂y
C
Take M = −y/2, N = x/2

∂M ∂N
= −1/2, = 1/2
∂y ∂x

1
Now xdy−ydx = −y/2 dx+(x/2)dy Figure 11.16
2
C C
Along BC
1 1 B = (1, 0), C = (0, 1);
= − − dxdy = A
2 2 Equation of BC is y = 1 - x; x varies from 1 to 0
R
where A denotes the area of the region R bounded dy = −dx

by C. 0
x2 y2 y dx − x dy = (1 − x)2 dx − x2 (−dx)
2 2
Equation of ellipse: + =1
a2 b2 BC 1
Parametric equations: x = a cos t, y = b sin t 0

1 = 1 − 2x + 2x2 dx
Area of the ellipse: xdy − ydx
2 1
C 0
2
2π = x − x2 + x3
1 3
= (a cos t · b cos t + b sin t · a sin t)dt 1
2 2 2
0 = − 1−1+ =− ·
2π 3 3
ab ab
= dt = · 2π = πab. Along CA
2 2 C = (0, 1), A = (−1, 0);
0
Equation of CA is y = 1 + x; x varies from 0

to −1
dy = dx
−1

y dx − x dy =
2 2
1 + 2x + x2 − x2 dx
CA 0
1
= x + x2 0 = −1 + 1 = 0
Figure 11.17
LHS = y2 dx − x2 dy
∂M ∂N
C = 1, = − sin x
∂y ∂x
2 2
= 0+ − +0=− The closed curve C composed of the line segments
3 3

∂N ∂M y=0 x = π/2 and y = 2x/π
RHS = − dxdy
R ∂x ∂y
and the region R is the triangular region bounded by
x varies from y − 1 to 1 − y and y varies from 0 to 1
⎡ 1−y ⎤ these lines.
1 2x
x varies from 0 to π/2 and y varies from 0 to
= ⎣ −2x − 2y dx⎦dy π

y=0 x=y−1
y − sin x dx + cos x dy
1
1−y C
= −x2 − 2xy2 1+y
dy
y=0 = (− sin x − 1) dxdy
R
1 ⎡ ⎤
π/2 2x/π
= [0 − 2y(1 − y)]dy = ⎣ (− sin x − 1) dy⎦ dx
y=0
x=0 y=0
4 3 1 π/2
= −2y + y
2
3 = (− sin x − 1) [y]2x/π
0 y=0 dx
4 2
= −2 + = − x=0
3 3 π/2
2x
LHS = RHS and Green’s theorem has been verified. = (− sin x − 1) .dx
π
0
Example 11.35 π/2
Apply Green’s theorem to evaluate (y − sin x) dx + 2
C =− (x + x sin x) dx
cos x dy where C is the plane triangle enclosed by π
0
the lines y = 0, x = π/2 and y = 2x/π. π/2
2 x2
[JNTU 1993, 1995, 1996S, 2003S, 2003] =− + x (− cos x) − 1 · (− sin x)
π 2 0

∂N ∂M 2 1 π2 π π π
Mdx + Ndy = − dxdy (1) =− − cos + sin
∂x ∂y π 2 4 2 2 2
R
C π 2 π 2
= − +0− =− + ·
Here M = y − sin x, N = cos x 4 π 4 π
Example 11.36 Along AQO

Verify Green’s theorem in the plane for √ 1
y = x, dy = √ dx; x varies from 1 to 0
2 2 x
3x − 8y2 dx + 4y − 6xy dy where C
C √ Mdx+Ndy
encloses the region bounded by y = x and y = x2 . AQO
2
= 3x −8y2 dx+ 4y−6xy dy
AQO
∂N ∂M
Mdx + Ndy = − dxdy (1) 0
R ∂x ∂y √ √ dx
C = 3x2 −8x dx+ 4 x−6x x √
2 x
1
0

= 3x2 −8x+2−3x dx
1
0
x3 x2
= 3 −11 +2x
3 2
1
11 5
=− 1− +2 =
2 2

5
LHS = Mdx + Ndy = −1 + = 3/2
2
C

∂N ∂M
Figure 11.18 RHS = − dxdy
R ∂x ∂y
Here M = 3x − 8y ,
2 2
N = 4y − 6xy ⎡ √ ⎤
1 x
∂M ∂N ⎢ ⎥
= −16y, = −6y = ⎣ −6y − −16y dy⎦ dx
∂y ∂x x=0 y=x2
The closed curve C composed of the arcs OPA and 1 1
y=x2
AQO is given by
= 5y 2
= 5 x − x4 dx
OPA: y = x2 , x varies from 0 to 1 √
x
√ x=0

0
AQO: y = x, x varies from 1 to 0. 5 2 x5
1
5
Along OPA = x −5 = · 1 − 1 = 3/2
2 5 0 2
y = x dy = 2xdx and x varies from 0 to 1
2 LHS = RHS
Green’s theorem has been verified.
2
Mdx+Ndy = 3x − 8y2 dx+ 4y − 6xy dy
OPA OPA
11.4.4 Green’s Theorem in
1 Vectorial Form
2
= 3x −8x4 dx+ 4x2 −6x3 2xdx If F̄ = M x, y ī + N x, y j̄ and r̄ = xī + yj̄, then
Green’s theorem,
in vectorial
form, is
x=0
1
8 x4 x5 F̄ · d r̄ = curl F̄ · k̄dS
= x − x5 + 8 − 12
3
5 4 5 C R
0
8 12 = ∇ × F̄ · k̄ dS
= 1− +2− = −1.
5 5 R
−a
where the symbols have their usual meaning. 1 3 2
= x + b2 x = − a3 − 2b2 a.
3 a 3
Example 11.37
Verify Green’s theorem for the scalar line integral of Along DA
F̄ = x2 + y2 ī − 2xyj̄ around the rectangle formed x = −a, dx = 0; y varies from b to 0
by the lines x = ±a, y = b. o
F̄ · d r̄ = 2ay dy
Solution The rectangle ABCD is composed of
DA b
the line segments AB, BC, CD, DA. o
⎛ ⎞ = ay2 = −ab2
b
LHS = F̄ · d r̄ = ⎝ + + + ⎠ F̄ · d r̄ Adding these we get

C AB BC CD DA 2
LHS = F̄ · d r̄ = a3 + (−ab)2
3
C

2 3
+ − a − 2ab + −ab2 2
3
= −4ab2
ī j̄ k̄
∂ ∂ ∂
Now, curl F̄ = = −4y k̄
∂x ∂y ∂z
x2 + y2 −2xy 0

RHS = curlF̄ · k̄dS
Figure 11.19 R
a b
Along AB = −4ydydx
y = 0, dy = 0; x varies from −a to a x=−a y=0
F̄ = x ī, d r̄ = dxī
2 a
b
a = −2y2 0 dx = −2b2 [x]a−a
1 a 2
F̄ · d r̄ = x2 dx = x3 = a3 . −a
3 −a 3
AB −a = −4ab2
Along BC LHS = RHS
x = a, dx = 0; y varies from 0 to b
Green’s theorem has been verified.
F̄ = −2ayj̄; d r̄ = dyj̄
EXERCISE 11.4
b
b
F̄ · d r̄ = −2ay dy = −ay2 o = −ab2 . 1. Verify Green’s theorem in the plane for the fol-
BC o lowing cases.
Along CD
Integral Closed curve C enclosing the
y = b, dy = 0; x varies from a to − a region R bounded by

F̄ = x2 +b2 ī−2bxj̄; d r̄ = dx ī+dy j̄
(a) (3x − 8y2 ) dx x = 0, y = 0 and x + y = 1
F̄ · d r̄ = x2 + b2 dx − 2bx dy C +(4y − 6xy) dy [JNTU 2003, 2005]
−a
2
F̄ · d r̄ = x + b2 dx + 0 (b) (xy + y2 ) dx + x2 dy y = x and y = x2
C
CD x=a

(c) (3x2 + 2y) dx the parallelogram with vertices where n̄ is the unit normal at any point of S drawn
C −(x + 3 cos y) dy (0, 0) (2, 0) (3, 1) (1, 1) in the path in which a right-handed screw would
advance when rotated in the sense of description of
(d) (2xy − x2 ) dx y = x2 , y2 = x C.
C +(x2 + y2 ) dy [JNTU 1999S]
Proof: Let S be represented by
Scalar line integral of the rectangle bounded by z = f (x, y) or x = g(y, z) or y = h(z, x)
(e) f¯ = (x2 − y2 ) ī + 2xy j̄ x = 0, a; y = 0, b which are continuous and differentiable functions.

Suppose that the projections of S on the xy-, yz- and

zx-planes are regions bounded by simple closed
2. Evaluate x2 − 2xy dx+ x2 y + 3 dy around
curves. If F̄ = F1 ī + F2 j̄ + F3 k̄ then the result
C of Stokes’ theorem (11.24) can be written as
the boundary of the region defined by y2 = 8x
and x = 2; (a) directly and (b) by using
Green’s theorem.
Ans: 128/5
3. A vector field is given by F̄ = sin yī + x(1 +

cos y)j̄. Evaluate the line integral over the circle
x2 + y2 = a2 , z = 0; (a) directly and (b) by
using Green’s theorem. [JNTU 1996]
Ans: πa2
4. A vector field is given by F̄ = (x − 2y)ī + xj̄.

Evaluate the line integral over the circle x2 +
y2 = 1, z = 0; (a) directly and (b) by using
Green’s theorem.
Ans: 3π
5. If C is the rectangle with vertices

(0,
0), (π, 0), (π, π/2), and (0, π/2), evaluate
Figure 11.20
e−x sin y dx + cos y dy using Green’s

C
∇ × F̄ · n̄dS
theorem. S
D

Ans: 2(1 − e−π ) = ∇ × F1 ī+F2 j̄+F3 k̄ · n̄ dS = F̄ ·d r̄
S
C
D
11.5 Stokes’ Theorem
= F1 ī + F2 j̄ + F3 k̄ · dx ī+dy j̄+dz k̄
11.5.1 Transformation Between C
Line and Surface Integral D
= F1 dx + F2 dy + F3 dz (11.25)
Statement Let F̄ be a vector point function hav-
C
ing continuous first partial derivatives in a domain
in space containing an open two-sided surface S We show that
D
bounded by a simple closed curve
D C then ∇ × F1 ī · n̄ dS = F1 dx (11.26)

∇ × F̄ · n̄ dS = F̄ · d r̄ (11.24) S
C
S
C
ī j̄ k̄ Applying Green’s theorem in the plane

∂ ∂ ∂ D D
Consider ∇ × F1 ī = ∂φ
∂x ∂y ∂z − dxdy = φ dx = F1 dx (11.33)
F1 0 0 R ∂y
C C
∂F1 ∂F1
= j̄ − k̄ since at each point (x, y) of C the value of φ is same
∂z ∂y as the value of F1 at each point (x, y, z) of C and dx
so that is same for both C and C . From Eqs. (11.32) and
∂F1 ∂F1 (11.33)
∇ × F1 ī · n̄dS = j̄· n̄− k̄ · n̄ dS (11.27)
∂z ∂y D

∇ × F1 ī · n̄ dS = F1 dx (11. 34)
If we take S to be given by z = f (x, y), the position S
C
vector of any point P on S is
Similarly, projecting on the yz- and zx-planes
r̄ = xī + yj̄ + z k̄ = xī + yj̄ + f (x, y)k̄ D

∇ × F2 j̄ · n̄ dS = F2 dy (11. 35)
∂r̄ ∂f S
so that = 0 · ī + j̄ + k̄ (11.28)
C
D
∂y ∂y
∇ × F3 k̄ · n̄ dS = F3 dz (11. 36)
∂r̄ S
C
Since is tangent to the surface S and n̄ is normal
∂y
∂r̄ Adding Eqs. (11.34), (11.35) and (11.36), we get
to it, we have n̄ · =0 Eq. (11.24).
∂y
∴ Taking dot product of Eq. (11.27) with n̄, we
Note
have
Stokes’ theorem in rectangular cartesian form
∂r̄ ∂f
0 = n̄ · = n̄ · j̄ + n̄ · k̄ ∂F2 ∂F1 ∂F3 ∂F2
∂y ∂y − cos γ + − cos α
∂x ∂y ∂y ∂x
∂f ∂z S

or n̄ · j̄ = − n̄ · k̄ = − n̄ · k̄ (11.29) ∂F1 ∂F3
∂y ∂y + − cos β ds
∂z ∂x
D
Substituting this in Eq. (11.26), we get
= F1 dx + F2 dy + F3 dz
∂F1 ∂z ∂F1
∇ ×F1 ī · n̄ dS = − + n̄· k̄ dS (11.30) C
∂z ∂y ∂y
where n̄ = cos α ī + cos β j̄ + cos γ k̄; (α, β, γ) being
Now on S F1 (x, y, z) = F1 (x, y, f (x, y)) = φ(x, y) the angles made by n̄ with the coordinate axes.
∂F1 ∂F1 ∂z ∂φ
so that + = (11.31) Example D11.38
∂y ∂z ∂y ∂y
Prove that r̄ · d r̄ = 0.
Using Eq. (11.31) in Eq. (11.30) C

∂φ
∇ × F1 ī · n̄ dS = − n̄ · k̄ dS Solution
D By Stokes’
theorem, we have
∂y
S
S F̄ · d r̄ = ∇ × F̄ · n̄ dS
∂φ
=− dxdy (11.32) S
R ∂y
C
Take F̄ = r̄
D
where R is the projection of S on the xy-plane and
r̄.d r̄ = (∇ × r̄) · n̄ dS = 0 ∴ ∇ × r̄ = 0̄ .
n̄ · k̄ dS = dxdy C
S
⎛ ⎞
ExampleD 11.39
⎜ ⎟
Show that f ∇f d r̄ = 0, C being a closed curve. = ⎝ + + + ⎠ F̄ · d r̄
C C1 C2 C3 C4
Solution
D By
Stokes’ theorem, we have where C1 , C2 , C3 , C4 are the line segments OA, AB,

F̄ · d r̄ = ∇ × F̄ · n̄ dS BC, CO, respectively.
S
C

Take F̄ = f ∇f · ∇ × F̄ = ∇ × f ∇f
= ∇f ×∇f +∇ ×∇f = 0+0 = 0
D

∴ f ∇f ·d r̄ = ∇ × f ∇f · n̄ dS = 0.
S
C
Example 11.40
Figure 11.21
Applying Stokes’ theorem show that ∇ × F̄ · Along C1
S
d s̄ = 0 where F̄ is any vector over the sphere x2 +y2 + y= 0, dy = 0; x varies
a from 0 to a
z 2 = 1. [JNTU 1998] 2 a3
F̄ · d r̄ = x − 0 dx + 0 = ·
3
C1 x=0
Solution Let S1 and S2 be the upper and lower
portions of the sphere x2 + y2 + z 2 = 1 cut by any Along C2
plane. Let C be the boundary for S1 and S2 . x = a, dx = 0;b y varies from 0 to b

2
F̄ · d r̄ = a − y2 · 0 + 2ay dy = ab2 .
∇ × F̄ · n̄dS
S C2 y=0

= ∇ × F̄ · n̄ dS + ∇ × F̄ · n̄ dS Along C3
S1 S2 y = b, dy = 0, x varies from a to 0
0
2
F̄ · d r̄ = x − b2 dx + 2bx · 0

= ∇ × F̄ · n̄ dS − ∇ × F̄ · n̄ dS, C3 x=a
S1 S2
x3 a3
reversing the direction of S2 to that of S1 = − bx2 0
a =− + b2 a.
3 3
Along C4
= F̄ · d r̄ − F̄ · d r̄ = 0. x = 0, dx = 0; 0 y varies from b to 0

C C
F̄ · d r̄ = 0 − y2 · 0 + 2 · 0 · ydy = 0.
Example 11.41
Verify Stokes’ theorem for the vector field F̄ = (x2 −
C b

y2 )ī + 2xyj̄ integrated around a rectangle in plane Adding these we get, LHS = F̄ · d r̄ = 2ab2
z = 0 formed by the lines x = 0, y = 0, x = a and C
y = b. [JNTU 1992, 2003S] Now consider
ī j̄ k̄
Solution We have
∂ ∂ ∂
F̄ = (x2 − y2 , 2xy), r̄ = (x, y), d r̄ = (dx, dy) ∇ × F̄ = = 4yk̄
D D ∂x ∂y ∂z
2
F̄ · d r̄ = x − y2 dx + 2xydy x2 − y2 2xy 0
C C

2π
RHS = ∇ × F̄ · n̄dS
S = (2 cos θ − sin θ) (− sin θ) dθ
⎛ ⎞
a b 0
dxdy ⎝ 2π π/2
= 4y k̄ · n̄ = 4ydy⎠dx
R k̄ · n̄ =− sin 2θdθ + 4 sin2 θdθ
x=0 y=0
a 0 0
b a
= 2y2 0 dx = 2b x 2
= 2ab 2 1 π
0 = 0+4· · =π (2)
0
2 2
ī j̄ k̄
LHS = RHS and Stokes’ theorem is verified. ∂ ∂ ∂
Now, curl F̄ = (3)
Example 11.42 ∂x ∂y ∂z
Verify Stokes’ theorem for the vector field F̄ = 2x − y −yz 2 −y2 z
(2x − y)i − yz 2 j̄ − y2 z k̄ over the upper half sur-
face of x2 + y2 + z 2 = 1 bounded by its projection = ī −2yz+2zy + j̄ (0 − 0)+k̄ (0+1)
on the xy-plane. [JNTU = k̄
1998S]
RHS = curl F̄ · n̄ dS = k̄ · n̄ dS
Solution By Stokes’ theorem, we have S

dxdy
= k̄ · n̄ = dxdy = π (4)
F̄ · d r̄ = curlF̄ · n̄ dS (1) R k̄ · n̄ R
S
C
(Area of circle x2 + y2 = 1.) From Eqs. (2) and (3)
Here F̄ = 2x − y, −yz 2 , −y2 z ; d r̄ = dx, dy, dz . LHS = RHS and Stokes’ theorem is verified.
The projection of the upper half of the sphere
x2 + y2 + z 2 = 1 in the xy-plane is the circle Example 11.43
x2 + y2 = 1, z = 0 and C is its boundary. Apply
Stokes’ theorem to evaluate
ydx + zdy + xdz where C is the curve of
C
intersection of x2 + y2 + z 2 = a2 and x + z = a.
[JNTU 1997]
Figure 11.22

LHS = F̄ · d r̄ = 2x − y dx−yz 2 dy−y2 zdz
C C Figure 11.23

Solution The intersection of the sphere
= 2x − y dx (∴ z = 0, dz = 0)
x2 + y2 + z 2 = a2 (1)
C
and the plane x+z=a (2)
Put x = cos θ, y = sin θ, dx = − sin θdθ,
dy = cos θdθ, θ varies from 0 to 2π is the circle in the plane (2) with AC as diameter.
EXERCISE 11.5
x z
+ =1
a a 1. Verify Stokes’ theorem in the following cases.
Hence OA = OC = a.
The coordinates of A = (a, 0, 0), C = (0, 0, a) Vector point function Surface, S
Length of the diameter (a) F̄ = y ī + z j̄ + x k̄ S: Part of the sphere
x2 + y2 + z 2 = 1
√ above the xy-plane
AC = a2 + 0 + a2 = a 2;
a (b) F̄ = −y3 ī + x3 j S: Circular disk :
Radius = √ x2 + y2 < 1, z = 0
2 [JNTU 1999]
Let F̄ · d r̄ = y dx + z dy + x dz (c) F̄ = xī + z 2 j̄ + y2 k̄ S: The plane x + y + z = 1
lying in the first octant.
= y ī + z j̄ + x k̄ · dx ī + dy j̄ + dz k̄ (d) F̄ = (y − z + 2) ī S: Surface of the cube x = 0,
+(yz + 4) j̄ − xz k̄ y = 0, z = 0, x = 2, y = 2,
so that F̄ = yī + z j̄ + xk̄ z = 2 above the xy-plane
ī j̄ k̄ 2. Using
Stokes’ theorem
evaluate

∂ ∂ ∂ x + y dx + (2x − z) dy + y + z dz
curl F̄ =
∂x ∂y ∂z C
y z x where C is the boundary of the triangle with

vertices (2,0,0), (0,3,0) and (0,0,6).
= ī (0 − 1) + j̄ (0 − 1) + k̄ (0 − 1) Ans: 21

= − ī + j̄ + k̄
3. If φ is a scalar point function, use Stokes’ theorem
∇S to prove that curl (grad φ) = 0.
Let n̄ = be the unit normal to the surface C
|∇S| 4. Evaluate yz dx + zx dy + xy dz where C is the
(circle) which lies in the plane x + z − a = 0 C
∇S = ∇ (x + z − a) = ī + k̄ curve x2 + y2 = 1, z = y2 by applying Stokes’
∇S i+k theorem.
⇒ n̄ = = √ Ans: 0
|∇S| 2

5. Evaluate F̄ · d r̄ where F̄ = y ī + xz 3 j̄ − zy3 k̄,
By Stokes’ theorem, we have C
where C is the circle x2 + y2 = 4, z = 3/2.
F̄ · d r̄ = curl F̄ · n̄ dS Ans: 19π/2
S
C
6. Evaluate
C x by Stokes’ theorem
ī + k̄ e dx + 2ydy − dz where C is the curve
=− ī + j̄ + k̄ · √ dS C
S 2
√ x2 + y2 = 4, z = 2.
= − 2S Ans: 0

where S is the area of the circle (2) 7. Apply Stokes’ theorem to evaluate y2 dx+x3 dy−
C
1 (x + z) dz where C is the boundary of the triangle
S = πa2
2 whose vertices are (0, 0, 0), (1, 0, 0) and (1, 1, 0).
√ πa2 πa2 1
= − 2× = −√ · Ans:
2 2 3
Laplace Transforms
12
12.1 Introduction A transformation T {f (t)} is called a linear trans-
Laplace transforms belong to a class of integral formation if
transforms, which include Fourier, Hankel, Legendre T {c1 f1 (t)+c2 f2 (t)} = c1 T {f1 (t)}+c2 T {f2 (t)} (12.3)
transforms and are very useful for engineers and
physicists for solving linear differential equations where f1 , f2 are two functions and c1 , c2 are any con-
with constant coefficients under appropriate initial stants.
and/or boundary conditions. Laplace transformation Let K(s, t) be some prescribed function of a
converts an ordinary differential equation into an variable t and a parameter s.
algebraic equation called an auxillary (or subsidiary) A linear integral transformation of functions f (t)
equation and a partial differential equation into one w.r.t the ‘kernel’ K(s, t) is given by
with one independent variable less; i.e., a partial b
differential equation with two independent variables T {f (t)} = K(s, t)f (t)dt (12.4)
x and b, say, becomes an ordinary differential a
equation. Solving these equations and then inverting provided the integral exists. It represents a function
the transformation, one gets the solution of the f (s) or F(s) called the image or transform of f (t)
problem. while f (t) itself is called the inverse image or inverse
One of the important applications of Laplace transform of f (s) or F(s).
Transformation is the solution of mathematical For particular choices of the kernel and the limits
models of physical or engineering systems in which a and b one gets one or the other of the transform of
the driving force is discontinuous or acts for a short Laplace: Fourier, Hankel, Legendre.
duration or is a periodic function.
The operation of differentiation of functions is a
transformation of derivable functions f (t) into their
12.2 Laplace Transformation
derivatives f (t). This can be written as If a = 0, b = ∞ and k(s, t) = e−st then Eq. (12.4)
becomes the Laplace transformation
D{f (t)} = f (t) (12.1)
∞
Here f (t) is the image or transform of f (t) under L{f (t)} = f (t)e−st dt (12.5)
the transformation. Integration is another example of 0
the transformation of functions. named so after the famous French mathematician
x
Laplace.1
I {f (t)} = f (t)dt (12.2)
0 1 Pierre Simon Marquis De Laplace (1749–1827) was a great
French mathematician who developed the foundation of potential
The simple operation of multiplication of func-
theory and made important contribution to Celestial Mechanics
tions by a constant is also a transformation of and Probability Theory. Laplace, known as Newton of France,
functions. taught Napoleon Bonaparte for a year.
∞ ∞
The symbol L is called the Laplace transformation te−st ∞ e−st
(ii) L{t} = te−st dt = − 1. dt
operator. Lf (t) is clearly a function of s. So,it is −s 0 −s
0 0
written as f (s). Inverting the transformation in 1 1 ∞
L{f (t)} = f (s), we have f (t) = L− 1{f (s)} which is = − (0 − 0) − 2 e−st 0
s s
called the inverse Laplace transform of f (s). 1
= 2 (s > 0)
s
Theorem 12.1 (Linearity of Laplace transforma- ∞
tion) (iii) L{t n } = t n e−st dt
Let f1 (t) and f2 (t) be any two function possessing 0
∞
Laplace transforms. If c1 , c2 are any constants then 1
= n+1 u(n+1)−1 e−u du
s 0
L{c1 f1 (t)+c2 f2 (t)} = c1 L{f1 (t)}+c2 L{f2 (t)}. (12.6)
where u = st, du = sdt
Limits: t = 0 → u = 0 ⇒ t = ∞ ⇒ u = ∞
Proof: By definition,
1
∞ = n+1 γ(n + 1)
s
L{c1 f1 (t) + c2 f2 (t)} = e−st [c1 f1 (t) + c2 f2 (t)]dt If n ∈ R, n = −1, −2, − 3, . . .
∞ 0
∞
n!
= c1 −st
e f1 (t)dt + c2 e−st f2 (t)dt = n+1 if n ∈ N ∪ {0},
0 0 s
= c1 L{f1 (t)} + c2 L{f2 (t)}. where γ(x) is the gamma function.
1 1
In particular, when n = − ,
Example 12.1 2 2
*
By direct application of the definition, we obtain −1/2 p(−1/2 + 1) p(1/2) π
Laplace transforms of the following elementary L t = = 1 =
s −1/2+1 s /2 s
functions for t ≥ 0: - 1 . p( 1 + 1) *
1
p( 1
) 1 π
L t 2 = 21 = 2 3/22 =
(i) f (t) = k (k is a non-zero constant) s 2 +1 s 2s s
∞ ∞
(ii) f (t) = t (iii) f (t) = t n −st
(iv) L{e } =
at
e ·e
at
dt = e−(s−a)t dt
0 0
(iv) f (t) = eat (v) f (t) = sin at 1 ∞ 1
= e−(s−a)t 0 = (s > a)
(s − a) s−a
(vi) f (t) = cos at (vii) f (t) = sinh at ∞
(v) L{sin at} = e−st sin at dt
(viii) f (t) = cosh at /
0
,∞
e−st
= [(−s)sin at −acos at]
Solution (−s)2 + a2 0
∞ ∞ a
1 = 2
(i) L{k} = ke−st dt = k − e−st s + a2
0 s 0 ∞
k
= − (0 − 1) =
k
(s > 0) (vi) L{cos at} = e−st cos at dt
s s /
0
,∞
e−st
= [(−s)cos at +asin at]
since, e−∞ = lim e−b = 0 and e0 = 1. (−s)2 + a2 0
b→∞
1 s
In particular L {1} = = 2
s s + a2
Laplace Transforms 12-3
Aliter 4 −s 3
= (e − 1) + (0 − e−s )
∞
e−(s−ia)t ∞ −s −s
I= eiat e−st dt = 4 − e−s
0 −(s − ia) 0 = .
1 s + ia s + ia s
= = 2 (s > a)
s − ia s + ia s + a2 Example 12.3
s ⎧
Re I =L {cos at} = 2 ⎨ 0 0<t<1
s + a2
a f (t) = t 1<t<2
Im I =L {sin at} = 2 ⎩
s + a2 0 t > 2.
∞
Solution
(vii) L{sinh at} = e−st sinh at dt ∞
∞ at
0
L{f (t)} = e−st f (t)dt
−st e − e−at 1
0
2
= e dt ∞
0 2 = 0·e−st dt + t ·e−st dt + 0·e−st dt
∞
1 ∞ −(s−a)t 0 1 2
= e dt − e−(s+a)t dt
2 0 e−st e−st 2

0
= t· − 2
1 1 1 −s s 1
= − −2s −s
2 s−a s+a 2e − e 1
= − 2 (e−2s − e−s )
= 2
a
(s > a) −s s

s − a2 1 1 −s 1 2 −2s
∞ = 2+ e − 2+ e .
(viii) L{cosh at} = e−st cosh at dt s s s s
∞ at
0
Example 12.4
−st e + e−at /
= e dt 10 0≤t<4
0 2 f (t) =
t ≥ 4.
1 ∞ −(s−a)t
∞ 0
= e dt + e−(s+a)t dt Solution ∞
2 0
0 L{f (t)} = e−st f (t)dt
1 1 1
= +
0

2 s−a s+a 4 ∞
s = 10 · e−st dt + 0 · e−st dt
= 2 (s > a). 0 4
s − a2 4
e−st 1 − e−4s
=10 = 10 .
12.2.1 Piece-Wise Continuous −s 0 s
Functions Example 12.5
/
Example 12.2 sin 2t 0<t<π
/ f (t) =
0 t ≥ π.
4 0<t<1
f (t) = . Solution
3 1<t<∞ ∞
Solution ∞ 0
π ∞
L{f (t)} = e−st f (t)dt −st
0
= e sin 2tdt + 0.e−st dt
∞
0 π
1
e−st π
= e−st · 4dt + e−st · 3dt = (−s sin 2t − 2 cos 2t)
0
1
1
∞ (−s)2 + 22 0
e−st e−st −πs
2(1 − e )
=4· +3 = .
−s 0 −s 1 s2 + 4
12.2.2 Linearity Examples (v) f (t) = sin2 2t = 12 (1 − cos 4t)

Example 12.6 1
Find the Laplace transform of each of the following L{f (t)} = (L{1} − L{cos 4t})
2
functions by applying the linearity theorem:
1 1 s 8
= − 2 =
(i) f (t) = eat cosh bt (ii) f (t) = (1 + t)3 2 s s + 16 s(s + 16)
2
(iii) f (t) = t +sin t +cos t (iv) f (t) = sin 2t cos 3t

(vi) cos3 t = 14 (3 cos t + cos 3t)
(v) f (t) = sin 2t
2
(vi) f (t) = cos t3
1
√ √ 3 L{f (t)} = (3L{cos t} + L{cos 3t})
(vii) f (t) = t − 1
√ (viii) f (t) = t− 1
√ . 4
t t 3 s 1 s
= 2 +
Solution We have 4 s + 1 4 s 2 + 32
ebt +e−bt s(s2 + 7)
(i) eat cosh bt = eat = 2
2
(s + 1)(s2 + 9)
1 (a+b) t √
e= + e(a−b) t (vii) L{f (t)} = L{ t − 1
√
1 1
} = L{t 2 } − L{t − 2 }
2 t
1
L{e cosh bt} = L e(a+b) t + L e(a−b) t
at
γ( 12 + 1) γ(− 12 + 1)
2 = −
1 1
1 1 1 s 2 +1 s− 2 +1
= + * * *
2 s−a−b s−a+b 1 π π 1 π
= − = ( − 1)
s−a 2s s s 2s s
=
(s − a)2 − b2
√ 3
(viii) f (t) = t− 1
√
t
(ii) L{(1 + t) } = L{1 + 3t + 3t + t }
3 2 3
= t 3/2 − 3t 1/2 + 3t −1/2 − t −3/2

=L{1} + 3L{t} + 3{t 2 } + L{t 3 } )
1 3 6 6 √ 1 3
= + 2+ 3+ 4 L{f (t)} = L t−√
s s s s t
= L{t 3/2 } − 3L{t 1/2 } + 3L{t −1/2 } − L{t −3/2 }
(iii) L{t +sin t +cos t} = L{t}+L{sin t}+L{cos t}
γ 32 + 1 γ 12 + 1 γ −1 +1
= 3
−3 1
+3 · 2
−1
1 1 s s 2 +1 s 2 +1 s 2 +1
= + 2 + 2
s 2 s +1 s +1 −3
γ 2 +1
s3 + 2s2 + 1 − −3
= 2 2 s 2 +1
s (s + 1)
. γ 12
3 1 1
γ 12 γ 12
= 2 2
− 3. 3/2 + 3. 1/2
2
1 s5/2 s s
(iv) f (t) = sin 2t cos 3t = (cos t − cos 5t) γ (−1/2)
2 −
s−1/2
1 √ √ √
L{f (t)} = (L{cos t} − L{cos 5t}) 3 π 3 π 3 π √ √
2 = − + 1/2 + 2 π. s
1 s s 4 s5/2 2 s3/2 s
= − 1 √ 1 √
2 s2 + 1 s2 + 25 (∴ γ( ) = π, γ(− ) = −2 π)
√ 2 2
12s
= 2 π 3 6 12
(s + 1)(s2 + 25) = ( − 3/2 + 1/2 + 8s1/2 )
4 s5/2 s s
EXERCISE 12.1 11. f (t) = et cosh t.

s−1
Find the Laplace transform of each of the following Ans:
s(s − 2)
functions: /
1. f (t) = eat − ebt . 1 0<t<2
12. f (t) = .
a−b t t > 2.
Ans:
(s − a)(s − b) 1 1 1
Ans: + + 2 e−2s
s s s
2. f (t) = cos2 kt. / t
e 0<t<1
1 1 1 s 13. f (t) =
Ans: · + · 2 0 t > 1.
2 s 2 s + 4k 2
e1−s − 1
Ans:
3. f (t) = (3e2t − 2)2 . 1 −⎧s
1 1 4 ⎨ t 0≤t<1
Ans: 9. − 12. + 14. f (t) = 2−t 1≤t <2
s−4 s−2 s ⎩
√ 0 t ≥ 2.
4. f (t) = cos t. 1 −s −2s
√ % (−1)n t n Ans: 2 (1 − 2e + e )
Hint: cos t = s
(2n)!
∞ (−1)n n! 12.2.3 Inverse Laplace Transforms
Ans: $
(2n)!sn+1 Using the definition, we can write the inverses of
n=0 the Laplace transforms of the elementary functions
5. f (t) = (sin t + cos t)l 2 . given above.
1 2 Table 12.1 Inverse Laplace Transforms
Ans: + 2
s s +4 1. L−1 { ks } = k, L−1 { 1s } = 1
6. f (t) = sin 3t cos 3t. 2. L−1 { snn!+1 } = t n n ∈ N ∪ {0},
Hint: sin 3t cos 3t = 1
2
sin 6t L− 1{ p(n+1)
sn +1 } = t
n n ∈ R, n > 0
3
Ans: 2 3. L−1 { s12 } = t
s + 36
4. L−1 { s−a
1
} = eat
7. f (t) = sin3 2t.
Hint: sin3 2t = 34 sin 2t − 14 sin 6t 5. L−1 { s2 +a
1
2} =
1
sin at
a
4b
Ans: 2 L−1 { s2 +a
s
(s + 4)(s2 + 36) 6. 2 } = cos at
8. f (t) = cosh2 t. 7. L−1 { s2 −a

1
2} =
1
a sinh at
Hint: cosh2 t = 12 (1 + cosh 2t) L−1 { s2 −a
s
8. 2 } = cosh at
s2 + 2s − 4
Ans:
2s(s2 − 4)
12.2.4 First Shifting Theorem or
9. f (t) = e−t sin t. s-Shift Theorem or
1 Translation on the s-Axis
Ans: 2
s + 2s + 2 (Replacement of s by (s − a) in the transform)
Theorem 12.2 If L{f (t)} = f (s) (s > α ≥ 0
10. f (t) = t cos 2t.
∞ and
Hint: Re 0 te2it · e−st dt
a ∈ R), then L{eat f (t)} = f (s − a) (s > a + α).
s2 − 4
Ans: 2 Proof: By definition, we have
(s + 4)2
∞
L{e f (t)} =
at
e−st [eat f (t)]dt Example 12.7
Find the Laplace transform of each of the following
0
∞
functions by applying the first shifting theorem:
= e−(s−a)t f (t)dt
0
(i) e−3t (2 cos 5t − 3 sin 5t) [JNTU 1999]
= f (s − a) (s > a + α) (12.7)
(ii) e2t cos2 t
(iii) 2et sin t cos t

f (s)
a (iv) t cos at [Assam 1999, Andhra 1999]
f (s)
[S.V.U. 1998S]
f (s–a)
(v) e4t sin 2t cos t
O s (vi) t 2 e−2t [Madras 2003]
(vii) cosh at sin at [Delhi, 2002]

Figure 12.1 Translation on the s-axis (first (viii) cosh at cos at.
shifting theorem)
Solution By first shifting theorem and linearity,
Thus, if we know the Laplace transform of f (t), we have
then the transform of [eat f (t)] can be readily obtained
by replacing s by (s − a) in the transform. Fig. 12.1 (i) L {e−3t (2 cos 5t − 3 sin 5t)}
shows the translation on the s-axis.
= 2L{e−3t cos 5t} − 3L{e−3t sin 5t}
Remark
s+3 5
Using Eq. (12.7), we can write =2· −3·
(s + 3) + 5
2 2 (s + 3)2 + 52
f (t) = e−at L−1 {f (s − a)} or 2s − 9
= 2
s + 6s + 34
L−1 {f (s)} = e−at L−1 {f (s − a)}
1
By application of first shifting theorem the follow- (ii) L{cos2 t} = L{1 + cos 2t}
ing useful results are obtained (s > a + α). 2
1
= (L{1} + L{cos 2t})
2
Table 12.2 Applications of the First 1 1 s
Shifting Theorem = + 2 by linearity
k
2 s s +4
1. L{keat } = s−a ∵ L{1} = ks
2. L{eat t n } = γ(n+1)
sn+1
(n ∈ R) ∵ L{t n } = γ(n+1)
sn+1
By first shifting theorem

n!
or sn+1 n!
(n ∈ N ∪ {0})or sn+1 1 1 s−2
L{e2t cos2 t} = +
2 s − 2 (s − 2)2 + 4
3. L{eat sin bt} = (s−a)b2 +b2 b
∵ L{sin bt} = s2 +b 2 (s − 2)2 + 2
=
4. s−a
L{eat cos bt} = (s−a)2 +b2
s
∵ L{cos bt} = s2 +b 2 (s − 2) [(s − 2)2 + 4]
b b
(iii) We have
5. L{eat sinh bt} = ∵ L{sinh bt} = s2 −b
(s−a)2 −b2 2
2 sin t cos t = sin 2t
6. s−a
L{eat cosh bt} = (s−a) s
∵ L{cosh bt} = s2 −b 2
2 −b2 2 L{2 sin t cos t} = L{sin 2t} =
s2 + 4
s
By first shifting theorem (viii) We have L{cos at} = .
s2 + a2
L{et 2 sin t cos t} = L{et · sin 2t} By first shifting theorem
2 2 1
= = 2 L{cosh at cos at} = (L{eat cos at} + L{e−at cos at})
(s − 1)2 + 4 s − 2s + 5 2

(iv) By first shifting theorem 1 s−a s+a
= +
1 (s + ia)2 2 (s − a)2 + a2 (s + a)2 + a2
L{t eiat } = = 2 1 (s +2a +2as)(s − a)+(s2 +2a2 −2as)(s+a)
2 2
(s − ia) 2 (s + a2 )2 =
(s2 − a2 ) + i(2as) 2 (s2 +2a2 )2 −4a2 s2
= 1 (s + 2a )2s + 2as(−2a)
2 2
s3
(s2 + a2 )2 = = .
L{t cos at} =L{Re t eiat } 2 s4 + 4a4 s4 + 4a4
s2 − a2 + i(2as) s 2 − a2 Example 12.8

=Re = If L{f (t)} = f (s), then
(s2 + a2 )2 (s2 + a2 )2
1
Equating the imaginary parts, we have (i) L{sinh at f (t)} = [f (s − a) − f (s + a)]
2
2as 1
L{t sin at} = . (ii) L{cosh at f (t)} = [f (s − a) + f (s + a)]
(s2 + a2 )2 2
Hence evaluate (a) L{sinh 3t sin 2t}
(v) We have (b) L{cosh 3t sin 2t}.
1
L{sin 2t cos t} = (L{sin 3t}+L{sin t}), by linearity Solution By shifting theorem and linearity
2
/ at ,
1
= ( 2
3
+ 2
1
) e − e−at
2 s + 32 s +1 (i) L{sinh at f (t)} = L f (t)
2
s2 + 5 1
= 2 = (L{eat f (t)} − L{e−at f (t)})
(s + 9)(s2 + 1) 2
By first shifting theorem 1
= [f (s − a) − f (s + a)]
(s − 4)2 + 5 2/ ,
L{e4t sin 2t cos t} =
eat + e−at
(s − 4)2 + 9 (s − 4)2 + 1 (ii) L{cosh at f (t)} = L f (t)
2 2
(vi) We have L{t 2 } = 3
s 1
By first shifting theorem = L{eat f (t)} + L{e−at f (t)}
2
2
L{t 2 e−2t } = 1
= [f (s − a) + f (s + a)]
(s + 2)3 2
a
(vii) We have L{sin at} = 2 .
s + a2 2
By first shifting theorem (a) f (s) = L{sin 2t} =
s2 + 4
1 2
L{cosh at sin at} = (L{eat sin at} + L{e−at sin at}) ⇒ f (s − a) = f (s − 3) =
2 (s − 3)2 + 4

1 a a 2
= + =
2 (s − a)2 + a2 (s + a)2 + a2 s2 − 6s + 13
a 2(s2 + 2a2 ) and f (s + a) = f (s + 3) =
2
= · 2 (s + 3)2 + 4
2 (s + 2a2 )2 − 4a2 s2
a(s2 + 2a2 ) 2
= 4 =
s + 4a4 s2 + 6s + 13

1 2 2 1 s−a s+a
L{sinh 3t sin 2t} = − L{cosh at cos bt} = +
2 (s − 3)2 +4 (s + 3)2 +4 2 (s−a)2 +b2 (s+a)2 +b2
1 1 s(s2 − a2 + b2 )
= 2 − = (1)
(s +13)−6s (s2 +13)+6s D
12s (α) − (β)
= 2 gives
(s + 13)2 − 36s2 2
12s 1 s−a s+a
= 4 L{sinh at cos bt} = −
s − 10s2 + 169 2 (s−a)2 +b2 (s+a)2 +b2
a(s2 − a2 − b2 )
(b) L{cosh 3t sin 2t} = (2)
D
1 2 2 (γ) + (δ)
= + gives
2 (s − 3)2 + 4 (s + 3)2 + 4 2
1 1 1 b b
= 2 + 2 L{cosh at sin bt} = +
(s + 13) − 6s (s + 13)2 + 6s 2 (s−a)2 +b2 (s+a)2 +b2
2(s2 + 13) b(s2 + a2 + b2 )
= 4 . = (3)
s − 10s2 + 169 D
(γ) − (δ)
Example 12.9 ∞ gives
2
Evaluate the complex integral 0 e(a+ib)t · e−st dt and
1 b b
deduce Laplace transforms of L{sinh at sin bt} = −
(i) cosh at cos bt (ii) sinh at cos bt 2 (s−a)2 +b2 (s+a)2 +b2
(iii) cosh at sin bt (iv) sinh at sin bt. 2abs
= (4)
D
Solution where D = s4 +2(b2 −a2 )s2 +(a2 +b2 )2 , as shown
∞ ∞
e−(s−a−ib)t below.
e(a+ib)t e−st dt =
0 −(s − a − ib) 0
(s − a)[(s + a)2 + b2 ] + (s + a)[(s − a)2 + b2 ]
1 s − a + ib
= = (s−a)(s+a)[s+a+s−a]+b2 [s−a+s+a]
s − a − ib s − a + ib
(s − a) + ib = (s2 − a2 ) · 2s + 2b2 s = 2s(s2 − a2 + b2 )
=
(s − a)2 + b2 (s − a)[(s + a)2 + b2 ] − (s + a)[(s − a)2 + b2 ]
Equating the real and imaginary parts and also = (s−a)(s+a)(s+a−s+a)+b2 (s−a−s−a)
replacing a by −a we obtain = (s2 − a2 ) · 2a − 2ab2 = 2a(s2 − a2 − b2 )
s−a b[(s + a)2 + b2 ] + b[(s − a)2 + b2 ]
L{eat cos bt} = (α)
(s − a)2 + b2 = b(2s2 + 2a2 + 2b2 ) = 2b(s2 + a2 + b2 )
s+a
L{e−at cos bt} = (β) b[(s + a)2 + b2 ] − b[(s − a)2 + b2 ]
(s + a)2 + b2
b = b(4as) = 4abs
L{eat sin bt} = (γ) [(s − a)2 + b2 ][(s + a)2 + b2 ]
(s − a)2 + b2
b = (s2 − a2 )2 + b2 [(s + a)2 + (s − a)2 ] + b4
L{e−at sin bt} = (δ)
(s + a)2 + b2 = s4 − 2a2 s2 + a4 + 2b2 (s2 + a2 ) + b4
(α) + (β)
gives = s4 + 2(b2 − a2 )s2 + (a2 + b2 )2 = D (say)
2
In particular, choosing a = 3, b = 2, we get end point from within the interval

from Eq. (1)
i.e., lim f (t) and lim f (t) exist.
t→c+ t→d −
s(s2 − 5)
L{cosh 3t cos 2t} = 4 ;
s − 10s2 + 169 Fig. 12.2 shows the graph of a function f (t) which
choosing a = 2, b = 1 we get from Eq. (2) is piece-wise continuous in [0, 6].
2(s2 − 5)
L{sinh 2t cos t} = ; Functions of exponential order
s4 − 6s2 + 25
choosing a = 3, b = 2, we get from Eq. (3) If the integral of e−st f (t) between the limits 0 and t0
2(s2 + 13) exists for every finite t0 the only remaining threat to
L{cosh 3t sin 2t} = ; the existence of the transform
s4 − 10s2 + 169 ∞
and choosing a = 2, b = 3 we get from Eq. (4) e−st f (t)dt (12.8)
12s 0
L{sinh 2t sin 3t} = 4 .
s + 10s + 169 is the behaviour of the integrand as t → ∞.
∞
12.3 Existence of Laplace We know that e−ct dt (12.9)
0
Transform
∞ converges for c > 0. So,we have to consider func-
Consider the integral 0 e−st f (t)dt.
tions f (t) which, for large t(t ≥ t0 ), are bounded by
In elementary calculus, we have found that finite
some exponential ebt so that the integrand in (12.12)
discontinuities or finite jumps of the integrand do
behaves like the integrand in (12.13) for large s.
not affect the existence of the integral. We therefore
introduce a phrase to describe functions which are Definition A function f (t) is said to be of expo-
continuous except for such jumps. nential order as t → ∞ if there exist constants M
and b and a fixed t-value to such that
12.3.1 Piece-Wise(or Sectionally) |f (t)| < Mebt for t ≥ t0 (12.10)

Continuous Functions of
If b is to be emphasized we say that f (t) is of order
Exponential Order
ebt as t → ∞. We write
Definition A function f (t) defined over an inter-
val [a, b] is said to be piece-wise (or sectionally) f (t) = O(ebt ) t → ∞ (12.11)
continuous over [a, b] if that interval can be divided
to mean that f (t) is of exponential order, the expo-
into a finite number of sub–intervals c ≤ t ≤ d such
nential being ebt as t → ∞.
that in each subinterval
Geometrically, the above condition implies that the
1. f (t) is continuous in the open interval (c, d) and
graph of f (t) (t > t0 ) does not grow faster than the
2. f (t) approaches a limit as t approaches each
graph of the exponential function g(t) = Mebt (t > 0).
For example,
f (t)
|t|, |t 2 |, |t n | ≤ et
|e−3t | < et | cos t| | sinh t| < et (t ≥ 0)

and so the function f1 (t) = t, f2 (t) = t , f3 (t) = t n ,
2
f4 (t) = e−3t , f5 (t) = cos t, f6 (t) = sinh t are all of

0 1 2 3 4 5 6 t exponential order. Now, consider
2
Figure 12.2 f (t) = eat (a > 0)
It is not possible to determine constants b and 12.3.2 Change of Scale Theorem

2 2
M such that eat ≤ Mebt . Therefore, eat is not of
Theorem 12.5 If L{f (t)} = f (s) then
exponential order.
1 s
Theorem 12.3 (Sufficient conditions for the L{f (at)} = f .
a a
existence
t0 −st of Laplace transforms) If the integral
e f (t)dt exists for every finite positive t0 and Proof: We have
0
if f (t) is of exponential order f (t) = (ebt ) as t → ∞ ∞
then the Laplace transform L{f (at)} = e−st f (at)dt

∞ 0
1 ∞ −( s )u
L{f (t)} = e−st f (t)dt = f (s) = e a f (u)du
0
a 0
Put at =u, adt = du
exists for s > b.
limits are t =0 ⇒ u = 0
∞
Proof: We write the integral 0 e−st f (t)dt as a t =∞ ⇒ u = ∞
sum of two integrals 1 s
∞ t0 ∞ ⇒ L{f (at)} = f
a a
e−st f (t)dt= e−st f (t)dt + e−st f (t)dt (12.12)
0 0 t0 The effect of multiplying the argument t of f by
s
a is to replace s by in the transform f (s) and
Now since f (t) is of exponential order there exist a
constants b and M such that 1
multiplying it by .
a
|f (t)| < Meb t for t ≥ t0
∞ ∞ Example 12.10
−st
∴ e f (t)dt < e−st |f (t)|dt Find the Laplace transform of each of the following
t0 t0 functions by applying the change of scale theorem:
∞
<M e−st · ebt dt (12.13) (i) sinh at (ii) cosh 5t (iii) cos at
t0
sin at sin t 1
(iv) given that L{ } = tan−1
M exp[−t0 (s − b)] t t s
= → 0 as t0 → ∞ for s > b.
(s − b) s2 + 2
(v) f (kt) given that L{f (t)} = 4
Therefore, the last integral in Eq. (12.13) is abso- s +4
lutely convergent for s > b (in the case of complex s, 2s
it converges for Re s > b), and thus the theorem is (vi) sinh kt sin kt given that L{sinh t sin t} =
s4 +4
proved. s+3
If f (t) is piece-wise continuous (vii) f (t/3) given that L{f (t)} =
Theorem 12.4 s2 + 6s + 18
over every finite interval in the range t ≥ 0 and if e−3/s
f (t) is of exponential order, f (t) = o(ebt ) as t → ∞, (viii) eat f (kt) given that L{f (t)} = 2 .
s
the Laplace transform L{f (t)} exists for s > b.
Solution
1
Proof: From Integral Calculus, we know that a (i) We have L{sinh t} = .
s2 − 1
function which is piece-wise continuous over an By applying change of scale theorem
interval is integrable over that interval.
∴ By Theorem 12.3, it follows that if f (t) is piece- 1 1
L{sinh at} = . s 2
wise continuous over every finite interval in the range a (a) − 1
t ≥ 0 and if f (t) is of exponential
∞ −st order f (t) = O(e )
bt
1 a2 a
as t→∞ then L{f (t)} = 0 e f (t)dt exists for s > b. = . = 2
a s 2 − a2 s − a2
s
(ii) We have L{cosh t} = . 12.3.3 Laplace Transforms of
−1s2 Derivatives
By applying change of scale theorem
Theorem 12.6 Theorem of transforms of deriva-
1 (s/5) 1 5s .52 s tives (multiplication by s) If f (t) is continuous for
L{cosh 5t} = . = . 2 = 2
5 (s/5) − 1
2 5 s −5 2 s − 25 t ≥ 0 and is of exponential order and f (t) is at
least piece-wise continuous and is also of exponential
s
(iii) We have L{cos t} = . order on [0, ∞] then
+1 s2
By applying change of scale theorem
L{f (t)} = sL{f (t)} − f (0). (12.14)
1 s/a 1 (s/a) · a 2
L{cos at} = . = . 2 Proof: Suppose f (t) is continuous on [0, ∞].
a (s/a)2 + 1 a s + a2
s Then
= 2
s + a2 ∞
/ , L{f (t)} = e−st f (t)dt

sin at 1 1 1 a 0

(iv) L = tan−1 = tan−1 ∞ ∞
at a s/a a s = e−st f (t) 0 + s e−st f (t)dt
0
By change of scale theorem, we have
= − f (0) + sL{f (t)} (s > b)
sin at a
⇒ L{ } = tan−1 Suppose now that f (t) is piece-wise continuous
t s
and f (t) has
a finite jump at t = t0 . Then
(v) By change of scale theorem ∞
1 (s/k) + 2 1 (s + 2k )/k
2 2 2 2 0
L{f (kt)} = · = . 4
k (s/k) + 4
4 k (s + 4k 4 )/k 4 t0 −t ∞
k(s + 2k )
2 2 = lim e−st f (t)dt + e−st f (t)dt
= 4 t→0 0 t0 +t
s + 4k 4 /
t0 −t t0 −t
= lim e−st f (t) +s e−st f (t)dt
(vi) By change of scale theorem, we have t→0 0
0
∞ ,
∞
1 2(s/k) 2k 2 s + e−st f (t) t +t +s −st
e f (t)dt
L{sinh kt sin kt} = . = 0
k (s/k)4 + 4 s4 + 4k 4 -
t0 +t
−st
∞ . t0 −t
(vii) By change of scale theorem = lim f (t) 0 + e−st f (t) t +t
e
t→0 0
t0 −t ∞
1 3s + 3 + s lim −st
e f (t)dt + −st
e f (t)dt
L{f (t/3)} = t→0
(1/3) (3s) + 6(3s) + 18
2

0

t0 +t
−st0
9(s + 1) s+1 =e f (t0− ) − f (t0+ ) − f (0+)
= = ∞
9(s2 + 2s + 2) (s + 1)2 + 1 −st
+s e f (t)dt
(viii) By change of scale theorem, we have 0
= e−st0 (0) − f (0+ ) + sL{f (t)}
−3k/s 2 −3k/s
e k e
L{f (kt)} = =
(s/k)2 s2 ∵ f (t0− ) = f (t0+ ) and f (0+) = f (0)
By first shifting theorem ⇒ L{f (t)} = sL{f (t)} − f (0).

k 2 e−3k/(s−a) Note
L{eat f (kt)} = .
(s − a)2 If f (t) satisfies the conditions of the above theorem
√
and f (0) = 0 then √ π −1/4s
(vi) If L{sin t} =
3/2
e
2s )
L{f (t)} =sf (s) or √
cos t π 1
2
d −1 prove that L √ = e−1/4s
−1
L {sf (s)} =f (t) = L {f (s)} t s
dt
/ , / ,
s s
where L{f (t)} = f (s). This result is useful in find- (vii) L−1 (viii) L−1
s2 + 4 s2 − 25
ing the inverse Laplace transform when the given
function φ(s) can be written as sf (s). Solution
(i) We have f (t) = t 3 , f (t) = 3t 2 , f (t) = 6t,
Laplace transforms of higher derivatives f (t) = 6 and f (0) = 0, f (0) = 0, f (0) = 0.
We have By the theorem on transforms of derivatives
L{f (t)} = sL{f (t)} − f (0) L{f (t)} = s3 L{f (t)} − s2 f (0) − sf (0) − f (0)
= s[sL{f (t)} − f (0)] − f (0) ⇒ L{6} = s3 L{t 3 } − s2 ; 0 − s, 0 − 0
= s2 L{f (t)} − sf (0) − f (0) 6 6
⇒ = s3 L{t 3 } ⇒ L{t 3 } = 4 .
L{f (t)} = sL{f (t)} − f (0) s s
= s3 L{f (t)}−s2 f (0)−sf (0)−f (0) (ii) We have f (t) = sin2 t,

f (t) = 2 sin t cos t = sin 2t and f (0) = 0
We can prove the following theorem by induction
By the theorem on transforms of derivatives
for the nth order derivative f (n) (t).
Theorem 12.7 Let f (t), f (t), ... f (n−1) (t) be L{f (t)} =sL{f (t)} − f (0)
continuous on [0, ∞] and be of exponential order. ⇒ L{sin 2t} =sL{sin2 t} − 0
Suppose f (n) (t) is at least piece-wise continuous on 1 2
[0, ∞] and is of exponential order. Then L{f (n) (t)} = ⇒ L{sin2 t} = L{sin 2t} =
s s(s2 + 4)
sn L{f (t)} − sn−1 f (0) − sn−2 f (0) − · · · − f (n−1) (0).
Note (iii) We have f (t) = teat ,
1. If f (0) = 0, f (0) = 0 then we have f (t) = eat + ateat = eat + af (t) and f (0) = 0.
L{f (t)} = s2 {f (t)} By the theorem on transforms of derivatives
d2
⇒ L−1 {s2 f (s)} = f (t) = 2 L−1 {f (s)} .
dt L{f (t)} = sL{f (t)} − f (0) = sL{f (t)}, ∵ f (0) = 0
2. Also, if f (0) = 0, f (0) = 0, ..., f (n−1) (0) = 0 ⇒ L{eat + af (t)} = sL{f (t)}
then we have L{f (n) (t)} = sn L{f (t)} 1
⇒ (s − a)L{f (t)} = L{eat } =
s−a
⇒ L−1 {sn f (s)} = f (n) (t). 1
⇒ L{f (t)} = L{teat } =
(s − a)2
Example 12.11
Find the Laplace transform of each of the follow- (iv) We have
ing functions using the theorem on transforms of
derivatives: f (t) = t cos at, f (t) = cos at − at sin at
(i) f (t) = t 3 (ii) f (t) = sin t (iii) f (t) = teat
2
f (t) = −2a sin at − a2 t cos at
1 = −2a sin at − a2 f (t)
(iv) f (t) = t cos at (v) f (t) = √
πt and f (0) = 0, f (0) = 1.
By the theorem on transforms of derivatives (vii) Using the theorem on transforms of

d
L{f (t)} = s2 L{f (t)} − sf (0) − f (0) derivatives, i.e., L−1 {sf (s)} = f (t)
dt
= s2 L{f (t)} − 1
1
Let f (s) =
⇒ L{−2a sin at − a f (t)} = s L{f (t)} − 1
2 2 s2 +4
/ ,
1
a Then f (t) = L−1 {f (s)} = L−1
⇒ (s + a )L{f (t)} = 1 − 2a 2
2 2
s2 + 4
s + a2
1
s −a
2 2
= sin 2t and f (0) = 0
= 2 2
s + a2
a / ,
∵ L{sin at} = 2 −1 s d
s + a2 Now, L = L−1 {s · f (s)} = f (t)
s +4
2 dt
s 2 − a2
⇒ L{f (t)} = L{t cos at} = 2 . d 1
(s + a2 )2 = sin 2t
dt 2
√
(v) We know that if g(t) = t then 1
1 = .2 cos 2t = cos 2t
g (t) = √ g(0) = 0 and that 2
2 t 1
1√
(viii) Let f (s) =
1 γ( 12 + 1) π s2 − 52
L{g(t)} =L{t 2 } = 1 +1
= 2 3/2 1
s2 s Then f (t) = L−1 {f (s)} = L−1 { }
s 2 − 52
Also, by the theorem of transforms of derivatives 1
L{g (t)} =sL{g(t)} − g(0) = sL{g(t)} sinh 5t and f (0) = 0
=
√ √ 5
π π / ,
=s. 3/2 = 1/2 s d
2s 2s Now, L−1 2 = L−1 {sf (s)} = f (t)
/ , √ s − 25 dt
1 π
⇒ L √ =sL{g(t)} = √ 1d
2 t 2 s = sinh 5 = cosh 5t.
/ , * 5 dt
1 π
⇒L √ =
t s 12.3.4 Laplace Transforms of
/ ,
1 1 Integrals (Division by s)
L{f (t)} =L √ =√
πt s In many electrical engineering problems, we have
√ to solve integro-differential equations. Consider for
cos t √
(vi) f (t) = √ . If g(t) = sin t then example, a series electrical circuit in Fig. 12.3.
t
√ Using Kirchoff ’s second law we derive the fol-
cos t lowing integro-differential equation for the current
g (t) = √ and g(0) = 0.
2 t flow I (t).
By the theorem on transforms of derivatives
√ dI (t)
L{g (t)} = sL{g(t)} − g(0) = sL{sin t} L + RI (t) + C −1 Q(t) = E0 cos wt
dt
√ )
cos t π 1 π 12 −1/4s t
L √ = s. 3/2 .e−1/4s = e where Q(t) = 0 I (t)dt is the quantity of electricity.
2 t 2s 2 s In order to apply Laplace transform technique for the
√ ) solution of such problems we need a formula for the
cos t π 1/2 −1/4s
L √ = e Laplace transform of an integral, which we derive
t s
below.
By the theorem on transforms of integrals

R
E(t) =
L
/ t ,
E0 cos wt 1
~ L f (u)du = f (s)
C 0 s
/ t ,
1 1
⇒L (1 − e−u )du = ·
0 s s(1 − s)
1
Figure 12.3 Series electric circuit = 2 .
s (1 − s)
E(t) = E0 cos wt, impressed voltage
R = resistance; L = inductance; Example 12.13
C = capacitance f (t) = sin at.
Theorem 12.8 (Theorem on transforms of inte- Solution

grals) Let f (t) be a piece-wise continuous function a
L{f (t)} = L{sin at} = = f (s)
for t ≥ 0 which is of exponential order. Then s 2 + a2
/ t , By the theorem on transforms of integrals
1
L f (u)du = L{f (t)} s > 0, s > b. / t ,
s 1
0
L f (u)du = f (s)
f (t) is piece-wise continuous for s
Proof: t t ≥ 0
0
/ t ,
and is of exponential order. So, φ(t) = 0 f (u)du 1 a a
⇒L sin audu = · 2 = .
is at least piece-wise continuous and is of exponen- 0 s s +a 2 s(s + a2 )
2
tial order.
Further, φ (t) = f (t) except for points at which Example 12.14
f (t) is discontinuous. / ,
sin t sin t 1
That is, φ (t) is piece-wise continuous on each f (t) = given that L = tan−1 .
t t s
finite subinterval. Applying the theorem on trans- / ,
forms of derivatives, we have sin t 1
Solution L{f (t)} = L = tan−1
t s
L{f (t)} = L{φ (t)} = sL{φ(t)} − φ(0), By the theorem on transforms of integrals
s > b = sL{φ(t)} ∵ φ(0) = 0 / t ,
/ t , sin u 1 sin t 1 1
1 L du = .L{ } = tan−1 .
⇒L f (u)du = L {f (t)} 0 u s t s s
0 s
/ , t
1 Example 12.15
or L−1 f (s) = f (u)du
s 0 f (t) = e−t sinh t.
where f (s) = L{f (t)} 1
Solution L{sinh t} = ,
Using the theorem on transforms of integrals find s2 −1
t
the Laplace transform of 0 f (u)du where f (t) is 1
L{e−t sinh t} =
given below. (s + 1)2 − 1
1
Example 12.12 = 2 , by shifting property
f (t) = 1 − e−t . s + 2s
Solution
By the theorem on transforms of integrals
/ t ,
1 1 1
L{f (t)} = L{1 − e−t } = − L e sinh udu = L{e−t sinh t}
−u
s s−1 0 s
1 1 1 1
= = f (s) = . 2 = 2 .
s(1 − s) s s + 2s s (s + 2)
/ , / ,
Using the theorem on transforms of integrals 1 −1 1 −1 1
- . = L −L
4 s −4
2 s2
t
i.e., L−1 1s f (s) = 0 f (u)du
1 1 1
= sinh 2t − t = (sinh 2t − 2t).
find inverse of the functions. 4 2 8
Example 12.16, EXERCISE 12.2

/
1 1. Find the Laplace transform of each of the follow-
L−1 .
s(s + 16)
2 ing functions:
Solution Let (a) t 2

1 1 2
f (s) = ⇒ f (t) = L−1 {f (s)} = sin 4t Ans:
s3
+ 16
s2 4
/ , t (b) tn
1
∴ L−1 f (s) = f (u)du n!
s 0 Ans:
t sn+1
1
= sin 4udu (c) cos2 2t
4 0
s2 + 8
cos 4u t 1 Ans:
=− = (1 − cos 4t). s(s2 + 16)
16 0 16
Example 12.17 (d) (sin t − cos t)2
/ ,
1 s2 + 2
L−1 . Ans:
s2 (s2 − 4) s(s2 + 4)
1 (e) t sin at
Solution Let f (s) = 2as
s2 − 4 Ans: 2
(s + a2 )2
1
Then f (t) = L−1 {f (s)} = sinh 2t (f) t sinh at
/ , / 2, 2as
−1 1 −1 1 Ans: 2
L f (s) = L (s − a2 )2
s s(s − 4)
2. Find the inverse Laplace transforms of the follow-
1 t 1
= sinh 2udu = [cosh 2u]t0 ing functions:
2 0 4
1 1
= (cosh 2t − 1) (a)
+s
s2
/ , 4 / , Ans: 1 − e−t
1 1
L−1 2 f (s) = L−1 2 2 1
s s (s − 4) (b)
t + 16s
s3
1 1
= (cosh 2u − 1)du Ans: (1 − cos 2t)
4 0 4
t
1 1
= sinh u − u (c)
1
4 2 0 + s2
s3
1 Ans: t + e−t − 1
= (sinh 2t − 2t)
8 3
Alternatively (d)
/ , s2 (s2 + 9)
1 1 1 1 1
L −1
= L{ 2 − 2} Ans: (3t − sin 3t)
s (s − 4)
2 2 4 s −4 s 9
12.3.5 Multiplication by t 1
(ii) We know that L{e3t } =
Differentiation of s−3
Transforms By the theorem on multiplication by t twice
2
Theorem 12.9 Let f (t) be piece-wise continuous d2 1 1·2
on [0, ∞] and be of exponential order. Then L{t 2 e3t } = (−1)2 2 =
ds s − 3 (s − 3)3
2
L{tf (t)} = −f (s) and =
n (s − 3)3
L{t n f (t)} = (−1)n f (s) (s > b)
3
dn (iii) We know that L{sin 3t} =
n
where f (s) = L{f (t)} and f (s) = n f (s). s2 +9
ds By shift theorem
Proof: Laplace transform of f (t) is defined by
3 3
∞ L{e−2t sin 3t} = = 2
(s + 2) + 9
2 s + 4s + 13
L{f (t)} = f (s) = e−st f (t)dt
0 By the theorem on multiplication by t, we have
By Leibnitz’s rule for differentiation under the
d 3
integral sign L{te−2t sin 3t} = −
ds s2 + 4s + 13
∞
d d 3(2s + 4)
f (s) = e−st f (t)dt = 2
ds ds 0 (s + 4s + 13)2
∞
∂ 6(s + 2)
= f (t)( e−st )dt = 2
∂s (s + 4s + 13)2
0 ∞
s
= f (t)(−te−st )dt (iv) We know that L{cos at} = 2
0
∞ s + a2
By the theorem on multiplication by t, we have
=− e−st [tf (t)]dt
0 d s
L{t cos at} = −
= −L{t f (t)} ds s2 +a2
s · 2s 1
By induction, we can prove the result for the nth = 2 2 2− 2 2
derivative. (s +a ) s +a
Note s 2 − a2
= 2 .
Inverting the results of the theorem, we have (s + a2 )2
n
L−1 {f (s)} = −tf (t) and L−1 {f (s)} = (−1)n t n f (t). Example 12.19
∞
3
Example 12.18 Show that te−t sin 3tdt = .
Find the Laplace transforms of the following func- 0 50
tions: (i) t sin 2t, (ii) t 2 e3t , (iii) te−2t sin 3t, Solution Let
(iv) t cos at, (Assam,1998; Andhra,1989) ∞ ∞
−t
Solution I= te sin 3tdt = e−t (tsin 3t)dt
(i) We know that L{sin 2t} =
2
0
∞
0
s2 + 4 If f (s) = e−st f (t)dt

By the theorem on multiplication by t we have 0

d 2 2 · 2s
L{t sin 2t} = − = 2 where f (t) = t sin 3t the required I = f (1)
ds s2 + 4 (s + 4)2 3
We know that L{sin 3t} = 2
= 2
4s s +9
(s + 4)2 By the theorem on multiplication by t we have
d 3 6s Under the assumption of the theorem, the order of

f (s) = L{sin 3t} = − ( 2 )= 2
ds s + 9 (s + 9)2 integration can be reversed. By doing so we obtain
6 3
I = f (1) = 2 = . ∞ ∞ ∞
10 50
f (u)du = e−ut du f (t)dt
EXERCISE 12.3 s 0 s
∞ −ut ∞
e
1. Find the Laplace transforms for the following = − f (t)dt
functions: 0 t s
∞
f (t)
(a) t 2 sin at = e−st dt
t
2a(3s2 − a2 ) 0
/ ,
Ans: f (t)
(s2 + a2 )3 =L s > b.
t
(b) t 3 e−at
6 Note
Ans:
(s + a)4 Inverting the transform, we obtain
/ ∞ ,
(c) t sin2 3t −1 f (t)
L f (u)du = or f (t)
54(s2 + 12) s t
Ans: 2 2 / ∞ ,
s (s + 36)2 −1
=tL f (u)du
(d) t 2 cos 3t s
2s(s2 − 27)
Ans: This result is useful when the integral of a
(s2 + 9)3
∞ transform is simpler than the transform itself.
3
2. Show that te−3t sin tdt = (Laplace
0 50 Example 12.20
transform of ∞t sin t with s = 3).
Find the Laplace transforms of
6
3. Show that e−2t sin3 tdt = (Laplace 1 − et e−at − e−bt sin t
0 65 (i) (ii) (iii)
t t t
transform of sin t with s = 2).
3
sin 3t cos t (cos at − cos bt)
(iv) (v) .
t t
12.3.6 Division by t (Integration of Solution
Transforms) 1 1
(i) We know that L{1 − et } = − .
Theorem 12.10 Let f (t) be piece-wise continu- s s−1
ous on (0, ∞) and be of exponential order. Suppose By the theorem on division by t, we have
that / , ∞
f (t) f (t) / , ∞
lim exists. Then L = f (u)du s > b, 1 − et 1 1
+ t t L = − du
t→0 s t s u u−1
where L{f (t)} = f (s).
=[log u − log(u − 1)]∞
s
∞
Proof: By definition u
∞ = log
f (s) = e−st f (t)dt u−1 s
0 s
=0 − log
s−1
Integrating both sides ∞s to ∞, wehave
∞ ∞ from s−1
f (u)du = e−ut f (t)dt du = log
s
s s 0
1 1 s s
(ii) We know that L{e−at − e−bt } = − . (v) We know that L{cos at−cos bt} = −
s+a s+b s2 +a2 s2 +b2
By the theorem on division by t, we have By the theorem on division by t, we have
/ , / ,
e−at − e−bt ∞
1 1 cos at − cos bt
L = − du L
t u+a u+b t
s ∞
∞ u u
= [log(u + a) − log(u + b)] . = − du
s u 2 + a2 u 2 + b2
s
s+a 1 ∞
= − log = log(u2 + a2 ) − log(u2 + b2 ) s
s+b 2
s+b 1 s 2 + a2
= log = 0 − log 2
s+a 2 s + b2

k +a s 2 + b2
∵ lim log =0 = log .
k→∞ k +b s 2 + a2
1 EXERCISE 12.4
(iii) We know that L{sin t} = .
s2 + 1
By the theorem on division by t, we have 1. Find the Laplace transforms for the following
functions:
/ , ∞
sin t 1
L = du (a) (e−4t − e−3t )/t
t s u2 + 1 s+3
∞ Ans: log
= tan−1 s s s+4
π
= − tan−1 s (b) (1 − cos at)/t
2 1
= cot−1 s Ans: log[(s2 + a2 )/s]
2
(iv) We know that (c) (sin2 t)/t
sin 3t cos t = 12 (sin 4t + sin 2t) and that 1 s2 + 4
Ans: log( )
4 s2
1 1 ∞
L{sin 3t cos t} = L{sin 4t} + L{sin 2t} sin2 t
2 2 2. Evaluate e−2t dt
2 1 ) t
= 2 + 0
2
s + 16 s2 + 4 sin2 t 1 s +4
Hint: L = log
t 4 s2
By the theorem on division by t, we have s=2
s=2
1
/ , ∞
= log 2
sin 3t cos t 2 1 4
L = 2 + 16
+ 2 du 1
t s u u +4 Ans: log 2
1 u u ∞ 4
= tan−1 + tan−1 ∞ −3t
2 4 2 s e − e−6t
1 π s 3. Show that
t
dt = log 2.
= − tan−1 0
2 2 4 [Hint: Put a = 3, b = 6 and s = 0 in Example 5].
π s
+ − tan−1 4. Show
2 2 t that the Laplace transform of
1 s s sin t 1
= cot−1 + cot−1 et · dt = cot−1 (s − 1).
2 4 2 0 t s
12.3.7 Initial Value Theorem 12.4 Heaviside Function or

Theorem 12.11 If L{f (t)} = f (s) then Unit-Step Function
lim f (t) = lim sf (s). The Heaviside function is a function H (t) defined by
t→0 s→∞
Proof: By the theorem on transformation of /
0 if t < 0
derivatives, we have H (t) =
1 if t ≥ 0.
L{f (t)} = sf (s) − f (0) It is also called a unit-step function and is denoted
∞
by U (t). If the jump discontinuity is at a point t = a
or e−st f (t)dt = sf (s) − f (0)
then
0
∞ /
or sf (s) = f (0) + e−st f (t)dt 0 if t < a
H (t − a) = U (t − a) =
0 1 if t ≥ a.
Now taking limit as s → ∞ we have The jump is of magnitude unity. The unit-step
∞ function U (t − a) is also denoted by u(a) (t) so that
lim sf (s) = f (0) + lim e−st f (t)dt H (t − a) = U (t − a) = Ua (t).
s→∞ s→∞ 0
∞

= f (0) + lim e−st f (t)dt f(t) f(t)
0 s→∞
1 1
= f (0) + 0 = lim f (t)
t→0
since lim e−st = 0.

s→∞
O t O a t
12.3.8 Final Value Theorem
Theorem 12.12 If L{f (t)} = f (s) then Figure 12.4 (a) Heaviside function H (t) or
unit-step function U0 (t) (b) Heaviside function
lim f (t) = lim sf (s)
t→∞ s→0 H (t − a) or unit-step function Ua (t)
Proof: By the theorem on transformation of Example 12.21

derivatives, we have We give below some examples of functions repre-
sented in terms of the unit-step function.
L{f (t)} = sf (s) − f (0) /
∞ 0 if t < π
1. f (t) = sin tUπ (t) =
or e−st f (t)dt = sf (s) − f (0) sin t if t ≥ π.
/
0
∞ 0 if t < 2
2. g(t) = tU2 (t) =
or sf (s) = f (0) + e−st f (t)dt t if t ≥ 2.
0
3. If a < b then a pulse function is defined
Now taking limit as s → 0, we have by (Fig. 12.6)
∞ ⎧
⎨ 0 if t < a
lim sf (s) = f (0) + lim e−st f (t)dt
s→0 s→0 0 p(t) = Ua (t) − Ub (t) = 1 if a ≤ t < b
∞ ⎩
0 if t ≥ b.
= f (0) + lim e−st f (t)dt
s→0 4. Express the following piece-wise function in terms
0 ∞
∞ of unit-step function:
= f (0) + f (t)dt = f (0) + f (t) 0
/
0
4t 0 ≤ t ≤ 3
= f (0) + lim f (t) − f (0) = lim f (t). f (t) =
t→∞ t→∞ 0 t ≥ 3.
f(t) (Putting t − a = u ⇒ dt = du and the limits for u

are u = 0 when t = a and U = ∞ when t = ∞.)
1 3 ∞
2 t
= e−s(u+a) f (u)du = e−as L{f (t)} = e−as f (s).
O 0
Note
Figure 12.5
1. L{f (t − a)Ua (1)} =e−as L{f (t)} = e−as f (s)
g(t)
⇒L−1 {e−as f (s)}
=f (t − a)Ua (t)
2. In particular when f (t) = 1, we have

O 2 t
1
g(t) = t u2 (t) L{Ua (t)} = e−as L{1} = e−as .
s
p(t) 3. If the function is in the form f (t)Ua (t), then
f (t)Ua (t) =f (t − a + a)Ua (t)

=g(t − a)Ua (t)
O t where g(t − a) =f (t) or g(t) = f (t + a).
a b
By the use of the above theorem

Figure 12.6
Solution f (t) can be expressed in terms of the L{g(t − a)Ua (t)} =e−as L{g(t)} = e−as L{f (t + a)}
step function U3 (t) as shown below: or L{f (t)Ua (t)} =e−as L{f (t + a)}.
f (t) = 4t − 4tU3 (t) = 4t[1 − U3 (t)] Example 12.22

= 4t[U0 (t) − U3 (t)] Find the Laplace transform of the function
/
We now take up a theorem which deals with the 0 0≤t≤2
f (t) =
Laplace transform of unit-step functions. It is called (t − 2)3 t ≥ 2.
the Second Translation (or Shifting) Theorem.
Solution We have f (t) = (t − 2)3 U2 (t) so that
6e−2s
12.4.1 Second Shifting Theorem or L{f (t)} = L{(t − 2)3 U2 (t)} = 4
s
Translation on the t-Axis 6
∵ L{t } = 4 .
3
Theorem 12.13 If L{f (t)} = f (s) (s > b) and s
a is any non-negative real number, then Example 12.23
−as Find the Laplace transform of f (t) = t u2 (t)
L{f (t − a)ua (t)} = e f (s).
Solution We have
Proof: We have by definition of Laplace transform
∞ L{tu2 (t)} =e−2t L{t + 2}
L{f (t − a)Ua (t)} = e−st f (t − a)Ua (t)dt =e−2t [L{t} + L{2}]
0 ∞
−2s 1 2
= e−st f (t − a)dt =e +
a s2 s
Example 12.24 Second Shifting Theorem or Translation on

Find the inverse Laplace transforms of the t-axis
e−2s −π s
Theorem 12.14
(i) (ii) 2e 4 /(s2 + 4) (iii) e−3s /s2 .
s+2 L−1 {f (s)} = f (t) ⇒ L−1 {e−as f (s)} = g(t)
/
Solution f (t − a) t > a
1 where g(t) =
(i) Here a = 2, f (s) = we have 0 t < a.
s+2
/ ,
−1 −1 1 Recall the results :
L {f (s)} =L = e−2t
s+2
/ −2s ,
−1 e L−1 {e−as f (s)} = f (t − a)ua (t)
∴L =e−2(t−2) u2 (t)
s+2 and L{ua (t)} = L{1 · ua (t)}
2 e−as
(ii) Here a = +π/4, f (s) = = .
s2 + 4 s
/ , Example 12.26
−1 2
⇒ f (t) =L = sin 2t Find the inverse Laplace transforms of the functions
s2 + 4
π
) e−3s 4e−(sπ/2) e−2s
−1 2e− 4 s π (i) , (ii) , (iii) .
L = sin 2 t − uπ/4 (t) s+2 s2 + 16 s2
s2 + 4 4
π Solution
= sin 2t − uπ/4 (t) (i) We have
2
= − cos 2t · uπ/4 (t) L−1 {e−as f (s)} = f (t − a)ua (t)
1 1
(iii) Here a = 3, f (s) = so that f (t) = t Here a = 3 and f (s) =
s2 s+2
/ , / −3s ,
e−3s e
L −1
=(t − 3)u2 (t). ∴ L−1 = e−2(t−3) u3 (t)
s2 s+2
/ ,
−1 −1 1
Example 12.25 ∵ L {f (s)} = L = e−2t
Find the Laplace transforms of the functions s+2
/
0, 0≤t<3 4
(i) f (t) = (ii) We have a = π/2 and f (s) = so that
(t − 3)2 , t ≥ 3. s2 + 16
/ ,
(ii) f (t) = t 2 U3 (t). 4
L−1 {f (s)} = L−1 2 = sin 4t
s + 42
Solution / −(sπ/2) ,
−1 4e
(i) Writing f (t) as f (t) = (t − 3)2 U3 (t), we have ∴L = sin 4(t − π/2)uπ/2 (t)
s2 + 16
L{f (t)} = L{(t − 3)2 U3 (t)} = sin(4t − 2π)uπ/2 (t)
= 2e−3s /s3 ∵ L{t 2 } = 2/s3 = sin 4t · uπ/2 (t)
(ii) Using the result given above 1
−3s −3s
(iii) Here a = 2, f (s) = . Hence f (t) = t.
L{t u3 (t)} =e
2
L{(t + 3) } = e (t + 6t + 9)
2 2 s2
/ ,
−3s 2 6 9 −1 e−2s
=e + 2+ L = (t − 2)u2 (t).
s3 s s s2
12.5 Unit Impulse function or In this case,

∞ 1/k
Dirac Delta Function
wk (x)dx = (k/2)dx = 1 (Fig. 12.7)
In many problems of physics and electrical and −∞ −1/k
mechanical engineering, the concept of impulsive k
force is encountered. An impulsive force as for (2) wk (k) = (k > 0)
π(1 + k 2 x2 )
instance, in the collision of two bodies may be inter-
preted as a very large force acting for a very short In this case,
∞ ∞
duration of time. To deal with such situations, we k dx
wk (x)dx =
introduce a function called the unit-impulse function −∞ π −∞ (1 + k 2 x2 )
∞
or Dirac2 delta Function. It is called a generalised 2 d(kx)
function, as it is function in the usual sense. We con- =
π 0 (kx) 2 +1
sider a few examples. 2 ∞
Apply unit force by pressing a coin of given thick- = · tan−1 (kx)
π 0
ness over the edge of a metal plate (−∞ < x < ∞). 2 π
We are interested in determining the pressure dis- = · =1
π 2
tribution w(x) introduced along the plate. The exact
shape of w(k) and stress field are determined as part As k → ∞ we get a “point” force of unit strength
of the problem. situated at x = 0. Denoting this force by δ(k) we
define δ(k) = lim wk (x)
n→∞
Consider now another example using the unit-strip
functions. Define δk (t) by
(a) /
O 1/k, 0≤t<k
x δk (t) =
0, Otherwise
which in terms of unit-step function is
1 1
w (x)
δk (t) = [u0 (t) − uk (t)] = [H (t) − H (t − k)]
k k
where H (t) is the Heaviside function.
O x
Figure 12.7
1/k
Pressure distribution shape of the curve w(x).
Since the unit force
is applied we have
∞
w(x)dx = 1
−∞
O k t
To study the general case of application of a
force we consider some specific function w(x) in the
form of sequences of function < wk (x) > called Graph of δk (t)
δ-sequences. For example,
/ let Figure 12.8
k/2, |x| < 1/k
(1) wk (k) =
0, |x| ≥ 1/k
The pulse is of height 1/k and duration k. As
2 Named after the English Physicist Paul Dirac (1902–1984)
k → 0 the amplitude of the impulse tends to infin-
who was awarded the Nobel prize in 1923 for his work in Quantum
Mechanics. ity. Now we define δ(t) = lim δk (t) so that we may
k→0+
interpret that δ(t) = 0 for t = 0 and δ(t) = ∞ at we can obtain its transform as follows:
t = 0. 1
The delta function can be made to act at any point δk (t − a) = [H (t − a) − H (t − a − k)]
k
other than the origin. Then, δ(t − a) acts at t = a and
1
is defined as = [Ua (t) − Ua+k (t)]
/ k
0, for t = a
δ(t − a) = Taking the Laplace transform
∞, for t = a.
1 e−as e−(a+k)s
Taking the limit as k → 0 we have L{δk (t − a)} = −
k s s
H (t) − H (t − k) e−as
δ(t) = lim δ(t) = lim = H (t) = (1 − e−ks )
k→0 k=0 k ks
Taking the limit as k → 0+ we obtain
H (t) is to be understood not as an ordinary deriva-
tives but as a generalised functions. Also, we have 1 − e−ks
H (t − a) = δ(t − a) lim =s (12.15)
k→0+ k
We now prove an important property of Dirac
Hence lim L{δk (t − a)} = L{δ(t − a)}
delta function. k→0+
e−as
Theorem 12.15 (Filtering property of Dirac delta = (s) = e−as (12.16)
s
function) If f (t) is continuous and integrable for
∞ When a = 0 we have
t ≥ 0 then f (t)δ(t − a)dt = f (a).
0
L{δ(t)} = 1 and L−1 {1} = δ(t)
Proof: We have
The above result
∞ (12.16) is obtainable by the fil-
∞
1 (a+k) tering property f (t)δ(t − a)dt = f (a) by taking
f (t)δk (t − a)dt = f (t)dt
k 0 ∞
0
/ a
1/k, a ≤ t < a + k f (t) = e−st since L{δ(t −a)} = e−st δ(t − a)dt =
since δk (t − a) =
∞, Otherwise 0
∞ a+k f (a) = e−as .

1
f (t)δk (t − a)dt = · f (t0 ) · dt
0 k a 12.6 Convolution Functions
1
= · f (t0 ) · [t]a+k
a
We now define an operation called the convolution
k operation on two functions of t that corresponds to
= f (t0 ) for a < t0 < a + k multiplying their transforms together. This operation
gives the inverse transform of the product of two
by the mean value theorem of integral calculus. transforms directly in terms of the original functions.
Taking the limit as k → 0 we obtain
∞ Definition Let f (t) and g(t) be defined for t ≥ 0.
f (t)δk (t − a)dt = f (a) ∵ f (t0 ) → f (a). Then the convolution of f (t) and g(t), denoted by
0
(f ∗ g)(t), is defined by
t
12.5.1 Laplace Transform of Dirac (f ∗ g)(t) = f (u)g(t − u)du (t ≥ 0)
Delta Function 0
The delta function does not satisfy the conditions of The following properties of the convolution oper-
the existence theorem. However, using its definition ation can be easily proved.
Commutative property y
u=t
f ∗g = g∗f
t t
i.e., f (t)g(t − u)du = g(u)f (t − u)du
0 0 t=u
Associative property t→∞
(f ∗ g) ∗ h = f ∗ (g ∗ h) O u=0 t
Distributive property
f ∗ (g + h) = f ∗ g + g ∗ h
Figure 12.9 Region of integration
Scalar multiplication
f ∗ (kg) = k (f ∗ g) Note
Taking L−1 on both sides, we have
Theorem 12.16 If f (t) and g(t) are piece-wise
continuous functions for t ≥ 0 and are of exponential L−1 {f¯ (s)ḡ(s)} = (f ∗ g)(t) = f (t) ∗ g(t).
order then
Example 12.27
L{(f ∗ g)(t)} = L{f (t)}L{g(t)}
Find the convolution t ∗ sin at.
= f¯ (s)ḡ(s) or (f¯ · ḡ)(s)
Solution We have, by definition,
Proof: We have, by definition, t
∞ ∞
t ∗ sin at = sin a(t − u)du
f¯ (s)ḡ(s) = e−su f (u)du · e−su g(v)dv 0
t
∞ ∞
0 0 1 1
= u· cos a(t − u)+1· 2 sin a(t −u)
= e−s(u+v) f (u)g(v)dudv a a 0
∞
0 0
∞ by integration by parts
= f (u)du e−s(u+v) g(v)dv t 1 at − sin at
0 0 = − 2 sin at = .
a a a2
Put u + v = t where u is fixed so that v = t − u
Example 12.28
and the limits for t are t = u and t = ∞
Find the convolution t 2 ∗ cos t.
∴ We have
∞ ∞ Solution
f¯ (s)ḡ(s) = f (u)du e−st g(t − u)dt t
0 u
t ∗ cos t =
2
u2 cos(t − u)du
The region of integration for this double integral 0
t
is 0 ≤ u < ∞, u ≤ t ≤ ∞ (Fig. 12.9) = u2(−sin(t −u))−2u(−cos(t −u))+2 sin(t −u)
Changing the order of integration, we obtain the 0
region of integration as 0 ≤ t < ∞, 0 ≤ u ≤ t. by integration by parts
∞ t = [t 2 · 0 − 2t(−1) + 0 − (0 − 0 + 2 sin t)]
∴ f¯ (s)ḡ(s) = e−st dt f (u)g(t − u)du
0 0 = 2t − 2 sin t.
∞ t
= e−st f (u)g(t − u)du dt Example 12.29
0 ∞ 0 Applying the convolution theorem find the inverse
= −st
e [(f ∗ g)(t)]dt transform of each of the following functions:
0
1 1
= L{(f ∗ ig)(t)}. (i) (ii)
s2 (s2 + a2 ) s(s + 1)(s + 2)
s2 (iii) Let L−1 {f¯ (s)} = f (t) and L−1 {ḡ(s)} = g(t)
(iii) . (VTU 2000, Madras 1999)
(s2 + a2 )(s2 + b2 ) s s
where f¯ (s) = 2 and ḡ(s) = 2
Solution s + a2 s + b2
- . - s .
(i) Let L−1 {f¯ (s)} = f (t) and L−1 {ḡ(s)} = g(t) so that L−1 f¯ (s) = L−1 2
s + a2
1 1 = cos at = f (t)
where f¯ (s) = 2 and ḡ(s) = 2
s s + a2 - . - s .
-1. and L−1 ḡ(s) = L−1 2
so that L−1 2 = t = f (t) and s + b2
s = cos bt = g(t)
- 1 . 1
L−1 2 = sin at = g(t)
s + a2 a By the convolution theorem, we have
By the convolution theorem, we have - s2 .
- . - 1 . L−1 2
L−1 f¯ (s)ḡ(s) = L−1 2 2 (s + a2 )(s2 + b2 )
s (s + a2 ) t
t = cos au cos b(t − u)du
1
= u · sin a(t − u)du 0

0 a 1 t- .
1 1 1 t = cos[(a−b)u+bt]+cos[(a+b)u−bt] du
= 2 · u · cos a(t −u)− 2 · sin a(t −u) 2 0

a a a 0
1 sin[(a − b)u + bt]
by integration by parts =
2 a−b
t 1
= 2 + 3 (0 − sin at) 1 sin[(a + b)u − bt] t
a a + du
at − sin at 2 a+b
= 0

a3 1 sin at − sin bt 1 sin at + sin bt
= +
(ii) We have 2 a−b 2 a+b
-1. - 1 . a sin at − b sin bt
= .
L−1 = 1, L−1 = e−t , a2 − b 2
s s+1
- 1 .
L−1 = e−2t Note
s+2 The above problems are solvable alternatively by the
By the convolution theorem, we have method of partial fractions.
- 1 . t
L−1 = 1 · e−u du = [−e−u ]t0 Example 12.30
s(s + 1) 0 Using the convolution theorem solve the integral
= 1 − e−t equations
t t
- 1 1 . (i) f (t) = cos t + e−t 0 f (u)eu du
L−1 = (1 − e−u ) − e−2(t−u) du t
s(s + 1) (s + 2) (ii) f (t) = at + 0 f (u) sin(t − u)du
0
t t
(iii) f (t) = 1 + t + 2 0 f (u) sin(t − u)du.
= e−2t (e2u − eu )du
0
t Solution
1 2u
= e−2t e − eu (i) The equation can be written
as t
2
f (u)e−(t−u) du
0
f (t) = cos t +
−2t 1 2t 1
=e e −e − +1
t 0
Noting that the integral is the convolution of f (t)
2 2
1 −t 1 −2t and e−t , we can write the equation as
= −e + e f (t) = cos t + f (t) ∗ e−t
2 2
Applying Laplace transform on both sides, we s2 + 1 A B c

Let = + 2+
obtain s (s − 1)
2 s s s−1
s 1 ∴ s2 + 1 = (As + B)(s − 1) + Cs2
f¯ (s) = 2 + f¯ (s) ·
s +1 s+1 Coefficient of s2 ⇒ 1 = A + C
1
∵ L{e−t } = putting s = 0 we get 1 = −B
s+1
∴ B = −1
1 s
⇒ f¯ (s) 1 − = 2 s = 1 we get 2 = c
s+1 s +1
s+1 ∴A=1
⇒ f¯ (s) = 2
s +1 1 1 2
∴ f¯ (s) = − − 2 +
s 1 s s s−1
= 2 + 2
s +1 s +1 Whose solution is f (t) = −1 − t + 2et .
Applying the inverse Laplace transform, we obtain
f (t) = cos t + sin t 12.7 Solution of Initial Value
Problems (IVPs) Using
(ii) The equation can be written as Convolution Theorem
t Example 12.31
f (t) = at + f (t) ∗ sin t Using the convolution theorem solve the IVPs for
0
the following:
Now applying Laplace transform on both sides, we
obtain (i) y (t) + y(t) = cos t, y(0) = y (0) = 0
a 1
f¯ (s) = 2 + f¯ (s) · 2
s s +1 (ii) y (t)+3y (t)+2y(t) = e−t , y(0) = 0 y (0) = −1
1 a
⇒ f¯ (s) 1 − 2 = 2 (iii) y (t) + y(t) = cos t, y(0) = y (0) = 0
s +1 s
a s2 + 1 1 1
⇒ f¯ (s) = 2 · 2
=a 2 + 4 (iv) y (t) + k 2 y(t) = sin at, y(0) = 1, y (0) = 0.
s s s s

whose solution is f (t) = a t + 6 t 1 3
Solution
(iii) The equation can be written as (i) Applying Laplace transforms on both sides of the
t differential equation and using the condition, we get
f (t) = 1 + t + f (t) ∗ sin t s
(s2 + 1)ȳ(s) = 2 where L{y(t)} = ȳ(s)
0 s +1
Taking Laplace transforms on both sides s 1 s
and L{cos t} = 2 or ȳ(s) = 2 · 2
1 1 1 s +1 s +1 s +1
f¯ (s) = + 2 + 2f¯ (s) · 2 / ,
s s s +1 −1 1 −1 s
∵L { 2 } = sin t and L = cos t
2 1 1 s+1 s +1 s2 + 1
or f¯ (s) 1 − 2 = + 2 = 2 ,
s +1 s s s We have
s2 − 1 ¯ s+1 1 s
or 2 f (s) = 2 y(t) = L−1 { 2 · }
s +1 s s + 1 s2 + 1
t
s + 1 s2 + 1
or f¯ (s) = 2 · = sin u cos(t − u)du
s (s + 1)(s − 1) 0
t
s2 + 1
= 2 (s = −1) = sin(t − u) cos udu,
s (s − 1) 0
by commutativity of convolution so that f (t) = L−1 {f¯ (s)}]

1 t = sin 3t and g(t)
= [sin u cos(t − u) + cos u sin(t − u)du]
2 0 = L−1 {ḡ(s)} = cos 3t
1 By the convolution theorem, we have
a = b = (a + b)
2
y(t) = L−1 {f¯ (s) · ḡ(s)}
1 t 1 t
= sin[u + (t − u)]du = t sin t
2 0 2 = f (u)g(t − u)du
0
(ii) Applying the Laplace transforms on both sides t
of the differential equations, we have = sin 3u · cos 3(t − u)du,
0
1
[s2 ȳ(s)−sy(0)−y (0)]+3[sȳ(s)−y(0)]+2ȳ(s) = t
s+1 = cos 3u · sin 3(t − u)du,
0
Using the initial condition, we obtain
1 by commutativity of convolution
(s + 1)(s + 2)ȳ(s) = − 1 or t
s+1 1
= [sin 3u cos 3(t − u) + cos 3u sin 3(t − u)du]
1 1 2 0
ȳ(s) = −
(s+1)2 (s+2) (s+1)(s+2) 1
/ , / , a = b = (a + b)
1 1 2
∵ L−1 = e−t and L−1 = e−2t
s+1 s+2 1 t 1
= sin(3u + 3t − 3u)du = t sin 3t
By the 2 0 2
/ convolution theorem,
, t we have
1 (iv) Applying Laplace transforms and using the
L−1 = e−u e−2(t−u) du
(s + 1)(s + 2) 0
initial conditions, we have
t a
= e−2t eu du = e−2t (et −1) s2 ȳ(s) − s + k 2 ȳ(s) = 2
s + a2
0
s a
= e −t
− e −2t ⇒ ȳ(s) = 2 + 2
/ , / , s +k 2 (s + k )(s2 + a2 )
2
−1 1 −1 1 1
L =L · Case (1) k = a
(s+1)(s+2) (s+1)(s+2) (s+1) - .
t a
−u −2u −(t−u)
L−1 2
= (e − e )e du (s + k )(s + a )
2 2 2
0
t 1 t
= sin ku · sin a(t − u)du,
= e−t (1 − e−u )du k 0
0
by convolution theorem
= e (t + e−t − 1)
−t

1 t
Hence the solution is = {cos[(k + a)u−at]−cos[(k −a)u+at]}du
−t −t −t −2t
2k 0
y(t) = e (t + e − 1) − (e − e ) 1 sin[(k + a)u − at] sin[(k − a)u + at] t
−t
= (t − 2)e + 2e −2t = −
2k k +a k −a 0
(iii) Applying Laplace transforms and using the 1 sin kt + sin at sin kt − sin at
= −
initial conditions, we have 2k k +a k −a

3s 3s 1 k sin at − a sin kt
s2 ȳ(s) + 9ȳ(s) = 2 or ȳ(s) = 2 =
s +9 s +9 k k 2 − a2

3 s 1 k sin at − a sin kt
Let f¯ (s) = 2 and ḡ(s) = 2 ∴ y(t) = cos kt +
s +9 s +9 k k 2 − a2
case (2) k = a 1
(c)
s a (s2
+ a2 )2
ȳ(s) = + 2 sin at − at cos at
+a
s2 2 (s + a2 )2 Ans:
a 1 a a3
L−1 { 2 } = L−1 { 2 · (s2 + a2 )}
(s + a2 )2 a (s + a2 ) s
(d)
1 t (s2 + a2 )2
= sin au · sin a(t − u) du
a 0 Ans: (t sin at/a2 ).
t
1
= [cos a(2u−t)−cos at] du
2a 0 4. Using the convolution theorem solve the
1 sin a(2u − at) t following:
= − u cos at
2a 2a 0
t
1 sin at (a) f (t) = t + 6 f (u)e(t−u) du
= − t cos at 0
2a a Ans: f (t) = (−6 + 7t + 6e7t )/49
1
= 2 (sin at − at cos at) t
2a (b) f (t) = 1 + f (u) sin(t − u)du
0
The solution of the equation is
Ans: f (t) = 1 + 1/2t 2
1
y(t) = cos at + 2 (sin at − at cos at).
2a t
(c) f (t) = et + 0
f (u)(t − u)du
EXERCISE 12.5
Ans: f (t) = (3e + 2tet + e−t /4)
t
1. Show that
t2 t
(a) 1 ∗ 1 = t, (b) 1 ∗ 1 ∗ 1 = , (d) f (t) = 1 + f (t − u)e−2u du
2 0
tn Ans: f (t) = 1 + 2t.
(c) 1 ∗ 1 ∗ ∗ ∗ ∗ ∗ 1 = (n times).
n!
2. Find the following convolutions: 5. Using the convolution theorem, solve the
following IVPs:
(a) 1 ∗ eat
Ans: (eat − 1)/a (a) y (t) − a2 y(t) = 0 (a = 0)
−at
(b) t ∗ eat Ans: y(t) = c1 e + c2 e
at
Ans: (eat − 1 − at)/a2

(b) y (t) + a2 y(t) = b
(c) e−at ∗ eat (a = b)
Ans: −(eat − e−bt )/(a − b). Ans: y(t) = c1 sin at + c2 cos at + b
a2
3. Use the convolution theorem to find the inverse

(c) y (t) + 2y (t) + 2y(t) = 0
Laplace transform of each of the following:
y(0) = 0, y (0) = 1
1 Ans: y(t) = et sin t
(a)
(s + a)(s + b)
Ans: −(eat − e−at )/(a − b) (d) y (t) + 4y(t) = sin t y(0) = 0, y (0) = 0
Ans: y(t) = 1
sin t − 16 sin 2t
1 3
(b) 2 2
s (s + a2 )
(e) y (t) + y(t) = 1 y(0) = 1, y (π/2) = 0
at − sin at
Ans: Ans: y(t) = 1 − sin t.
a3
6. Show that the solution of the system of the f (t)

differential equations
1
x (t) − 2y (t) = f (t)
x (t) − y (t) + y(t) = 0
O
Under the conditions x(0) = x (0) = y(0) a 2a 3a 4a 5a 6a t
(a) Triangular wave
= y (0) = 0 so that f (0) = 0 is
t t
x(t) = f (u)du − 2 f (u) cos(t − u)du f (t)
0 0
t
y(t) = f (u) cos(t − u)du
0
O a 2a 3a 4a 5a t
12.8 Periodic Functions
Introduction
(b) Square wave (Type I)
A function is called periodic if its graph repeats
itself at regular intervals. Such functions occur f (t)
frequently in science and engineering. A function
f (t) is called periodic if there exists a number
p such that f (t + p) = f (t) for all t > 0 and
t ∈ dom f . The smallest positive number p with
this property is called the period of f (t). Notice that O a 2a 3a 4a 5a t
f (t) = f (t + p) = f (t + 2p) = · · · f (t + np).
The simplest examples of periodic functions are
the circular functions. sin t, cos t are periodic with
(c) Square wave (Type II)
p = 2π and tan t, cot t are periodic with p = π. A
few other examples are the following. f (t)
(a) Triangular wave

/
t/a, 0≤t<a
f (t) =
(2a − t)/a, a ≤ t ≤ 2a O a 2a 3a 4a t
f (t + 2a) = f (t), p = 2a
(b) Square wave (Type I) (d) Saw-tooth wave

/
k, 0≤t<a Figure 12.10
f (t) =
−k, a ≤ t ≤ 2a
f (t + 2a) = f (t), p = 2a
(d) Saw-tooth wave
(c) Square wave (Type II)
f (t) = t, 0 ≤ t < a; f (t + a) = f (t), p=a
/
k, 0≤t<a
f (t) = We can easily observe that the triangular wave
0, a ≤ t ≤ 2a
function is continuous while the other three functions
f (t + 2a) = f (t), p = 2a viz. square wave functions (Type I and Type II) and
the saw-tooth wave function are piece-wise continu- Solution By the theorem on Laplace transforms
ous. The Laplace transform of a periodic function is of periodic functions, we have
obtained by integrating it over a single period. p
1
12.8.1 Laplace Transform of (1 − e−sp ) 0
Periodic Functions The given function f (t) is periodic with p = T so
Theorem 12.17 (Theorem of Laplace transforms that
p −st
of a Periodic function) If f (t) is piece-wise continu- e f (t)dt
ous for t ≥ 0, is of exponential order and is periodic L{f (t)} = 0
1 − e−sp
with period p then T
p −st 1 −st t
= e dt
e f (t)dt 1 − e−st 0 T
L{f (t)} = 0 (s > 0).
1 − e−sp 1 1 e−st e−st T
= t · − 1 · dt
Proof: By definition 1 − e−sT T s (−s)2 0
∞
1 1 Te−st − 0 1 −st
L{f (t)} = e−st f (t)dt = − (e − 1)
1 − e−sT T −s s2
0
p 2p −sT
1 e 1 − e−sT
= −st
e f (t)dt + e−st f (t)dt + · · · = − +
0 p
1 − e−sT s Ts2
(r+1)p 1 1 e−sT
+ e−st f (t)dt + · · · = 2 −
sT s 1 − e−sT
rp
∞ 1 1 (1 − e−sT ) − (1 + e−sT )
% (r+1)p = 2 + ·
= e−st f (t)dt sT 2s 1 − esT
r=0 rp 1 1 e sT /2
+ e−sT /2
= 2 + 1 − sT /2
Put t = u + rp ⇒ dt = du. The limits are sT 2s e − e−sT /2
1 1
t = rp ⇒ u = 0, t = (r + 1)p ⇒ u = p = 2 + [1 − coth(sT /2)].
sT 2s
and note that f (u + rp) = f (u) for all r
∞ Example 12.33
% p
The square wave (or meander) function
= e−rsp e−st f (u)du /
r=0 0 1, 0 < t < a/2
f (t) =
p
−1, a/2 < t < a
= (1 − e−sp )−1 e−st f (t)dt
0 and f (t + a) = f (t). [JNTU 1999]
% 1
−rsp −sp −2sp
∵ e =1+e +e + ··· = ; Solution By the theorem on Laplace transforms
1 − e−sp of a periodic function f (t) with period p
and replacing the dummy variable u by t.
1 p −st
L{f (t)} = e f (t)dt
2 0
12.8.2 Examples
Find the Laplace transform of each of the following Hence f (t) is periodic with period p = a
a
periodic functions: 1 a/2 −st
L{f (t)} = −as
e ·1dt + e−st ·(−1)dt
1−e 0 a/2
Example 12.32
1 e−st a/2 e−st a
The saw-tooth wave function given by = +
t 1 − e−as −s 0 s a/2
f (t) = for 0 < t < T 1 1−e −as/2
e−as − e−as/2
T = +
and f (t + T ) = f (t) with p = T 1−e −as s s
1
= (1−e −as/2
)−e−as/2
(1−e−as/2
) Example 12.35
s(1−e−as ) The periodic function
(1 − e−as/2 )2 /
= sin at, 0 < t < π/a
s(1 − e−as/2 )(1 + e−as/2 ) f (t) = .
0, π/a < t < 2π/a
∵ 1 − e−as = 1 − (e−as/2 )2
(Madras 2003, Coimbatore 1999)
= (1 − e−as/2 )(1 + e−as/2 )
1 1 − e−as/2 1 eas/4 − e−as/4 Solution By the theorem on Laplace transforms
= = · of periodic function f (t) with period p, we have
s 1 + e−as/2 s eas/4 + e−as/4
Multiplying the numerator and denominator by eas/4 , p
1
1 as L{f (t)} = e−st f (t)dt
= + tanh . 1 − e−sp 0
s 4
Example 12.34 In the present case, f (t) is periodic with period 2π/a
The triangular wave function defined by
L{f (t)}
/ 2π/a 2π/a
t, 0<t<a 1
f (t) = = −st
e sin atdt + −st
e ·0dt
2a − t, a < t < 2a 1−e−2πs/a 0 π/a
and f (t + 2a) = f (t) 1 e−st (−s sin at − a cos at) π/a
=
(VTU 2003s, Machas 2000 PT Delhi 1997) 1 − e−2πs/a s 2 + a2 0
−πs/a
ae +a
Solution The given function is of period 2a. =
By the theorem on Laplace transforms of periodic (s2 + a2 )(1 − e−2πs/a )
functions, we have a(1 + e−πs/a )
=
1 2a (s2 + a2 )(1 − e−πs/a )(1 + e0πs/a )
−st
L{f (t)} = e f (t)dt a
1 − e−2as 0 =
2a (s + a )(1 − e−πs/a )
2 2
1 a −st
= e ·tdt + e−st · (2a − t)dt a eπs/2a
1−e −2as = ·
-0 a
a2 + s2 eπs/2a − e−πs/2a
1 e−st e−st .a
= t · − 1 a 1 (eπs/2a −e−πs/2a )+(eπs/2a +e−πs/2a )
1 − e−as −s (−s)2 0 = · ·
- s +a 2
2 2 eπs/2a −e−πs/2a
e −st
e−st .2a a πs
+ (2a − t) · − (−1) = 1 + coth .
−s (−s)2 a 2(s + a )
2 2 2a
−as
1 ae − 0 1
= − 2 (e−as − 1) Example 12.36
1−e −2as −s s Using the theorem on Laplace transforms of periodic

ae−2as − ae−as e−2as − e−as functions verify that L{cos at} = 2
s
+ + s + a2
.
−s s2
Solution The function f (t) = cos at is peri-
1 1 −as −as −as
= · 1 − e + e (e − 1) odic with period 2π/a. By the theorem on Laplace
1 − e−2as s2 transform of periodic functions with period p, we
1 1 have
= · (1 − e−as )2
1 − e−2as s2 p
1 1 − e−as 1 eas/2 − e−as/2 1
= 2 = · L{f (t)} = e−st f (t)dt
s 1 + e−as s2 eas/2 + e−as/2 1 − e−sp 0
2π/a
1 as 1
= 2 tanh . = e−st cos atdt
s 4 1 − e−2πs/a 0
1 e−st 2π/a
= (−s cos at + a sin at) 12.9 Inverse Transforms—By The
1 − e−2πs/a s2 + a2 0 Method of Partial Fractions
1 −s −2πs/a s We have found Laplace transforms of some elemen-
= (e − 1) = 2 .
1 − e−2πs/a s2 + a2 s + a2 tary functions using the definition. The use of prop-
EXERCISE 12.6 erties such as linearity, first shifting, change of scale
etc. allow us to extend the list. In all these cases,
1. Find the Laplace transform of the periodic function inverse transforms of the corresponding functions
f (t) given by have been given already. To reinforce the knowledge,
/ we state below the formulas for the inverse trans-
1, 0<t<a
f (t) = and f (t + 2a) = f (t). forms and work out a few examples.
0, a < t < 2a
But first we observe that L{f (t)} is a rational
q algebraic function. We know how to express such
Ans: f¯ (s) =
s(1 + e−sa ) a function into partial fractions. If each fraction is
linear then inverse Laplace transform of such a func-
2. Find the Laplace transform of the periodic function
tion could be found by Heaviside’s partial fraction
g(t) given by
expansion theorem. But in the general case, we can
/
1, 0<t<2 put the given function into partial fractions which
g(t) = and g(t + 4) = g(t). fall into one or the other of the standard forms and
−1, 2 < t < 4
then inverse Laplace transforms could be found, as
1 illustrated below.
Ans: ḡ(s) = tanh s
s Example 12.37
3. Find the triangular wave function f (t) defined by Find the inverse Laplace transforms of the following
/ functions:
t/a, 0≤t≤a
f (t) = 3s2 + 4s + 2 (s2 + 1)(s − 1)
(2a − t)/a, a ≤ t ≤ 2a (i) (ii)
s3 s4
and f (t + 2a) = f (t). Find its Laplace Transform. s+5 (s + 1)
(iii) 2 (iv) 3
s − 3s + 2 s (s − 1)
1 as (2s2 −6s+5)(s3 −6s2 +11s−6)
Ans: f¯ (s) = tanh (v) (Andhra, 1999)
as2 2 s4
4. Find the Laplace transform of the periodic 2s + 3 3s − 4 1
(vi) 2 (vii) (viii) e−1/s .
function defined by the saw-tooth wave s +9 (s − 3)(s − 4)
2 s
Solution
f (t) = t, 0 ≤ t ≤ a and f (t + a) = f (t).
- 3s2 + 4s + 2 . -1. -1.
(i) L−1 = 3L−1 + 4L−1 2
1 a as s3
Ans: f¯ (s) = 2 + 1 − coth s
- .
s
s 2s 2 −1 1
+ 2L , by linearity
5. Find the Laplace transform of full wave rectifica- s3
tion of f (t) = | sin t| given by = 3 + 4t + t 2
/ - (s2 + 1)(s − 1) . - 3 .
sin at, 0 < t < π/a −1 s − s + s − 1
2
f (t) = and p = 2πa (ii) L−1 = L
− sin at, π/a < t < 2π/a s4 s4
-1. -1. -1. -1.
a πs = L−1 − L−1 2 + L−1 3 − L−1 4 ,
Ans: f¯ (s) = coth s s s s
s 2 + a2 2a
By linearity
6. Using the Laplace transform of a periodic function
a 1 1
verify that L{sin at} = 2 . = 1 − t + t2 − t3
s + a2 2 6
(iii) Putting the function into partial fractions, 1 t 5

=e − e2t + e3t
we have 2 2
- . - s . - 3 .
s+5 A B −1 2s + 3 −1 −1
= + (vi) L = 2L + L
s − 3s + 2
2 s−1 s−2 s2 + 9 s2 + 9 s2 + 9
s+5 s+5 = 2 cos 3t + sin 3t
where A = = −6; B = =7
s − 2 s=1 s − 1 s=2 (vii) We have
- s+5 . - 1 . - 1 .
L−1 2 = −6L−1 +7L−1 3s − 4 3(s − 3) + 5
s −3s+2 s−1 s−2 =
= 6e + 7e
t 2t (s − 3)(s − 4)
2 (s − 3)(s2 − 4)
3 1
(s + 1) = 2 +5
(iv) First we write the function as the sum (s −4) (s−3)(s−2)(s+2)
s3 (s − 1)
of partial fractions Now
(s + 1) (s + 1) + 2 1 2 1 A B C
= 3 = 3+ 3 = + +
s3 (s − 1) s (s − 1) s s (s − 1) (s−3)(s−2)(s+2) (s−3) (s−2) (s+2)

1 2 s3
= 3 − 3 1 + s + s2 + ,
s s 1−s 1 1
by actual division where A = = ;
(s − 2)(s + 2) s=3 5
1 2 2 2 2 1 1
= 3− 3− 2− + B= =− ;
s s s s s−1 (s − 3)(s + 2) s=3 4
−1 2 2 2 1 1
= 3 − 2− + C= =
s s s s−1 (s − 3)(s − 2) s=−2 20
- s+1 . - 1 . -1.
Now L−1 3 = 2L−1 − 2L−1 - 3s − 4 .
s (s − 1) s−1 s L−1
-1. -1. (s − 3)(s2 − 4)
− 2L−1 2 − L−1 3 3 - 2 . - 1 .
s s = L−1 2 + L−1
1 2 2 s −4 s−3
= 2e − 2 − 2t − t
t
5 −1 - 1 . 1 −1 - 1 .
2 − L + L
(v) Here the denominator is s − 6s + 11s − 6 =
3 2 4 s−2 4 s+2
(s − 1)(s − 2)(s − 3). The given function is 3 5 1
= · sin 2t + e3t − e2t + e2t
(2s2 − 6s + 5) 2 4 4
f¯ (s) =
(s − 1)(s − 2)(s − 3) (viii) We have
A B C
= + + %∞
−1 %∞
(−1)n 1
s−1 s−2 s−3 1
e− s =
n
/n! =
1 s n! sn
where A = (s − 1)f¯ (s)|s=1 = ; n=0 n=0
2 %∞
1 −1 (−1)n 1
B = (s − 2)f¯ (s)|s=2 = −1; ∴ e s =
s n=0
n! sn+1
5
C = (s − 3)f¯ (s)|s=3 = / , % ∞
(−1)n −1
/ ,
2 −1 1 −1/s 1
- . 1 - 1 . - 1 . and L e = L
−1 ¯ −1 s (n!)2 sn+1
∴ L f (s) = L − L−1 n=0
2 s−1 s−2 %
∞
5 −1 - 1 . (−1)n
+ L = tn
2 s−3 n=0
(n!)2
/ , /
12.9.1 First Shift Theorem −1 s 1 −1 1
L = L
L−1 {f¯ (s)} = f (t) ⇒ L−1 {f¯ (s − a)} = eat f (t) s + 4a4
4 4a s − 2as + 2a2
2
,
1
Example 12.38 − 2
s+3 s s + 2as + 2a2
(i) (ii) / ,
s2 − 4s + 7 (s + a)2 + b2 1 1 −1 a
= · L
a s 4a a (s − a)2 + a2
(iii) (iv) / ,
(s + a)3/2 s4 + 4a4 1 1 −1 a
− · L
4a a (s + a)2 + a2
Solution 1 1
(i) Here = · eat sin at
4a a
s+3 s+3 1 1 −at
= − · e sin at
s2 − 4s + 7 (s2 − 4s + 4) + 3 4a
ata −at
(s + 2) + 5 1 e −e
= √ = 2 sin at
(s − 2)2 + ( 3)2 2a 2
/ , / ,
s+3 s−2 1
L−1 2 =L−1 √ = 2 sinh at sin at.
s − 4s + 7 (s − 2)2 + ( 3)2 2a
√ )
5 −1 3
+√ L √ 12.9.2 Change of Scale Property
3 (s−2)2 +( s)2
√ √ 1
L−1 {f¯ (s)} = f (t) ⇒ L−1 {f¯ (ks)} = f (t/k)
=e2t cos 3t + e2t sin 3t k
√ √
=e2t (cos 3t + sin 3t)
Example 12.39
(ii) Here 4 1 8
(i) (ii) (iii)
s (s + a) − a 9s2 + 16
/ , 16s 4 −1
√ √ 81s 4 − 16
−1/s
= −1 e cos 2 t
(s + a)2 + b2 (s + a2 + b2 ) (iv) If L √ = √ , then show that
s+a a b / −a/s , 3 √ πt
= − · e cos 2 at
(s + a) + b
2 2 b (s + a)2 + b2 L−1 √ = √ (a > 0).
/ , / , s πt
−1 s −1 s+a
∴L =L
(s + a)2 + b2 (s + a)2 + b2 Solution / ,
/ , a
− L
a −1 b (i) We know that L−1 2 = sin at. Taking
b (s+a)2 +b2 / s +, a2
4
a
=e−at · cos bt e−at sin at a = 4, we have L−1 2 = sin 4t = f (t)
b s + 16
1
(iii) Here By change of scale property, L−1 {f¯ (ks)} = f (t/k)
k
/ , * ) / ,
−1 1 −at −1 1 −at t −1 4 1 4
L =e L =e ·2 ∴L = sin t
(s + a) 3/2 s 3/2 π (3s) + 16
2 3 3
*
t (ii) We know that
=2 · eat / , / , / ,
π −1 1 1 −1 1 1 −1 1
L = ·L − ·L
(iv) Here the denominator is s4 −1 2 s2 −1 2 s2 +1
s4 + 4a4 = (s2 + 2a2 )2 − (2as)2 = (s2 + 2as + 1
= (sinh t − sin t)
2a2 )(s2 − 2as + 2a2 ) 2
By change of scale property Example 12.40

s s−2
1 (i) (ii)
L {f¯ (ks)} = f (t/k) Here k = 2
−1 (s2
+a2 )2 (s2
− 4s + 5)2
/ , k s−a 1 s 2 + b2
1 1 (iii) log (iv) log 2 .
∴ L−1 = f (t/2) s−b 2 s + a2
(2s)4 − 1 2
Solution
1 1
= · (sinh t/2 − sin t/2) (i) We know that
2 2 / ,
1 −1 1
= (sinh 4/2 − sin t/2) L = sin t and
4 s 2 + a2
/ ,
d 1 −2s
(iii) We know that = 2
/ , ds s + a2 2 (s + a2 )2
a3 / , / ,
L−1 {f¯ (s)} =L−1 ∴L −1 s
=L −1 1 d
− ·
1
s 4 − a4 (s2 + a2 )2 2 ds s2 + a2
/ ,
1 a a 1
= L−1 2 − = t sin t
2 s − a2 s 2 + a2 2a
1
= (sinh at − sin at) = f (t)
1 / , / ,
(s − 2) s
/ , 2 / , 2 (ii) L−1 = e 2t −1
L
8 23 [(s − 2)2 + 1]2 (s2 + 1)2
∴L −1
=L −1 / ,
81s4 − 16 (3s)4 − 24 1 d 1
= e2t L−1 −
1 2 ds s2 + 1
=L−1 {f¯ (3s)} = f (t/3) 1 1
3 = e2t · t sin t = e2t · t sin t
by change of scale property 2 2
1 2t
1 1 2 2t = te sin t
= · sinh t − sin 2
2 3 3 3
(iii) Let
1 2t 2t
= sinh − sin
6 3 3 L−1 {f¯ (s)} =f (t) so that L−1 {f¯ (s)} = −tf (t)
s−a
(iv) In the given result replace s by s/a and use change If we take f¯ (s) = log
of scale property, s−b
1 1
we have f¯ (s) = −
L−1 {f¯ (s/a)} = a f (at) s−a s−b
/ −a/s , √ d
∴L −1
√
e
=a· √
cos 2 at and − tf (t) = L−1 {f¯ (s)}
s/a πt ds / ,
/ −a/s , √ 1 1
e cos 2 at =L−1 − L−1
⇒ L−1 √ = √ . s−a s−b
s πt
=eat − ebt

12.9.3 Inverse Transforms of −1 s−a eat − ebt
Hence f (t) =L log =
Derivatives s−b t
e −e
bt at
=
L−1 {f¯ (s)} = f (t) t
⇒ (i) L−1 {f¯ (s)} = −tf (t); (iv) Let L−1 {f¯ (s)} = f (t) so that
(ii) L {f−1 ¯ (n) (s)} = (−1) t f (t)
n n
L−1 {f¯ (s)} = −tf (t)
1 s 2 + b2 (ii) Denote the given function by ḡ(s) so that

If we take f¯ (s) = log 2 s+a
2 s + a2 ḡ(s) = log
1 1 s+b
= |n(s2 + b2 ) − (s2 + a2 )
2 2 d 1 1
¯ s s Let f¯ (s) = − ḡ(s) = −
then f (s) = 2 − 2 ds s+b s+a
s + b2 s + a2
- s . / 2 ,
s 1 1
∴ −tf (t) = L−1 {f¯ (s)} = L−1 + 2 −L−1 + 2 We observe that f¯ (u) = −
s b s a u+b u+a
= cos t − cos at Integrating this between s and ∞ we have
/ ,
−1 −1 1 s 2 + b2 ∞
u+b ∞ s+a
⇒ f (t) = L L log 2 f¯ (u)du = log = log (1)
2 s + a2 s u+a s s+b
cos bt − cos at −1 ¯
But f (t) =L {f (s)}
= / ,
−t 1 1
cos at − cos bt =L−1 −
= . s+b s+a
t
=e−bt − e−at (2)
12.9.4 Inverse Transforms of k +g
∴ log → 0 as k → ∞
Integrals of the form k +a
∞
¯ (s)ds
f
s Taking L−1 on both sides of Eq. (1)
/ ∞ ,
−1 ¯ −1 ¯ f (t) / , / ∞ ,
L {f (s)} =f (t) ⇒ L f (s)ds = (s > b) s+a
t L−1 log =L−1 f¯ (u)du
/ ∞
s
, s+b s
or f (t) =t L−1 f¯ (u)du f (t) e−bt − e−at
s = = (3)
t t
Example 12.41 f (t) sinh at
s s+a 1 s+a (iii) Let = so that f (t) = sinh at
(i) cot−1 (ii) log (iii) log t t
a s+b 2 s−a
1 s f¯ (s) =L{f (t)} = L{sinh at}
(iv) (v) a
s(s + 4)2 (s2 − 4)2 = 2
s s s − a2
(vi) a cot−1 − b cot−1 .
a b
∞ ∞
du
Solution Now f¯ (u)du =a 2 − a2
(i) We know that u
s s

a 1 ∞ 1 1
L{f (t)} =L{sin at} = 2 = f¯ (s) =a· − du
s + a2 2a s

u−a u+a

/ , ∞
f (t) 1 u−a ∞
and L = f¯ (u)du = n
t 2 u+a s
∞
s
∞
1 1 u 1 s−a
=a du = a. tan−1 = 0 − log
s u +a
2 2 a a s
2 s+a
π −1 s −1 s 1 s+a
= −L = cos t = log
2 a a 2 s−a
- s . f (t) sin at s−a
∴ L−1 cot−1 = = ∵ log → 0 as k → ∞
a t t k +a
/ ,
(iv) Denote the function by ḡ(s) so that −1 s 1
⇒ f (t) = L = sinh 2t
(s − 4)2
2 4
1
ḡ(s) =
s(s + 4)2 (vi) Let f (t) = sin at − sin bt then

1 1 1 4
= − − (4) f¯ (s) =L{f (t)} = L{sin at − sin bt}
16 s s + 4 (s + 4)2
a b
when we put it into partial fractions note that = − 2
s2 +a 2 s + b2

1 1 1 1 / , / ,
= − f (t) sin at − sin bt
s(s + 4) 4 s s+4 Now L =L
t t
1 1 1 1 ∞
⇒ = −
s(s + 4)2 4 s(s + 4) (s + 4)2 = f¯ (u)du

1 1 1 4 ∞
s
∞
= − − du du
16 s s + 4 (s + 4)2 =a − b
u 2 + a2 u 2 + b2
d s s
Let f¯ (s) = − ḡ(s) (5) −1 u −1 u ∞
ds
=a tan − b tan
1 1 1 8 a b s
= − − π −1 s π s
16 s 2 (s + 4) 2 (s + 4)2 =a −tan −b −tan−1
2 a 2 b
Now, integrating Eq. (5) from s to ∞ −1 s −1 s
=a cot − b cot
∞ ∞ a b
d ∞ - .
f¯ (u)du = − ḡ(u)du = −ḡ(u) s s sin at − sin bt
s s du s
∴ L−1 a cot−1 − b cot−1 =
a b t
1
= , by Eq. (4)
s(s + 4)2
12.9.5 Multiplication by s
Therefore, we have
L−1 {f¯ (s)} = f (t)
f (t) = L−1 {f¯ (s)}
/ , ⇒(i) L−1 {sf¯ (s)} = f (t) if f (0) = 0
1 −1 1 1 1
= L − − , (ii) L−1 {sn f¯ (s)} = f (n) (t) if f (n) (0) = 0
10 s2 (s + 4)2 (s + 4)3
1
= (t − te−4t − 4t 2 e−4t ) Multiplication by s amounts to differentiation w. r. t
/
16
, ‘t’ of f (t) for r = 0, 1, 2, · · · (m − 1).
−1 1 f (t) 1
∴L = = (t −te−4t −4t 2 e−4t )
s(s+4) 2 t 16t Example 12.42
1 s s(a2 − b2 )
= (1 − (1 + 4t)e−4t ) (i) (ii)
16 s + 16
2 s4 − (a2 + b2 )s2 + a2 b2
/ , s s
s (iii) 2 (iv)
(v) Let L−1 = f (t) so that s + a2 (s − a)(s − b)
(s2 − 4)2
/ ∞ , / , ∞ Solution
f (t) −1 udu −1 −1
=L =L 1
t s (u2 −4)2 2(u2 − 4) s (i) Let f¯ (s) = .
/ , / , s2 + 16
1 −1 1 1 1 −1 1 1
= L = · L − 1 4
2 s2 −4 2 4 s−2 s+2 Then f (t) =L−1 {f¯ (s)} = L−1 { 2 }
4 s + 42
1 1 1 1
= (e2t − e−2t ) = sinh 2t = sin 4t and f (0) = sin 0 = 0
8 4 4 4
s s / ,
∴ L−1 { } =L−1 {s · 2 } L−1
s
=
1 d at
(e − ebt )
s2 + 16 s + 16 (s − a)(s − b) a − b dt
=L−1 {s · f¯ (s)} = f (t) 1
1 d = (aeat − beat ).
= · (sin 4t) = cos 4t a−b
4 dt
(ii) We have 12.9.6 Division by s
)
s4 − (a2 + b2 )s2 + a2 b2 = (s2 − a2 )(s2 − b2 ) −1 −1 f¯ (s) t
L {f (s)} = f (t) ⇒ L = f ((u)du
s
∴ The given function 0
Division by s amounts to integration of f (t) w.r.t. ’t’.

s(a2 − b2 )
f¯ (s) =
s4 − (a2 + b2 )s2 + a2 b2 Example 12.43
s(s2 −b2 )−s(s2 −a2 ) s s 1 s
= = 2 2− 2 2 (i) (ii)
(s2 −a2 )(s2 −b2 ) s −a s −b s(s2 + 4) s2 (s + 2)
1 1
We note that (iii) (iv) 2
/ , s(s + 4) (s + a2 )2
1 1
L−1 2 = sinh at and sinh 0 = 0 Solution
s −a 2 a
/ , (i) We know that
1 1 / ,
and L−1 2 = sinh bt and sinh 0 = 0 1 1
s −b 2 b L−1 2 = sin 2t
/ , s +4 2
−1 s(a − b2 )
2
/ ,
∴L −1 1 1 t
s4 − (a2 + b2 )s2 + a2 b2 ∴L = sin 2udu
/ , s(s2 + 4) 2 0
1 1
=L −1
s· 2 −s· 2 −1
s − a2 s − b2 = [cos 2u]t0
4
d 1 d 1 1
= ( sinh at) − ( sinh bt) = (1 − cos 2t)
dt a dt b 4
= cosh at − cosh bt
Aliter:
/ , / , / ,
1 1 1 1 −1 1 s
(iii) We know that L−1 2 = sin at and L −1
= L − 2
s +a 2 a s(s + 4)
2 4 s s +4
sin 0 = 0. On multiplying by s
/ , / , 1
s 1 = (1 − cos 2t)
L −1
=L −1
s· 2 / , t4
s 2 + a2 s + a2 1
(ii) L−1 = e−2u du
d sin at s(s + 2)
= = cos at t
0
dt a 1 1
= − e−2u = (1 − e−2t )
(iv) We know that 2 0 2
/ , / , t
1 1
L −1 1 L−1 2 = (1 − e−2u )du
(s − a)(s − b) s (s + 2) 2 0
/ , / ,
1 1 1 1 1 −2u t
= L −1
−L −1 = u+ e
a−b s−a s−b 2 2
0

1 1 1 −2t− 1
= (eat − ebt ) → 0 as t → 0 = t+ e 2
a−b 2 2
1 EXERCISE 12.7
= (2t − 1 + e−2t )
4
1. Find the inverse Laplace transforms of the follow-
(iii) We know that ing function:
/ ,
−1 ¯ −1 1
L {f (s)} = L = e−4t = f (t) s2 − 5s + 6
s+4 (a)
s3
By the property of derivatives t2
/ , Ans: 5t + 6 ·
1 2!
L−1 {f¯ (s)} = L−1 − = tf (t) = te−4t
(s + 4)2 s(s2 − 2)2
(b)
Now, by division property 2s5
/ , t 3 1
1 Ans: − 3t 2 + t 4
L−1 = − ue−4u du 2 4
s(s + 4)2 0
−4u t π
ue 1 (c)
=− − e −4u s2 + π 2
−4 (−4)2 0 Ans: sin πt
t −4t
= e + (e − 1)
1 4t s+3
(d)
4 16 (s − 1) + (s + 2)
1
= [−1 + (1 + 4t)e−4t ] 1
16 Ans: [4et − e−2t ]
3
(iv) We know that
/ , 2 6 5
1 1 (e) + 2−
−1 ¯ −1 s3 s s
L {f (s)} = L = sin at = f (t)
s +a
2 2 a Ans: t + 6t − 5
2
By the property of derivatives s

(f)
L−1 {f¯ (s)} = − tf (t) (s + 1)(s + 2)(s − 3)
/ , 1 2 3 3
−1 −2s t Ans: et − e−2t + et
⇒L = − sin at 4 5 20
(s2 + a2 )2 a
/ ,
−1 s t 2. First shift:
⇒L = sin at
(s2 + a2 )2 2a s
(a)
By the property of division s −5
2
√
/ , / , Ans: cosh t
1 −1 1 s
L−1 = L · s
(s2 + a2 )2 s (s2 + a2 )2 (b)
t (s + a)2 + b2
1 a
= u sin audu Ans: e−at (cos bt − sin bt)
2a 0 b

1 − cos au sin au t s+5
= u + (c)
2a a a2 0 s2 + 8s + 25

1 t 1 1
= − cos at + 2 sin at Ans: e−4t (cos 3t + sin 3t)
2a a a 3
/ ,
1 1
⇒ L−1 (d)
(s2 + a2 )2 (s2 + 2s + 5)2
1 e−t
= 3 (sin at − at cos at) Ans: (sin 2t − 2t cos 2t)
2a 16
3. Change of scale: 6. Multiplication of s:

1 2
(a) s2
2s − 5 (a)
1 5t s−3
Ans: e 2 Ans: 3e34 (2 + 3t)
2
s
3s − 8 (b)
(b) 2s2 −1
4s2 + 25 1 √
3 5t 4 5t Ans: cosh(t/ 2)
Ans: cos − sin 2
4 2 5 2
4. Inverse transforms of derivatives: s+3
(c) [JNTU 2005S (Set 1)]
s+ (s2 + 6s + 13)2
(a) 1 −3t
s2 + 2s + 2 Ans: e cos 2t
1 4
Ans: e−t t sin t
2 s
(d) 2 [JNTU 2005 (Set 4)]
1 s2 + 7 s − a2
(b) |n 1
2 s2 + 3 Ans: cosh at
1 √ √ a
Ans: [cos 7t − cos 3t]
t 7. Division by s:
s(s + 1) s2 + 2
(c) log 2 (a)
s +4 s(s2 + 4)
1
Ans: (cos 2t − e−t − 1) 1
t Ans: (1 + cos 2t)
2
s+1 s
(d) log (b) [JNTU 2006 (Set 1, 3)]
s−1 (s62 + 4)2
2 1
Ans: sinh t
t Ans: sin 2t
4
5. Inverse transforms of integrals:
/ ∞ , 1
1 1 (c) [JNTU 1995, 2001]
(a) Evaluate L−1 − du s(s2 + a2 )
s u u+1 1
[Hint: L−1 1s − s+1 1
= 1 − e−t ] Ans: 2 (1 − cos at)
−t
a
1−e
Ans: 1
t (d) [JNTU 200]
/ ∞ , s2 (s2
+ a )
2

u+2 1 1
(b) Evaluate L−1 |n du Ans: 2 t − sin at
u+1
−1
u+2s e−t −e−2t a a
[Hint: L |n u+1 = t
]
e−t − e−2t 12.10 Solution of Linear
Ans:
t2 Differential Equation by
1 the Method of Laplace
(c) Transforms
s(s + 1)
Ans: 1 − e−t To solve a linear differential equation with constant
% n
1 coefficients of the form ar y(r) (t) = Q(t), we
(d)
(s + 1)2 r=0
Ans: te−t proceed as follows:
1. Take the Laplace transform on both sides of the Substituting the values of y(0) and y (0), we obtain
equations using the formulas for derivatives and the
given initial conditions. (s2 + 4)ȳ(s) = s + 6
s+6 s 6
or ȳ(s) = 2 = 2 +
2. Transpose to the right terms with negative sign s +4 s + 4 s62 + 4
(i.e., terms not involving ȳ).
Applying L−1 on both sides
3. Divide both sides by the coefficient of ȳ and y(t) =L−1 {ȳ(s)}
Q̄(s) / , / ,
obtain ȳ(s) = with partial fractions and take =L −1 s
+ L
6 −1 2
, by linearity
p̄(s) s2 + 4 2 s2 + 4
the inverse transforms on both sides.
y(t) = cos 2t + 3 sin 2t
Q̄(s)
4. Put the rational function into partial frac- (ii) Taking Laplace transform of the given differential
P̄(s) equation
tions and take the inverse transforms on both sides.
[s2 ȳ(s) − sy(0) − y (0)] + 2[sȳ(s) − y(0)] + 5ȳ(s)
Example 12.44
1
Solve the following initial value problems by the =
application of Laplace transform method: (s + 1)2 + 1
(i) y + 4y = 0 y(0) = 1, y (0) = 6 ⇒ [s2 ȳ(s) − s · 0 − 1] + 2[sȳ(s) − 0] + 5ȳ(s)
1
= 2
(ii) y + 2y + 5y = e−t y(0) = 0, y (0) = 1 s + 2s + 2
[JNTU 2003, 2006S (Set 2)]
1
⇒ (s2 + 2s + 5)ȳ(s) = 1 +
(iii) (D + 1)x = t cos 2t given that x = 0, Dx = 0 at
2 + 2s + 2
s2
d s2 + 2s + 3
t = 0 where D = [JNTU 2002S] = 2
dt s + 2s + 2
d 2x dx
(iv) 2 + 2 + x = 3te−t given that x(0) = 4, s2 + 2s + 3
dt dt ⇒ ȳ(s) =
dx (s2 + 2s + 2)(s2 + 2s + 5)
= 0 at t = 0 [JNTU 2003]
dt 1 (s2 + 2s + 5) + 2(s2 + 2s + 2)
= ·
(v) 4y + π2 y = 0, y(0) = 2, y (0) = 0 3 (s2 + 2s + 2)(s2 + 2s + 5)

[JNTU 2001] 1 1 2
= + ,
3 s2 + 2s + 2 s2 + 2s + 5
d 2y dy resolving in to partial fractions
(vi) 2
+ 2 − 3y = sin t
dt dt
y=
dy
= 0 when t = 0 [JNTU 2000S] Applying L−1 on both sides
dt / , / ,
1 −1 1 −1 1
Solution
y(t) = L +2L
3 s2 +2s+2 s2 +2s+5
(i) Taking Laplace transform of the given differential / , / ,
1 −1 1 −1 1
equation, we obtain = L +2L
3 (s+1)2 +1 (s+1)2 +4
/ , / ,
1 −t −1 1 2 −1 2
[s2 ȳ(s) − sy(0) − y (0)] + 4ȳ(s) = 0 = e L + L
3 s2 + 1 2 s2 + 4
−t
where ȳ(s) = L{y(t)} e
= (sin t + sin 2t) by first shift theorem
3
(iii) Taking Laplace transform of the given differen- 3

⇒ (s + 1)2 x̄(s) = + 4s · ts
tial equation, we obtain (s + 1)2
3 4(s + 1) 4
d s ⇒ x̄(s) = · + +
[s x̄(s)−s(x(0))−x (0)]+ x̄(s) = (−1)
2
(s + 1)4 (s + 1)2 (s + 1)2
ds s2 + 4
(by the property of multiplication by t) Applying L−1 on both sides
/ , / ,
s2 − 4 −1 1 −1 1
⇒ (s2 + 1)x̄(s) = x(t) = 3L + 4L
(s2 + 4)2 (s + 1)4 s+1
/ ,
s2 − 4 1
⇒ x̄(s) = + 4L−1
(s62 + 1)(s2 + 4)2 (s + 1)2
t3
d s (s2 + 4) · 1 − s · 2s = 3 · e−t · + 4e−t + 4e−t · t
∴ =−
ds s + 4
2 (s2 + 4)2 3!
1 3
s2 − 4 = e−t t + 4t + 4
= 2 2
(s + 4)2
(v) Applying Laplace transform on the equation, we
Applying L−1 on both sides 2s
/ , / , get ȳ(s) = 2
8 1 5 1 4s + π2
x(t) = L−1 + L−1 2 Applying L−1 on both sides
3 (s + 4)
2 9 s +4
/ , / ,
5 1 1 s 1 πt
− L−1 2 y(t) = L−1 2 = cos
9 s +1 2 s + (π/2)2 2 2
8 1 5 5
= · (sin 2t − 2t cos 2t) + sin 2t − sin t (vi) Applying Laplace transform on both sides
3 2·8 9·2 9
4 1 5 [s2ȳ(s)−sy(0)−y (0)]+2[sȳ(s)−y(0)]−3ȳ(s)
= sin 2t − t cos 2t − sin t 1
9 3 9 = 2
(please see Example 12.34) s +1
/ , 1
−1 1 1 ⇒ (s2 + 2s − 3)ȳ(s) = 2
L = (sin at − at cos at) s + 1
(s2 + a2 ) 2a3
1
Partial fractions: ⇒ ȳ(s) = 2 (1)
(s + 1)(s + 3)(s − 1)
s −4
2
1 8 5
= − 2 Applying L−1 on both sides of Eq. (1)
(s + 4)(s + 1)
2 2 3 s +4 s +1
2
/ ,
1
1 y(t) = L−1 (2)
Multiply by both sides (s2 + 1)(s + 3)(s − 1)
s2 + 4 - s+2 . - 1 .
1 1
s2 −4 1 8 5 1 = − L−1 2 − L−1
⇒ 2 = · − · 10 s +1 40 s+3
(s +4)2 (s2 +1) 3 (s2 +4)2 3 (s2 +4)(s2 +1) / ,
1 −1 1
8 1 5 1 5 1 + L
= + · 2 − · 8 s−1
3 (s2 + 4)2 9 s + 4 9 s2 + 1 1 1 1 1
(iv) Taking Laplace transform of the given differen- = − cos t − sin t − e−3t + et
10 5 40 8
tial equation, we obtain
Partial fractions:
[s2 x̄(s) − sx(0) − x (0)] + 2[sx̄(0) − x(0)] + x̄(s) As + B C D 1
+ + = 2
3 s +1
2 s+3 s−1 (s + 1)(s + 3)(s − 1)
=
(s + 1)2 = ȳ(s)
(As+B)(s+3)(s−1)+(s2 +1)[C(s−1)+D(s+3)] = 1 sides,

(3) 2
s ȳ − sy (0) − y (0) + n2 ȳ
1 = aL {sin nt cos α + cos nt. sin α}
C = (s + 3)ȳ(s)s=−3 = − ;
40
a n s
1 ⇒ ȳ = 2 cos α + sin α
D = (s − 1)f (s)ȳ = (4) s + n2 s 2 + n2 s 2 + n2
s=1 8
1 s
= an cos α · 2 + a sin α · 2
Putting s = 0 in Eq. (3) s 2 + n2 s + n2
On inversion we get
−3B − C + 3D = 1 ⇒ 3B = −C + 3D − 1 a at
1 3 1 y = a cos α 2 (sin nt − nt cos nt) + sin α sin nt.
= + −1⇒B=− (5) 2n 2n
40 8 5
Simultaneous linear differential equations
3
Coefficient of s in Eq. (3) with constant coefficients
1 Example 12.47
A+C +D =0 ⇒ A=− . (6) Solve the pair of equations using Laplace transform
10
method:
Example 12.45 x (t)−x (t)+5y (t) = t y (t)−4y (t)−2x (t) = −2
Solve by the method of Laplace transforms
(y − 6y + 11y − 6y = 0) given that with the initial conditions:
y (0) = y (0) = 0 y (0) = 2.
x (0) = x (0) = 0, y (0) = y (0) = 0.
Solution
Solution Let L {x (t)} = x̄ (s) and L {y (t)} = ȳ (s)
Let L {y (t)} = ȳ (s). Taking Laplace transforms on
Application of Laplace transforms converts the
both sides, we have
problem into that of solving a pair of algebraic
equations.
s3 ȳ − s2 y (0) − sy (0) − y (0) − 6 s2 ȳ − sy (0)
1
−y (0) + 11 sȳ − y (0) − 6ȳ = 0 s2 x̄ − sx (0) − x (0) − x̄ + 5sȳ − y (0) =
s2
⇒ s3 − 6s2 + 11s − 6 ȳ = 2 2
s2 ȳ − sy (0) − y (0) − 4ȳ − 2sx̄ − 2x (0) = −
2 s
⇒ ȳ =
(s − 1) (s − 2) (s − 3) Using the initial conditions these equations reduce
2 1 2 1 2 1 to
= · + · + ·
(−1)(−2) s−1 (−1) s−2 2 · 1 s − 3
1 2
s2 − 1 x̄ + 5sȳ = 2 , −2sx̄ + s2 − 4 ȳ = −
s s
On inversion, we get y (t) = et − 2e2t + e3t .
By Cramer’s rule
Example 12.46 1 B
d 2y s2
5s s2 − 1 5s
x̄ =
Solve 2 + n2 y = a sin (nt + α) given that − 2s s2 − 4 −2s s2 − 4
dt
dy 11s2 − 4
y= = 0 at t = 0. =
dt s2 s2 + 1 s2 + 4
Solution 1 5 4
L {y (t)} = ȳ (s). Taking Laplace transforms on both =− 2+ 2 −
s s + 1 s2 + 4
B
s2 − 1 1
s2 s2 − 1 5s (i) y −3y +2y = 4t+e3t , y(0) = 1, y (0) = −1
ȳ =
−2s − 2s −2s s2 − 4 [JNTU 2003]
1 3t
−2s2 + 4 Ans: y(t) = 2t + 3 + (e − e ) − 2e2t
t
= 2
s s2 + 1 s2 + 4
(j) (D2 + 1)x = t cos 2t, x(0) = x (0) = 0
1 2s s
= − 2 + 2 [JNTU 2000]
s s +1 s +4 Ans: x(t) = 49 sin 2t − 59 sin t − 3t cos 2t
Taking inverse Laplace transforms, we get
(k) y + 2y + 5y = e−t sin t, y(0) = 0, y (0) = 1
x (t) = −t + 5 sin t − 2 sin 2t; et
Ans: y(t) = (sin t + sin 2t)
y (t) = 1 − 2 cos t + cos 2t. 3
(l) x + 9x = cos 2t, x(0) = 1, x(π/2) = −1
EXERCISE 12.8
Ans: x(t) = 15 [4 cos 3t + 4 sin 3t + cos 2t]
1. Solve the following equations by Laplace trans- 2. Solve the pairs of linear differential equations by
form method: Laplace transforms:
(a) x (t) + 4x (t) + 4x (t) = 4e−2t ; (a) x (t) − 3x (t) − y (t) + 2y (t) = 14t + 3;
x (0) = −1, x (0) =4 x (t) − 3x (t) + y (t) = 1
Ans: x (t) = e−2t 2t 2 + 2t − 1 x (0) = x (0) = 0, y (0) = 13/2
1 1
(b) x (t) + x (t) = t cos 2t; Ans: x (t) = 2 − et − e3t − e−2t ;
2 2
x (0) = 3, x (0) = 1 5
y (t) = 7t + 5 − et + e−2t
Ans: x (t) = 5 cos t + sin t − 2 cos 2t 2
(c) y (t) − y (t) = 5 sin 2t; (b) 2x (t) + 2x (t) + y (t) − y (t) = 3t; x (t) +
y (0) = 0, y (0) = 1 x (t) + y (t) + y (t) = 1 x (0) = 1, y (0) = 3.
Ans: y (t) = 3 sinh t − sin 2t Ans: x (t) = t + 3e−t − 2e−3t ;
y (t) = 1 − t + 2e−3t
(d) y (t) + 6y (t) + 9y (t) = 6t 2 e−3t ;
dx dy
y (0) = 0, y (0) = 0 (c) = 2x − 3y, = y − 2x;
1 dt dt
Ans: y (t) = t 4 e−3t x(0) = 8, y(0) = 3
2 Ans: x(t) = 5e−t + 3e4t ; y(t) = 5e−t − 2e4t
(e) x (t) + 4x (t) = 2t − 8; x (0) = 1, x (0) = 0
1 1 dx d 2y dx
Ans: x (t) = 3 cos 2t − sin 2t + t − 2 (d) − 2 2 = et , + 2x − y = 1;
4 2 dt dt
dt
x(0) = y(0) = y (0) = 0
(f) x (t) + 2x (t) = 8t; x (0) = 0, x (0) = 0 Ans: x(t) = 1 + e−t − e−at − e−bt ;
Ans: x (t) = 2t 2 − 2t + 1 − e−2t y(t) = 1 + e−t − be−at − ae−bt
(g) y + 2y − y − 2y = 0, y(0) = y (0) = 6 Unit-step function

[JNTU 1995]
−2t There arise practical situations in which values of
Ans: y(t) = et − 3e−t+ze
functions change suddenly. We define a function
which will suppress a given term up to a certain value
(h) (D2 + 4D + 3)y = e−t , y(0) = 1
of t and insert the term for all large t. The function
[JNTU 2002S]
7 −t 3 −3t t −t we are about to introduce leads to a powerful tool in
Ans: y(t) = e − e + e constructing inverse transforms.
4 4 2
/ ,
12.11 Heaviside’s Partial Function 2s2 + 5
(ii) L−1
Expansion Theorem /(s + 1)(s
, + 2)(s + 3)
Theorem 12.18 Let f¯ (s) and d̄(s) be two poly- (iii) L −1 1
.
nomials in s without common zeros. Let the lead- s3 − 1
ing coefficient of ḡ(s) be unity and deg f¯ (s) < Solution
deg ḡ(s)(= n). If α1 , α2 , αn are n simple zeros, we (i) We have
can write ḡ(s) = (s − α1 )(s − α2 ) · · · (s − αn ). Then
3s + 4 A B
f¯ (s) = +
the inverse Laplace transform of is given by (s − 1)(s − 2) s − 1 s − 2
ḡ(s)
3s + 4
) where A = = −7
f¯ (s) %n ¯
f (αi ) αi t s − 2 s11
−1
L = e . 3s + 4
ḡ(s) ḡ(αi ) B= = 10
i=1
s − 1 s=2
/ ,
Proof: By hypothesis f¯ (s)/ȳ(s) is a proper frac- 3s + 4
∴ L−1
(s − 1)(s − 2)
tion. So, there exist n constants Ai (i = 1, 2, · · · n) / , / ,
1 1
such that = −7L−1 + 10L−1
s−1 s−2
f¯ (s) f¯ (s)
= = −7e + 10e
t 2t
ḡ(s) (s − α1 )(s − α2 ) · · · (s − αn )
A1 A2 Ai An (ii) We have
= + +· · ·+ +· · ·+
s−α1 s−α2 s−αi s−αn 2s2 +5 A B C
= + +
f¯ (s) (s − αi ) (s+1)(s+2)(s+3) s+1 s+2 s+3
⇒ Ai = lim (s − αi ) = f¯ (αi ) · lim
s→αi ḡ(s) x→αi ḡ(s) 2s2 + 5 7
where A = = ;
0 (s + 2)(s + 3) s=−1 2
∴ f (αi ) = 0; form
0 2s2 + 5
B= = −13;
1 (s + 1)(s + 3)
= f¯ (αi ) · lim −1 , by L’ Hospital’s rule s=−2
x→αi g (s)
2s2 + 5 23
f¯ (αi ) C= =
= (s + 1)(s + 2) s=−3 2
ḡ(αi )
f¯ (s) % f¯ (αi )
n / ,
∴ = ·
1
−1 2s2 +5 7 23
(α ) s − α ∴L = e−t −13e−2t + e−3t
ḡ(s) i=1
ḡ i i (s+1)(s+2)(s+3) 2 2
) / ,
f¯ (s) %n
f¯ (αi )
⇒ L−1 = ·L −1 a
, (iii) Here f¯ (s) = 1, ḡ(s) = s3 − 1;
ḡ(s) i=1
ḡ (αi ) s · · · αi
∴ ḡ (s) = 3s2 . √
by linearity
) 1 1 3
% Zeros of ḡ(s) are α2 = 1, v2 = w = − + ,
f¯ (s) f¯ (αi ) αit
n
or L−1 = e ··· . √ 2 2
ḡ(s) i=1
ḡ (αi ) α3 = ω 2 = − −
1 i 3
which are the cube roots of
2 2
Example 12.48 unity. By Heaviside’s formula we have
Using Heaviside’s expansion formula find / ,
/ , 1 1 1 1
−1 3s + 4 L −1
= e αi t + eα2 t + eα3t
(i) L ; s −1
3 ḡ (α1 ) ḡ (α2 ) ḡ (α3 )
(s − 1)(s − 2)
s
1 et 1 ωt 1 ω2 t (a)
= + 2e + 4e −1
s2
3 12 ω ω
1
1 t 2 Ans: (et + e−t ) = cosh t
= e + ωeωt + ω2 eω t 2
3
2s + 7
(b) 3
s −s
1 1 9 5
∴ ω3 = 1, 2 = ω, 4 = ω2 Ans: −7 + et + 4−t
ω ω √ 2 2
1 t 1 3 −t/2 −i √3t 3s2 + 5
= e + − +i e ·e 2 (c)
3 2 2 (s + 1)(s − 2)(s − 3)
√ 2 17 2t
1 3 −t/2 −i √3 t Ans: e−t − e + 8e3t
+ − −i e ·e 2 3 6
2 2 4
√ √ (d)
1 t 3t √ 3t (s + 1)(s + 2)
−t/2
= e −e cos + 3 sin . Ans: 4(e−t − e−2t )
3 2 2
1
EXERCISE 12.9 (e)
+1
s3 √ √
1 −t 3t √ 3t
1. Using Heaviside’s expansion formula find the Ans: e − et/2 cos − 3 sin
inverse Laplace transform of 3 2 2
Table 12.3 A Short Table of Transforms
S.No f (t) f¯ (s) α (s > α)
1
1. 1 0
s
1
2. eat a
s−a
n!
3. tn 0
sn+1
(n = 1, 2, · · · )
n!
4. t n eat a
(s − a)n+1
(n = 1, 2, · · · )
a
5. sin at 0
s 2 + a2
s
6. cos at 0
s 2 + a2
s
7. sinh at |a|
s 2 − a2
s
8. cosh at |a|
s 2 − a2
b
9. e−at sin bt −a
(s + a)2 + b2
s+a
10. e−at cos bt −a
(s + a)2 + b2
*
1 1 π
11. t2 0
2s s
*
1 π
12. t− 2 0
s
(b + 1)
13. t b (b > −1) 0
sb+1
(b + 1)
14. t b eat(b>−1) a
(s − a)b+1
e−as
15. Ua (t) 0
s
a−b
16. eat − ebt (a > b) a
(s − a)(s − b)
1 1 b2 − a2
17. sin at − sin bt 0
a b (s2 + a2 )(s2 + b2 )
(b2 − a2 )s
18. cos at − cos bt 0
(s2 + a2 )(s2 + b2 )
Table 12.4 Properties of Laplace Transforms of Functions of t
Property Function of t Laplace transform of function of t

∞
Definition f (t) e−st f (t)dt s > 0
0
Linearity c1 f1 (t) + c2 f2 (t) c1 L{f1 (t)} + c2 L{f2 (t)}
1¯ s
Change of scale f (at) f
a a
First shift theorem e−at f (t) ¯ a)

f (s +
Transforms of derivatives f (t) sf¯ (s) − f (0)

f (t) s2 f¯ (s) − sf (0) − f (0)
··· ···
f (n) (t) sn f¯ (s) − sn−1 f (0) − − − f (n−1) (0)
t 1¯
Transforms of integrals f (t)dt f (s)
0 s
d ¯
Multiplication by t tf (t) − f (s)
ds n
d
tn t n f (t) (−1)n f¯ (s)
ds
∞
f (t)
Division by t f¯ (s)ds
t s
−as ¯
Second shifting theorem f (t − a)Ua (t) e f (s)
T
0 e−st f (t)dt
Periodic function f (t)with period T
1 − e−st
Initial value theorem lim f (t) lim sf¯ (s)

t→0 s→∞
Final Value Theorem lim f (t) lim

t→∞ s→0f¯ (s)
t
Convolution theorem f (u)g(t − u)du f¯ (s)ḡ(s)
0
Table 12.5 Properties of Laplace Transforms of Functions of s
Property Function of s Laplace transform of function of s
Definition f (s) f (t) = L−1 {f¯ (s)}
Linearity c1 f¯1 (s) + c2 f¯2 (s) c1 L−1 {f¯1 (s)} + c2 L−1 {f2 (s)}
First shifting theorem f¯ (s − a) L−1 {f¯ (s − a)} = eat f (t)
s
Change of scale f¯ L−1 f¯ as = af (at)
a
Inverse transform of f¯ (s) L−1 {f¯ (s)} = −tf (t)

derivative f¯ (n) (s) L−1 {f¯ (n) (s)} = (−1)n t n f (t)
∞ ∞ f (t)
Inverse transform of f¯ (s)ds L−1 { f¯ (s)ds} =
s s t
integral
Multiplication by s sf¯ (s) L−1 {sf¯ (s)} = f (t)

provided f (0) = 0
f¯ (s) -¯ . t
f (s)
Division by s L−1 s = 0 f (u)du
s
/
f (t − a), t>a
Second shifting theorem e−as f¯ (s) L−1 {e−as f¯ (s)} =
0, t<a
t
Convolution theorem f¯ (s)ḡ(s) L−1 {f¯ (s)ḡ(s)} = f (u)g(t − u)du
0
Question Bank
Mulitple Choice Questions

Chapters 1, 2 Ordinary & Linear (c) sin y/x (d) cos y/x Ans: (c)
Second and Higher
Order Differential 6. The orthogonal trajectory of y = ax + c where
Equations ‘c’ is a parameter is
1. The order and the degree of the differential (a) x + ay = c (b) x − ay = c
2 3/2
d 2y dy (c) ax = y + c (d) ax + y = c Ans: (a)
equation k 2 = 1 + are
dx dx
(a) (1, 2) (b) (2, 1) 7. The general solution of D3 − D y = 0 is
(c) (2, 2) (d) (2, 3) Ans: (c) (a) y = c1 + c2 x + c3 x2
(b) y = c1 + c2 ex + c3 e−x
2. The differential equation obtained by eliminat- (c) y = c1 + c2 cos x + c3 sin x
ing ‘c’ from y = cx + c2 − c3 is (d) none of the above Ans: (b)
(a) y = xy1 + y12 − y13
8. The complete solution of xdx + ydy +
(b) y = xy1 + y1 − y13 xdy − ydx
(c) y = xy + y12 − y13 = 0 is
x2 + y2
(d) none of the above Ans: (a)
x2 + y2
(a) − tan−1 (y/x) = c
3. The differential equation representing the fam- 2
ily of straight lines through the origin x2 + y2
(b) + tan−1 (y/x) = c
2
(a) xy1 = y2 (b) y = xy1 (c) x2 + y2 − tan−1 (y/x) = c
(c) xy + y1 = 0 (d) xy1 + y = 0 Ans: (b) (d) x2 + y2 + tan−1 (y/x) = c Ans: (b)
4. The differential equation obtained by elimi- dy y

nating A and α from y = A cos(mx + α) is 9. The complete solution of = +
dx x
(a) y2 + m2 y = 0 (b) y2 − m2 y = 0 1 + (y/x)2 is cx =
(c) y2 − my = 0 (d) y2 + my = 0 Ans: (a) y
(a) + 1 + (y/x)2
x
dy y y y
5. The general solution of = + tan is (b) − 1 + (y/x)2
dx x x x
cx = y
(a) sinh y/x (b) cosh y/x (c) + 1 − (y/x)2
x
A-2 Engineering Mathematics I
x
(d) + 1 + (x/y)2 Ans: (a) (c) xy = sin x + c
y (d) none of the above Ans: (c)
dy y+x−2
10. The complete integral of = is
dx y−x−4 17. The general solution of (x2 +y2 )dx+xydy = 0 is
(a) x2 −y2 + 2xy−4x+8y+c = 0 (a) x4 + 2x2 y2 = c
(b) x2 +y2 +2xy−4x+8y+c = 0 (b) x3 + x2 y2 = c
(c) x4 + x2 y2 = c
(c) x2 −y2 −2xy−4x+8y+c = 0
x4
(d) x2 −y2 +2xy+4x−8y+c = 0 Ans: (a) (d) + x2 y2 = c Ans: (c)
2
Hint: Write the equation as: xdy + ydx + (x −
2)dx − (y − 4)dy = 0 , 2d(xy) + (2x − 4)dx − ydx − xdy
18. The general solution of = 0 is
(2y − 8)dy = 0 and integrate y2
(a) xy = c (b) x = cy
dy y−x
11. The complete integral of = (c) x = cy2 (d) x2 = cy Ans: (b)
dx y−x+2
(a) x2 − y2 + 2xy + 4y + c = 0 dy
19. The complete solution of + xy = x is
(b) x2 + y2 + 2xy + 4y + c = 0 dx
x2 + y2 − 2xy + 4y + c = 0
2
(c) (a) y = 1 + ce−x
(d) x2 − y2 − 2xy + 4y + c = 0 Ans: (c) (b)
2
y = 1 + cex /2
2
(c) y = 1 + ce−x /2
Hint: Write the equation as: ydx + xdy − xdx − 2
(d) y = ce−x /2 Ans: (c)
(y + 2)dy = 0, ⇒ −2d(xy) + 2xdx + (2y +
4)dy = 0 and integrate 20. The orthogonal trajectories of the family of
curves r x = ax cos nθ are
12. If x2 − ay dx + y2 − 3x dy = 0 is exact
(a) r n = cn sin θ
then a =
(b) r n = cn sin nθ
(a) −3 (b) 3 (c) 0 (d) 1 Ans: (b)
(c) r n = cn sec nθ
13. The orthogonal trajectories of xy = c are (d) none of the above Ans: (b)
(a) y/x = c (b) x2 − y2 = c
(c) y + 2x = c (d) x2 − 2y = c Ans: (b)
2 21. The complete solution of D2 − 1 y = 0 is y =
dy y (a) C1 ex + C2 e−x
14. An integrating factor of − = x is (b) C1 cos x + C2 sin x
dx x
1 1 1 1 (c) (C1 + C2 x) ex
(a) − (b) − 2 (c) (d) 2 Ans: (c) (d) (C1 + C2 x) cos x Ans: (a)
x x x x
x y
15. The differential equation obtained from + =
a b 22. The complete solution of (D − 1)2 y = 0 is y =
1 by eliminating the constants a and b is (a) (C1 + C2 x) e−x
d 2y d 2y (b) (C1 + C2 x) cos x
(a) = 0 (b) =1
dx2 dx2 (c) (C1 + C2 x) ex
d 2y d 2y (d) C1 ex + C2 ex Ans: (b)
(c) = x (d) =y Ans: (a)
dx2 dx2 1
23. The P.I. of =
16. The complete solution of the differential equa- (D − 1)2 ex
tion xy1 + y = cos x is x2 e−x
y (a) xex (b)
(a) = sin x + c 2
x x2 x
(b) xy = cos x + c (c) e (d) x2 ex Ans: (c)
2
Question Bank A-3
1 Chapter 3 Rolle’s Theorem and

24. The P.I. of sin ax =
D2+ a2 Mean Value Theorems
1 1
(a) sin ax (b) cos ax 1. The value of c in Rolle’s theorem for f (x) =
2a 2a2 sin ax in (0, πa) is
1 1
(c) cos ax (d) − cos ax Ans: (d) 1 π π π
2a 2a (a) (b) (c) (d) Ans: (c)
1 a 4a 2a 6a
25. The P.I. of x= 2. The value of c in Rolle’s theorem for f (x) = x2
(D + 1)2
(a) x + 1 (b) x − 2 in (−1, 1) is
(c) x + 2 (d) x Ans: (b) (a) 0 (b) 0.5 (c) 0.25 (d) −0.5 Ans: (a)
1
26. sin x = 3. The value of c is Rolle’s theorem for f (x) =
D2 x2 − x (0, 1)
(a) cos x (b) sin x
(c) − sin x (d) − cos x Ans: (c) (a) 0 (b) 0.5 (c) 0.25 (d) −0.5 Ans: (b)
1 4. The value of c in Lagrange’s mean value theo-

27. e−x sin x =
(D + 1)2 rem for f (x) = ex in (0,1) is

(a) ex cos x (b) ex sin x (a) log e − e−1 (b) log e
(c) −e sin x (d) −ex cos x
x
Ans: (c) (c) log(e + 1) (d) log(e − 1) Ans: (d)
1 5. The value of c in Cauchy MVT for f (x) = ex
28. xe2x =
(D − 2)2 and g(x) = e−x in (3, 7) is
x3 2x
(a) e (b) x3 e2x (a) 4 (b) 5 (c) 4.5 (d) 6 Ans: (b)
6

x2 2x x3 2x e7 −e3 e7 −e3
(c) e (d) e Ans: (a) Hint: − e = −7 −3 = e
2e 10
= −e10
2 12 e −e e3 −e7
1 θ = 0 ⇒ f (x + h) = f (x) + hf (x)
29. The P.I. of sin x =
D2 +D θ = 1 ⇒ f (x + h) = f (x) + hf (x + h)
1
(a) (cos x + sin x) (x + h)2 = x2 + h · 2(x + θh)
2 1
1 h2 = 2θh2 ⇒ θ =
(b) − (cos x + sin x) 2
2
1 6. The value of θ if f (x) = x2 and f (x + h) =
(c) (cos x − sin x)
2 f (x) + hf (x + θh)
1 (a) −0.5 (b) 0.25 (c) 0 (d) 0.5 Ans: (d)
(d) (− cos x + sin x) Ans: (b)
2
7. The value of c in Rolle’s theorem for f (x) =
1
30. The P.I. of 2x = sin x/ex in (0, π) is
(D − 3) (a) π/4 (b) π/2 (c) π/3 (d) 2π/3 Ans: (a)
2x 2x
(a) (b)
x log 2 + 3 x log 2 − 3 8. The value of c is Cauchy’s Mean Value Theo-
2x 2x √ 1
(c) (d) Ans: (c) rem for f (x) = x and g(x) = √ in (1, 4) is
log 2 − 3 log 2 + 3 x
(Hint: Write 2x = ex (log 2)) (a) 1.5 (b) 2 (c) 2.5 (d) 3 Ans: (b)
9. The value of c in Lagrange’s Mean Value 17. Maclaurin’s expansion for log(1 + x) is
Theorem for f (x) = log x in [1, e] is x2 x3 x4
(a) (e − 1)−1 (b) e + 1 (a) x − + − + ...
2 x 4
(c) e − 1 (d) e Ans: (c)
x2 x3 x4
(b) x + + + + ...
10. Lagrange’s MVT is not applicable to the func- 2 3 4
tion f (x) = x1/3 in [−1, 1] because x2 x3 x4
(c) x + + + + ...
(a) f (−1) = f (1) 2! 3! 4!
(b) f is not continuous in [−1, 1] x2 x3 x4
(c) f is not derivable in (−1, 1) (d) x − + − + ... Ans: (a)
2! 3! 4!
(d) f is not a bijective function Ans: (c)
11. Lagrange’s MVT is not applicable to the func- 18. Maclaurin’s expansion of cos x is
tion defined on [−1, 1] by f (x) = x sin 1x (x = 0) %∞
x2r
and f (0) = 0 because (a)
r=0
(2r)!
(a) f (−1) = f (1) %∞
(−1)r x2r
(b) f is not continuous in [−1, 1] (b)
(c) f is not derivable in (−1, 1) r=0
(2r)!
(d) f is not a one-to-one function Ans: (c) %∞
(−1)r x2r+1
(c)
r=0
(2r + 1)!
12. The value of c for Lagrange’s MVT for the %∞
x2r+1
function f (x) = cos x in [0, π/2] is (d) Ans: (b)
(2r + 1)!
(a) cos−1 (2/π) (b) sin−1 (2/π) r=0
(c) sin−1 (1/π) (d) cos−1 (1/π) Ans: (b)
19. The expansion of ex in powers of (x − 1)
13. The value of c of Rolle’s theorem for f (x) = ∞
% (x − 1)r
(x − a)(x − b) in [a, b] is (a) e
a+b √ r!
(a) − (b) ab r=0
2 % ∞
(x − 1)r
a+b (b) e−1
(c) a + b (d) Ans: (d) r!
2 r=0
%∞
(−1) (x − 1)r
r
14. The value of c of Lagrange’s Mean Value The- (c) e
r!
orem for f (x) = (x − 2)(x − 3) in [0, 1] is n=0
%∞
(−1)r (x − 1)r
(a) 0.5 (b) 1 (c) 2.5 (d) 2 Ans: (a) (d) Ans: (a)
r=0
r!
15. The value of c of Rolle’s theorem for f (x) =
(x − 1)(x − 2) in [0, 3] is 20. The expansion for sinx in powers of(x−π/2) is
(a) 1.5 (b) 2.5 (c) 3 (d) 2 Ans: (a) 1 1
(a) 1 − (x − π/2)2 + (x − π/2)4 − . . .
2 4
16. The value of c of Cauchy’s Mean Value 1
(b) x + (x − π/2) + (x − π/2)3 + . . .
Theorem for f (x) = sin x and g(x) = cos x in 3!
[0, π/2] 1 1
(c) 1 + (x − π/2) + (x − π2)4 + . . .
2
2 4!
(a) π/8 (b) π/6 (c) π/4 (d) π/3 1
(d) x − (x − π/2) + (x − π/2)3 + . . .
3!
Ans: (c)
Ans: (a)
Question Bank A-5
Chapter 4 Functions of Several 11. The

√ degree of the homogeneous function
√ √ √
Variables x − y / x + y is
∂u ∂u ∂u
1. If u = log(x2 +y2 +z 2 ) then x +y +z = 1
∂x ∂y ∂z (a) 1 (b) 0 (c) −1 (d) Ans: (b)
2
(a) 0 (b) 1 (c) 2 (d) 2u Ans: (c) ∂u ∂u
12. If u = f (y/x) then x +y =
∂x ∂y
2. If u = log(x3 + y3 +
z − 3xyz) then (x + y +
3
∂u ∂u ∂u (a) 0 (b) 1 (c) u (d) 2u Ans: (a)

z) + + =
∂x ∂y ∂z dy
13. If x3 + y3 − 3xy = 0 then =
(a) 0 (b) 1 (c) 2 (d) 3 Ans: (d) dx
x2 − y x −y
2
−1 x ∂u ∂u (a) (b) − 2
3. If u = tan then x + y = y2 − x y −x
y ∂x ∂y y2 − x y2 − x
(a) 0 (b) 1 (c) 2 (d) 3 Ans: (a) (c) 2 (d) − 2 Ans: (b)
x −y x −y
If z =
4. B f (x + ay) + φ(x − ay) then ∂(u, v)
∂2 z ∂2 z 14. If u = x2 − y, v = x + y then =
= ∂(x, y)
∂y2 ∂x2 (a) 2x − 1 (b) 2x + 1
(a) 0 (b) a (c) a2 (d) 1 Ans: (c) (c) 1 (d) 0 Ans: (b)
∂z ∂z ∂(u, v)
5. If z = (x2 + y2 )/(x + y) then + = 15. If u = xy and v = y then =
∂x ∂y ∂(x, y)
(a) 4xy/(x+y)2 (b) 2xy/(x+y)2 (a) −2y/x (b) 2y/x
(c) 4xy/(x+y) (d) 4xy/(x−y)2 Ans: (a) (c) −y/x (d) y/x Ans: (b)
∂u ∂u ∂u ∂(x, y)
6. If u = f (y−z, z−x, x−y) then + + = 16. If x = r cos θ, y = r sin θ then =
∂x ∂y ∂z ∂(r, θ)
(a) 0 (b) 1 (c) 2 (d) −1 Ans: (a) (a) −r (b) r (c) 0 (d) r −1 Ans: (b)
∂u ∂u
7. If u = xy then x +y = 17. The necessary conditions for a function f (x, y)
∂x ∂y to have an extreme value are
(a) u log(u/e) (b) u log u
∂f ∂f
(c) u (d) u log(eu) Ans: (d) (a) > 0, >0
∂x ∂y
∂f ∂f
∂u ∂u (b) = 0, =0
8. If u = x2 + 2xy then x +y = ∂x ∂y
∂x ∂y ∂f ∂f
(a) u (b) 2u (c) 0 (d) 1 Ans: (b) (c) = 0, = 0
∂x ∂y
∂f ∂f
∂u ∂u (d) < 0, <0 Ans: (b)
9. If u = xy/(x + y) then x +y = ∂x ∂y
∂x ∂y
(a) u (b) 2u (c) 0 (d) 1 Ans: (a) 18. A function f (x, y) has a maximum value at (a, b)
∂2 f ∂2 f ∂2 f
∂u ∂u if—, where... r = 2 , s = , t= 2
10. If u = log(x2 /y) then x +y = ∂x ∂x∂y ∂y
∂x ∂y
(a) u (b) 2u (c) 1 (d) 0 Ans: (c) (a) rt − s2 < 0, r < 0
(b) rt − s2 > 0, r < 0 r2 + r

(c) rt − s2 > 0, r > 0 (d) Ans: (a)

(d) rt − s2 = 0 Ans: (b) r 2 − 2r 2 + rr
5. The envelope of the family of straight lines
19. If rt − s2 < 0 at a point P = (a, b) then P is a a
(a) maximum point y = mx + where m is a parameter is
m
(b) minimum point (a) y2 +4ax = 0 (b) y2 = 4ax
(c) saddle point (c) y2 = 2ax (d) y2 +2ax = 0 Ans: (b)
(d) none of the above Ans: (c)
6. The radius of curvature of the circle x2 + y2 −
x+y 2x − 2y = 0 at (0, 0) is
20. If u = and v = tan−1 x + tan−1 y then √
1 −
xy
u, v x, y (a) 4 (b) 2 (c) 1 (d) 2 Ans: (d)
J ·J =
x, y u, v 7. The radius of curvature at the origin of the curve
(a) 0 (b) 1 (c) −1 (d) 2 Ans: (b) x3 + 3x2 y − 4y3 + y2 − 6x = 0 is
(a) 2 (b) 3 (c) 1 (d) 0 Ans: (b)
Chapter 5 Radius of Curvature
8. The radius of curvature of the Folium of Des
1. The radius of curvature of the catenary s =
Cartes x3 + y3 = 3axy at (0, 0) is
c tan ψ is ρ =
(a) c2 sec2 ψ (b) sec2 ψ (a) 3a (b) 3a/2 (c) 3a/4 (d) 3/2 Ans: (b)
(c) c sec2 ψ (d) c sec ψ Ans: (c)
9. The envelopes of the family of parabolas y2 =
2. The curvature of the circle x2 + y2 = r 2 at any 2a(x − a) where ‘a’ is a parameter is
point on it is (a) x = ±ay (b) x = ±2ay
(c) x = ±2y (d) x = ±y Ans: (c)
(a) 1/r (b) 1/r 2 (c) 0 (d) 1 Ans: (a)
10. The envelope of the family of straight lines
3. The Cartesian formula for the radius of curva- x cos α+y sin α = c sec α (α is a parameter) is a
ture is ρ =
3 2 3/2 (a) circle (b) elipse
1 + y12 1/y1 (c) parabola (d) hyperbola Ans: (c)
(a) (b) 2
y2

2 3/2
y2 2
1 + y1 1 + y1
(c) (d) Ans: (c) Chapter 6 Curve-Tracing
y2 y2
1. If the powers of y in the equation of a curve are
4. The polar formula for the radius of curvature is all even then the curve is symmetric about the
ρ= (a) x-axis (b) y-axis

3/2 (c) origin (d) lines y = ±x Ans: (a)
r2 + r 2
(a)
r 2 + 2r 2 − rr 2. If the powers of x in the equation of a curve are

3/2 all even then the curve is symmetric about the
r2 − r 2 (a) x-axis (b) y-axis
(b)
r 2 + 2r 12 − rr (c) origin (d) lines y = ±x Ans: (b)

3
r2 + r 2 3. If the powers of x and y in the equation of a
(c) 2

r2 + 2r − rr curve are both even then the curve is symmetric
about
Question Bank A-7
(a) the x-axis (b) the y-axis (d) cusp Ans: (d)
(c) the origin (d) both the axes Ans: (d)
a−x
4. Among the following, the curve which is sym- 12. For the curve y2 = x2 , the origin is a/an
a+x
metrical about the line y = x is
(a) isolated point
(a) y2 (2a − x) = x3
a+x (b) point of symmetry
(b) y2 = x2 (c) node
a−x
(c) x3 + y3 = 3axy (d) cusp Ans: (c)
(d) 3ay2 = x(x − a)2 Ans: (c)
13. The straight lines y2 = x2 are tangents to the
5. No part of the curve y = x /(1−x ) lies between
2 2 curve
(a) y2 (a + x) = x2 (a
− x)
(a) y = 0, −1 (b) y = 0, 1 (b) y = x2 / 1 − x2
(c) y = 1, 2 (d) y = −1, −2 Ans: (a) (c) x3 + y3 = 3axy
(d) 3ay2 = x(x − a)2 Ans: (a)
6. The parametric equations of the curve
(x/a)2/3 + (y/b)2/3 = 1 are 14. The equations of the asymptotes of the curve
(a) x = cos3 θ, y = sin3 θ y = (x2 + 1)/(x2 − 1) are
(b) x = a cos3 θ, y = b sin3 θ (a) x = 1, y = ±1
(c) x = a cos θ, y = b sin θ (b) x = ±1, y = 1
(d) x = a2 cos3 θ, y = b2 sin3 θ Ans: (b) (c) x = −1, y = 1
(d) x = 2, y = ±1 Ans: (b)
7. No part of the curve y = (x2 + 1)/(x2 − 1) lies
between 15. The straight lines y2 = x2 are tangents to the
(a) y = 2, −2 (b) y = ±1√ curve
(c) y = ±3 (d) y = ± 2 Ans: (b) (a) x2 y2 = a2 y2 − x2
(b) y2 = 4ax
8. The curve ay2 = x3 is symmetrical about (c) x2 + y2 = a2
(a) the x-axis (b) the y-axis (d) 3axy = x3 + y3 Ans: (a)
(c) both the axes (d) the line y = x Ans: (a)
16. The region of the curve y2 (1 − x2 ) = x2 (1 − x2 )
9. One of the points of intersection of the curves is that
x2 + y2 = 2ay and y = x is (a) for which x < −1
(a) (−a, a) (b) (a, −a) (b) which lies between x = ±1
(c) (a, a) (d) (−a, −a) Ans: (c) (c) for which y > 1
(d) for which x > 1, y > 1 Ans: (b)
10. For the folium of Des Cartes x3 + y3 = 3axy the
origin is a/an 17. The asymptote of the curve (a − x)y2 = x3 is
(a) isolated point (a) x + a = 0 (b) x = a
(b) point of symmetry (c) y = 0 (d) y = a Ans: (a)
(c) node
(d) cusp Ans: (c) 18. The curve y = x3 − 3x2 − 9x + 9 has point of
inflexion at
11. For the curve y2 (2a − x) = x3 , the origin is a/an (a) x = 3 (b) x = −4
(a) isolated point (c) x = 1 (d) x = 2 Ans: (c)
(b) point of symmetry
(c) node 19. The lines θ = ±π/4 are tangents to the curve
(a) r 2 = a2 cos 2θ 4. The arc of the upper half of the cardioid is

(b) r = a + b cos θ(a > b) bisected at θ =
(c) r = a(1 + cos θ)
(a) π/3 (b) π/2 (c) π/4 (d) π/6 Ans: (a)
(d) r = a cos θ Ans: (a)
20. The equation of the oblique asymptote of the 5. The length of the arc between x = 0 to a of the
curve x3 + y3 = 3axy is x
catenary y = c cosh is given by
(a) y = x c
(b) x + y + a = 0 a
x
(c) x − y + a = 0 (a) sinh dx
(d) x + y − a = 0 Ans: (b) c
0
a
x
(b) cosh dx
Chapter 7 Applications of c
Integration 0
a
1. The formula for the length of the arc in Cartesian x
(c) 1 + cosh dx
coordinates is c
0
b (d) none of the above Ans: (b)
(a) 1 + y1 dx
a 6. The area of one loop of the curve r = a cos 3θ is
b +
(a) πa2 /4 (b) πa2 /6
(b) 1 + y13 dx
(c) πa2 /8 (d) πa2 /12 Ans: (d)
a
b +
(c) 1 + y12 dx x2
7. The volume of the solid generated by curve +
a
a2
(d) none of the above Ans: (c) y2
= 1 about the y-axis is
b2
2. The formula for the length of the arc in para- 2 2
(a) πa2 b (b) πab2
metric coordinates is 3 3
t 2 4 4
(c) πab2 (d) πa2 b Ans: (c)
(a) x2 + y2 dt 3 3
t1
t 2 8. The area enclosed by the curve r 2 = a2 cos2θ is

(b) x2 − y2 dt (a) 2a2 (b) a2
t1
(c) a2 /2 (d) (3/2)a2 Ans: (b)
t 2
(c) 1 + y2 /x2 dt 9. The perimeter of the curve r = 2a cos θ is
t1 (a) 3aπ (b) 2aπ
(d) none of the above Ans: (a) (c) aπ (d) 4aπ Ans: (b)
3. The perimeter of the loop of the curve

10. The length of the arc of the curve x = a cos θ, y =
3ay2 = x(x − a)2 is
√ a sin θ from θ = 0 to θ = π/3 is
(a) 4a/3
√ (b) 4a/ 3 (a) aπ/2 (b) aπ/6
(c) 2a/ 3 (d) 2a/3 Ans: (b) (c) 2aπ/3 (d) aπ/3 Ans: (d)
Question Bank A-9
Chapter 8 Multiple Integrals a6 a6

(c) (d) Ans: (c)
1 2 48 42
1. x2 ydxdy = √
a a2 −x2
0 0
8. In polar coordinates the integral (x2 +
2 1 4 8
(a) (b) (c) (d) Ans: (c) 0 0
3 3 3 3 y2 )dydx is
1 1 a π/2 a π/2
2. (x + y)dxdy = (a) r 2 drdθ (b) r 3 drdθ
0 0 0 0 0 0
(a) 1 (b) 2 (c) 3 (d) 4 Ans: (a) a π/4 a π/4
(c) r 2 drdθ (d) r 3 drdθ Ans: (b)
2 3
0 0 0 0
3. (4 − y2 )dxdy =
0 0 2 3 2
(a) 16/3 (b) 8/3 (c) 5/3 (d) 16 Ans: (d) 9. xy2 z dzdydx =
0 1 1
4. In polar coordinates the integral (a) 24 (b) 28 (c) 20 (d) 26 Ans: (d)
∞ ∞
2 2
e−(x +y ) dxdy = ∞
2
0 0 10. e−x dx =
π/2∞ π/4∞ 0
(a) e −r 2
drdθ (b) e−r rdrdθ √
(a) 1 (b) π/2 (c) π/2 (d) 0 Ans: (c)
0 0 0 0
π/2∞ π/2∞
2
(c) e−r rdrdθ (d) e−r drdθ Chapter 9 Sequences and Series
0 0 0 0 1. Which one of the following is the nth term ‘sn ’
Ans: (c) of a bounded sequence?
π acos θ (a) 1 + (−1)n (b) (−1)n−1 n
5. r sin θdrdθ = (c) (2n − 1) (d) −n Ans: (a)
0 0
2. Which one of the following is the nth term ‘sn ’
(a) −a2 /3 (b) −a2 /2
of the sequence unbounded above?
(c) a2 /3 (d) a3 /3 Ans: (c)
(a) 2n /3n (b) sin n
∞ y (c) −n2 (d) (n2 + 1)/n Ans: (d)
e−y
6. dxdy =
y 3. Which one of the following is a monotonically
0 0
increasing sequence ‘{sn }’ if sn =
(a) +1 (b) 0
1 1
(c) −1 (d) 2 Ans: (a) (a) n (b) 1 − n
2 2
a x y (c) −n (d) (−1)n n Ans: (b)
7. xyz dxdydz
4. Which one of the following is a convergent
0 0 0 sequence?
a6 a6 EnF < =
(a) (b) (a) (b) n2
24 36 m
@ A @ A $
n 2 n+1 2 (a) l > 1 ⇒ $ un is divergent
(c) (d) Ans: (c) (b) l = 1 ⇒ $ un is convergent
n+1 n
(c) l < 1 ⇒ $ un is divergent
$
m
(d) l = 1 ⇒ un is convergent Ans: (a)
sn
x=1
% ∞ %∞
5. If lim sn = l then lim = n n+1
n→∞ nn→∞ 13. The series (i) and (ii)
(a) 0 (b) ∞ (c) 1/l (d) l Ans: (d) n=1
n+1 n=1
n
are such that
1 1 1 (a) (i) converges (ii) diverges
6. + + ... + =
1.2 2.3 n(n + 1) (b) (i), (ii) both converge
(n + 1) (c) (i), (ii) both diverge
(a) 1 (b)
n (d) (i) diverges (ii) converges Ans: (c)
n 1
(c) (d) 1 + Ans: (c)
(n + 1) n+1 1 1 1
14. The series 1 − 2
+ 2 − 2 + . . . is
7. The nth
partial sum of the series 2 3 4
% 1 (a) conditionally convergent
is (b) convergent
(n + 1)(n + 2)
1 n (c) divergent
(a) (b) (d) absolutely convergent Ans: (d)
2(n + 2) n+2
1 n
(c) (d) Ans: (d) 1 1 1
n+2 2(n + 2) 15. The series 1 − √ + √ − √ + . . . is
% % 2 3 4
8. The geometric series r n and (2r)n are (a) conditionally convergent
both convergent if (b) convergent
1 (c) divergent
(a) |r| < 1 (b) |r| < (d) absolutely convergent Ans: (a)
2
1
(c) |r| ≤ 1 (d) |r| ≤ Ans: (b) %
∞

2 16. The series (−1)n−1 (n + 1)1/2 − n1/2
% 1 n=1
9. The series is convergent if and only if (a) diverges
nk+1
(a) k = 1 (b) k < 1 (b) converges
(c) k > 0 (d) k > 1 Ans: (c) (c) absolutely convergent
(d) conditionally convergent Ans: (d)
%
10. un is a series of positive terms such that
un % %
∞
(−1)n−1
lim = l then un is 17. The series is absolutely convergent
un + 1 n=1
nk
(a) l < 1 (b) l > 1 if
(c) l = 1 (d) l > 0 Ans: (a) (a) k = 1 (b) k > 1
(c) 0 < k < 1 (d) k = 0 Ans: (b)
11. un is the nth term of a convergent series than
(a) lim un > 0 (b) lim un = 0 %
∞
(−1)n−1
(c) lim un = 0 (d) lim un > 1 Ans: (b) 18. The series is conditionally conver-
n=1
nk
12. If un is the nth term of a positive series such that gent if
lim un1/n = l then the statement which is true (a) k = 1 only (b) k > 1
among the following is
Question Bank A-11
(c) 0 < k < 1 (d) k = 0 Ans: (c) 2 3 4 5

(a) − 2 + 2 − 2 + ...
%∞ 2 1 2 2 3 4
1 −n −n 1 1 1 1
19. The series (−1) n−1
n+ is (b) 2 − 2 + 2 − 2 + . . .
n=1
n 1 2 3 4
1 1 1 1
(a) convergent (c) √ − √ + √ − √ = . . .
(b) divergent 1 1 2 2 3 3 4 4
1 1 1 1
(c) absolutely convergent (d) 2 + 2 + 2 + 2 + . . . Ans: (a)
(d) conditionally convergent Ans: (c) 1 2 3 4
26. The series which is absolutely convergent
%
∞
− k 2 +3k+3
20. The series n is convergent if k among the following.
1 1 1
n=1 (a) 1 + 2 + 3 + 4 + . . .
(a) ∈
/ (−2, −1) (b) ∈
/ [−2, −1] 3 3 3
(c) ∈ (−2, −1) (d) ∈ [−2, −1] Ans: (b) 1 1 1
(b) −1 + 2 − 3 + 4 . . .
3 3 3
%
∞
1 1 1
21. The series (3x)n is convergent if (c) 1 − + − + . . .
n=0 2 3 4
1 1 1 1 1
(a) −1 ≤ x ≤ 1 (b) − < x < (d) 1 + + + + . . . Ans: (b)
3 3 2 3 4
1 1
(c) − ≤ x ≤ (d) −3 < x < 3 Ans: (b) %
∞ %
∞
3 3
27. Between the two series (2x)n and (3x)n
22. Which one of the following is not convergent? n=1 n=1
1 1 1 one is convergent and the other is divergent if
(a) √ + √ + √ + . . .
1 1 2 2 3 3 1
1 1 1 (a) |x| > (b) 2 < |x| < 3
(b) 1 − 2 + 2 − 2 + . . . 2
1 1 1
2 3 4 (c) < |x| < (d) |x| < Ans: (c)
1 1 1 3 2 3
(c) 1 − + − + . . .
2 3 4 %
∞
(d) 1 + x2 + x4 + . . . (0 < x < 1) Ans: (c) 28. If the series
2
n−(k−4) is convergent then
n=1
%
∞
1
23. The series (a) k ∈
/ (3, 5) (b) k ∈ (3, 5)
n=2
n(log n)p (c) k ∈
/ [3, 5] (d) k ∈ [3, 5] Ans: (c)
(a) converges for all real p Hint: (k − 4) > 1 ⇒ (k − 4 + 1)(k − 4 − 1) >
2
(b) diverges for all real p 0 ⇒ (k − 3)(k − 5) > 0

(c) converges for all positive values of p
%
∞
(d) converges for all p > 1 Ans: (d) 29. If the series
2
n−k is convergent then
n=1
%
∞
1
24. The series is convergent then (a) k ∈
/ (−1, 1) (b) k ∈
/ [−1, 1]
n=1
n2k 2 −3k+1 (c) k ∈ (−1, 1) (d) k ∈ [−1, 1] Ans: (b)
(a) k < 0 or k > 3/2
(b) k > 0, k < −3/2 %
∞
2
30. If the series n−(k−a) is convergent then
(c) k ∈ (0, 3/2)
n=1
(d) k ∈ [0, 3/2] Ans: (a)
(a) k ∈ (a − 1, a + 1)
25. The series which is conditionally convergent (b) k ∈ [a − 1, a + 1]
among the following. (c) k ∈
/ (a − 1, a + 1)
(d) k ∈
/ [a − 1, a + 1] Ans: (d) (c) (ī + j̄ + k̄)/3
1
%
∞
2 −5k+9) (d) √ (ī + j̄ + k̄) Ans: (a)
31. The series n−k/(k is 3
n=1
4. If φ(x, y, z) = x2 +y2 +z 2 −3xyz then curl (grad
(a) convergent for all real k
φ) =
(b) divergent for all real k
1 (a) īx + j̄y (b) j̄y + k̄z
(c) convergent for all k ∈ − , 1 (c) 0̄ (d) īx + k̄z Ans: (c)
11
1
(d) divergent for all k ∈ − , 1 Ans: (b) 5. ∇r =
11
(a) r̄ (b) r̄/r
Hint: The series converges ⇔ (c) r̄/r 2 (d) r̄/r 3 Ans: (b)
2 k 2 − 5k + 9
k/ k − 5k + 9 > 1 ⇔ <1⇔
k 6. The directional derivatives of φ(x, y, z) = xy +
(k − 3) < 0 which is impossible
2
yz at P = (1, 1, 0) in the direction of ē = ī −
%
∞
3j̄ + 5k̄ is
32. The series n−(2k−1)/(k−2) is convergent if and √ √
(a) 1/√35 (b) 2/√35
n=1
only, (c) 3/ 35 (d) 4/ 35 Ans: (c)
(a) k > 1 (b) k > −1
(c) k < 1 (d) k < 0 Ans: (b) 1
7. ∇ = −r̄/r k ⇒ k =
r
(2k − 1)
Hint: The series converges ⇔ > (a) 0 (b) 1 (c) 2 (d) 3 Ans: (d)
k −2
1 ⇔ 2k − 1 > k − 2 ⇔ k > −1
8. If φ(x, y, z) = x2 yz then ∇φ|p where p =
%
∞
33. The series n−(k−a)
2 /(k−b)2
is convergent and (1, −2, 1) is
n=1 (a) −2ī+ j̄−2π (b) −4ī + j̄ − 2k̄
if only if (c) −ī+ j̄−2k̄ (d) −4ī − j̄ − 2k̄ Ans: (b)
a+b
(a) k > (b) k > 0
2 9. The unit normal to the surface φ(x, y, z) = xy +
a+b
(c) k < 0 (d) k < Ans: (a) yz + zx = 3 at P = (1, 1, 1) is
2 √
(a) (ī + j̄ + k̄)/ √3
(b) (ī − j̄ + k̄)/√ 3
Chapters 10, 11 Vector Calculus
(c) (ī + j̄ − k̄)/ √3
1. ∇(xyz) = (d) (−ī + j̄ + k̄)/ 3 Ans: (a)
(a) yz ī + zxj̄ + xz k̄

(b) yz ī + zxj̄ + yz k̄ 10. Curl x2 ī + y2 j̄ + z 2 k̄
(c) yz ī + zxj̄ + xyk̄ (a) ī + j̄ + k̄ (b) j̄ + k̄
(d) xzi + yz j̄ + zxk̄ Ans: (c) (c) ī + k̄ (d) 0̄ Ans: (d)
2. ∇(log r) =
11. Curl xyī + y2 j̄ − yz k̄ =
(a) r̄/r (b) r̄/r 2
(c) 2r̄/r (d) r̄ Ans: (b) (a) z ī + yj̄ (b) −z ī + xk̄
(c) −z ī − xk̄ (d) −z ī + yj̄ Ans: (c)
3. ∇(x + y + z) =
12. If ā is a constant vector then ∇ × (ā × r̄) =
(a) ī + j̄ + k̄
(b) 2(ī + j̄ + k̄) (a) ā (b) 2ā (c) 3ā (d) 0̄ Ans: (b)
Question Bank A-13
13. ∇ · r̄ = (a) 0 (b) 2F̄

(a) 1 (b) 2 (c) 3 (d) 0 Ans: (c) (c) F̄ (d) F̄.d r̄ Ans: (d)
C
14. If ∇ 2 (r n ) = 6 then n =
25. By
Green’s theorem the value of the integral
(a) 0 (b) 1 (c) 2 (d) 3 Ans: (c)
2xy − x2 dx + (x2 + y2 )dy where C is the
15. If ∇ 2 (r n ) = 0 then n = 0 is c
curve bounded by y = x2 and y2 = x is
(a) 1 (b) 2 (c) −2 (d) −1 Ans: (d)
(a) 2 (b) 1 (c) 0 (d) 12 Ans: (c)
1
16. ∇ 2 =
r 26. ∇ 2 r 3 =
1 1 (a) r (b) 3r (c) 6r (d) 12r Ans: (d)
(a) − 2
(b) 2 (c) 1 (d) 0 Ans: (d)
r r
17. ∇ 2 (log r) = 27. If F̄ is a solenoidal vector then curl(curl F̄) =
(a) ∇ 2 F̄ (b) 0
1 1 1 2
(a) − (b) (c) (d) Ans: (c) (c) −∇ 2 f¯ (d) F̄ Ans: (c)
r r r2 r3
18. (ē · ∇)r̄ = 28. If grad φ = yī + xj̄ then φ =
(a) ē (b) r̄ (c) r̄/r (d) 0 Ans: (a) (a) x + y (b) x − y
1 2
(c) xy + c (d) x + y2 Ans: (c)
19. Ifφis a harmonic function (i.e.∇ 2φ = 0) thenφis 2
(a) solenoidal only
(b) irrotational only 29. If F̄ = (2+y)ī+axj̄+2z k̄ is irrotational then a =
(c) neither solenoidal nor irrotational (a) 1 (b) 2 (c) 3 (d) 4 Ans: (a)
(d) solenoidal and irrotational Ans: (d)
30. If f¯ (x, y, z) = xy2 ī + 2x2 yj̄ − 3ayz k̄ at (1, 1, 1)
20. (ē × ∇) × r̄ = is solenoidal then a =
(a) 1 (b) 0 (c) 2ē (d) −2ē Ans: (b) (a) 0 (b) 2 (c) 1 (d) −1 Ans: (c)

21. (ē × ∇) × r̄ =
31. r̄ · n̄ds =
(a) 2ē (b) −2ē (c) ē (d) −ē Ans: (b) S
22. If f¯ (x, y, z) = (2x + y + az)ī + (bx + 2y + 3z)j̄ + (a) v (b) 2v (c) 3v (d) 4v Ans: (c)
(2x + cy + 3z)k̄ is irrotational then (a, b, c) =
(a) (1, 2, 3) (b) (3, 2, 1) 32. The vector f (r)r̄ is
(c) (2, 1, 3) (d) (3, 1, 2) Ans: (c) (a) irrotational
(b) solenoidal
23. If f¯ = F1 (y, z)ī + f2 (z, x)j̄ + f3 (x, y)k̄ then f¯ is (c) both irrotational and solenoidal
(a) irrotational only (d) none of the above Ans: (a)
(b) solenoidal only
(c) both irrotational and solenoidal
(d) neither irrotational nor solenoidal Ans: (b)
Chapter 12 Laplace Transforms
1. If a function f (t) defined for t ≥ 0 is piecewise
24. For any closed surface S curl F̄ · n̄ds = and continuous over every finite interval and is
of exponential order then f (t) is said to be of 1 a

(c) (d) Ans: (a)
class s2 − a2 s2 − a2
(a) B (b) C (c) A (d) D Ans: (c) 8. L{cos at} =
a 1
(a) 2 (b)
2. Given that f (t) is a function defined on s + a2 + a2
s2
s s
[a, b] (a > 0) such that (c) 2 (d) 2 Ans: (c)
(i) f (t) is piecewise continuous on [a, b]. s + a2 s − a2
(ii) f (t) is of exponential order i.e. 9. L{sin hat} =
lim e−st f (t) = 0. a 1
t→∞ (a) 2 (b) 2
(a) (i) and (ii) are necessary conditions for the s −a 2 s − a2
s s
existence of Laplace transform of f (t). (c) 2 (d) 2 Ans: (a)
s − a2 s + a2
(b) (i) is necessary (ii) is sufficient conditions
for the existence of Laplace transform 10. L{cosh at} =
of f (t). 1 s
(a) 2 (b) 2
(c) (i) and (ii) are sufficient conditions for the s + a2 s + a2
existence of Laplace transform of f (t). 1 s
(c) 2 (d) 2 Ans: (d)
(d) (i) is sufficient and (ii) is necessary s −a 2 s − a2
conditions for the existence of Laplace
transform of f (t). Ans: (c) 11. n N ⇒ L{t n } =
(a) n!/sn (b) n!/sn+1
3. Among the following, the function which is not (c) (n + 1)!/s n
(d) (n + 1)!/sn+1 Ans: (b)
of exponential order is
2 12. If H (t − a) is the Heaviside’s unit step function,
(a) et (b) t 2
L{Ht − a)} =
(c) cosh t (d) none of the above Ans: (a) (a) e−as/s (b) eas /s
−as
(c) e /(s − a) (d) eas /(s − a) Ans: (a)
4. L{C1 f1 (t) + C2 f2 (t)} = C1 L{f1 (t)} + C2 L{f2 (t)}.
This property in respect of Laplace transforms 13. If δ(t −a) is the Dirac delta function then
is called the L{δ(t −a)} =
(a) shifting property
(b) distributive property (a) eas (b) e−as (c) e−s/a (d) es/a Ans: (a)
(c) symmetricity property
(d) linearity property Ans: (d) 14. If L{f (t)} = f¯ (s) then by the first shifting
property
(a) L{f (eat f (t))} = f¯ (s − a)
5. L{1} =
(b) L{e−at f (t)} = f¯ (s − a)
1 1
(a) 1 (b) (c) 0 (d) − Ans: (b) (c) L{et/a f (t)} = f¯ (s − a)
s s (d) L{eat f (t/a)} = f¯ (s − a) Ans: (a)
6. L{eat } =
a a 15. If L{f (t)} = f¯ (s) then by the change of scale
(a) (b)
s−a s property
1 1
(c) (d) Ans: (c) (a) L{f (at)} = f¯ (s/c)
s−a s+a 1
(b) L{f (at)} = f¯ (s/a)
7. L{sin at} = a
s¯
a 1 (c) L{f (at)} = f (s/a)
(a) 2 (b) a
s + a2 s 2 + a2 (d) none of the above Ans: (a)
Question Bank A-15
/
f (t − a), (t > a) 1 s−b
16. If L{f (t)} = f¯ and F(t) = (b) log
0 (t < a) 2 s−a
Then (by the second shifting property) (c) log[(s − b)/(s − a)]
L{F(t)} =
1 s−a
(a) eas f (s/a) (b) e−as f¯ (s/a) (d) log Ans: (c)
2 s−b
(c) eas f¯ (s) (d) e−as f¯ (s) Ans: (d) / ,
sinh t
24. L =
17. If L{f (t)} = f¯ (s) then L{f (t)} = t
(a) sf¯ (s) − f (0) (b) sf¯ (s) − f (0)
s+1 1 s+1
(c) f¯ (s) − f (0) (d) sf¯ (s) − f (0) Ans: (a) (a) log (b) log
s−1 2 s−1
s−1 1 s−1
18. l{te2t } = (c) log (d) log Ans: (b)
s+1 2 s+1
1 1 / ,
(a) (b) −
(s − 2)2 (s − 2)2 −1 1
25. L =
1 s
(c) (d) none of the above Ans: (b)
(s − 1)2 (a) 1 (b) et (c) t (d) sin t Ans: (a)
19. If L{f (t)} = f¯ (s) then, by the initial value / ,
−1 1
theorem, lim f (t) = 26. L
t→0 s−a
(a) e−at (b) et
(a) lim f¯ (s) (b) lim f¯ (s)
s→0 s→∞ (c) eat (d) ret−a Ans: (c)
(c) lim sf¯ (s) (d) lim sf¯ (s) Ans: (d)
s→0 s→∞
27. If L−1 {f¯ (s)} = f (t then L−1 {sf¯ (s)} =
20. If L{f (t)} = f¯ (s) then, by the final value ∞
theorem lim f (t) = lim sf¯ (s) = (a) f (t − a) (b) f (t)dt
t→∞ s→0
(a) lim sf¯ (s) (b) lim f¯ (s) t
s→∞ s→∞ (c) f (t) (d) none of the above Ans: (c)
(c) lim sf¯ (s) (d) lim f¯ (s) Ans: (c)
s→0 s→0
¯
28. If L−1 {f¯ (s)} = f (t) then L−1 [ f (s) ]=
21. If L{f (t)} = f¯ (s) then L{t n f (t)} =
s
t ∞
dn dn ¯ 1
(a) (−1)n nf¯ (s) (b) f (s) (a) f (u)du (b) f (u)du
ds dsn u
n
d dn 0 t
(c) − nf¯ (s) (d) (−1)n−1 nf¯ (s) Ans: (a) t2 t
ds ds
(c) f (u)du (d) f (u)du Ans: (d)
22. If L{f (t)} = f¯ (s) then L{ 1t f (t)} = t1 0
∞ ∞
/∞ ,
(a) f¯ (s)ds (b) f¯ (s)ds −1

¯f (s) = f (t) then L−1

29. If L f (u)du =
0 s s
∞ ∞
(a) tf (t) − f (0) (b) tf (t)
(c) (d) f¯ (s)ds Ans: (b) (c) f (t)/t (d) f (t) Ans: (c)
−∞ a
/ ,
/ , −1 1
e −e at bt 30. L =
23. L = s(s − 1)
t
(a) et − 1 (b) et + 1
s−a (c) et + 1 (d) −et + 1 Ans: (a)
(a) log
s−b
Fill in the Blanks
Chapter 1 Ordinary Differential x c1 1 1

Equations Hint: y1 = + ⇒ y1 = + c1 /x2
2 x x 2
1. The differential equation of the family of para- 1 c1 1
y 2 = − 2 y 2 + y1 = 1
bolas with vertex at the origin and the y-axis 2 x x
as its axis is _______. Ans: xy = 2y
d dy
12. The general solution of x = 1 is
dx dx
2. The differential equation of the system of unit
_______. Ans: y − x = A log x + B
circles is _______. Ans: (1 + y12 )3 = y22
dy 2x − 2y + 3
dr θ 13. The general solution of = is
3. +rtan = 0 is the differential equation of the dy 2x − 4y + 1
dθ 2 _______. Ans: x2 +2y2 −2xy +3x −y +c = 0
family of curves whose equation is _______.
Ans: r = a(1 + cos θ) Hint: 2(xdy+ydx)−(4y−1)dy−(2x+3)dx=0
⇒ 2d(xy) − 2y2 − x2 + y − 3x + c = 0
4. The differential equation (x +y)(dx −dy) = dx +
dy has the general solution _______. 14. The general solution of (x + y)(ydx − xdy)
Ans: x + y = Cex−y = y2 dx is _______. Ans: (1 + x) = Aex/y
5. The orthogonal trajectories of x2 y = e are dθ
_______. Ans: 2y2 − x2 = a 15. If ‘θ’ satisfies the differential equation =
dt
−kθ (k: constant) and θ = 50 when t = 0, then
dy x + 2y − 3 θ = _______. Ans: 50e−kt
6. The general solution of = is
dx 2x + y − 3
c(y − x)3 = _______. Ans: c(x + y − 2) 1
16. The type of the differential equation y + y = y2 x
x
7. Self-orthogonal system of curves is _______. is _______ equation. Ans: Bernoulli’s
x2 y2 dy
Ans: 2 + 2 =1 17. The type of the differential equation −
a +λ b +λ dx
8. The orthogonal trajectories of r = a(1+cos θ) are tan y
= (1 + x)ex sec y is _______ equation.
_______. Ans: r = a(1 − cos θ) 1+x
Ans: reducible to linear
9. The differential equation representing the curves dy
orthogonal to the st. lines through the ori- 18. The differential equation (x + 2y3 ) y is
dx
gin is _______. Ans: yy1 + x = 0 _______. Ans: linear in x
dy y(1 − x)
10. The general solution of + = 0 19. It is given that the I.F. of (y + xy2 )dx − xdy = 0
dx x(1 − y) is a function of y only then it is _______.
is Axy = _______. Ans: ex+y
1
Ans: 2
11. The differential equation whose general solution y
x2 20. It is given that the I.F. of (x2 y2 + xy + 1)ydx
is y = + c1 log x + c2 is _______.
4 +(x2 y2 − xy + 1)xdy = 0 is of the form (xy)
1 1
Ans: y2 + y1 = 1
x then it is _______. Ans: 2 2
xy
Question Bank A-17
21. The complete solution of y2 dx + 2xydy = 1

8. A particular value of cos3 x is _______.
x2 dy +2xydx is f (x, y) = c where f (x, y) = D
1
_______. Ans: xy(y − x) Ans: (sin 3x + 3 sin x)
12
Hint: Grouping the term d(xy2 ) = d(x2 y) ⇒ 1
9. A particular value of 2 x is _______.
xy(y − x) = c D −4 x
Ans: −
22. The type of the equation (3x−2y+5)dx+ (−2x+ 4
1 − cos 3x
4y + 1)dy = 0 is _______. Ans: exact 10. cos 3x = _______. Ans:
D2 − 9 18
1
23. The complete solution of (3x2 − 2xy − 5)dx− 11. x = _______. Ans: −x
(x2 + y2 − 2y)dy = 0 is _______. D2 − 1
1 2x − 1
Ans: 3x3 − 3x2 y − y3 + 3y2 − 15x + c = 0 12. x = _______. Ans:
D+2 4
Group terms and write d(x − 5x)+
3
Hint:
1 x
1
d y2 − y3 − d(x2 y) = 0 13. cos ax = _______. Ans: − cos ax
3 D 2 + a2 2a
1
14. cos x = _______. Ans: sin x
D2 + D + 1
Chapter 2 Linear Differential 1 1
Equations of Second Hint: 2 cos x = 2 cos x
D +D+1 −1 +D+1
and Higher Order
1
1. The solution of the I.V.P. y + y = 0, y(0) = 2, = cos x = cos xdx = sin x
D
y (0) = −1 is y = _______. 1
Ans: 2 cos x − sin x 15. e sin x = _______. Ans: −ex sin x
x
(D − 1)2
2. The general solution of (9D2 +6D+1)y = 0 is 16. The P.I. of the equation (D2 + 2D + 1)y =
_______. Ans: (c1 + c2 x)e−x/3 e−x cos x is _______. Ans: −e−x cos x
1 1
3. If ex and e−x are two L.I. solutions of (aD2 + Hint: e−x cos x = e−x 2 cos x
bD + c)y = 0, then _______. (D + 1)2 D
Ans: b = 0, c/a = −1 = −e−x cos x
4. If sin x and cos x are two L.I. solutions of (aD2 + 1 x x

17. ex = _______. Ans: e
bD + c)y = 0, then_______. D2 −1 2
Ans: b = 0, c/a = 1 Hint: Since f (1) = 0 replace
5. The two L.I. solutions of (D2 + a2 )y = 1ex xex xex
= = by F2 on p. __
0 are _______. Ans: cos ax, sin ax f (D) f (1) 2x
√ √ √ 1 x2
6. If e−ax/ 2 (cos ax/ 2, sin ax/ 2) are two L.I. 18. e 2x
= _______e 2x
. Ans:
solutions of (D4 + a4 )y = 0, then the other two (D − 2)2 (D − 1) 10
1
L.I. solutions are _______.
√ √ √ Hint: e2x
(D − 2)2 (D − 1)
Ans: eax/ 2 (cos ax/ 2, sin ax/ 2)
x2 e2x x2 e2x
1 = = by F3 on p. __
7. A particular value of e2x is 2! 5 10
D2 − 3D + 2
8614 √3x
_______. Ans: e2x (x − 1) 19. If f (D)y = [D2 −1]31 = 5614, then e =
f (D)
1 1 √ √ √ x3
Hint: e2x = e2x 3x3 e 3x
= _______e 3x
. Ans: √
(D − 1)(D − 2) D(D + 1) 288 3
Hint: f (D) = D6 − 3D 4
+ 9D2 − 1 d 2y dy
3 3
√
3x
√ 30. The equation x2 − x + y = 2 log x when
xe xe x3 e 3x dx 2 dx
√ = √ √ = √ transformed to linear equation with constant
f 3 (6.5.4( 3)2 −3.4.3.2) 3 288 3 coefficients becomes _______.
1 1 Ans: (θ − 1)2 y = 2z
20. cos ax = _______. Ans: cos 2x − 2
D 4 − a4 16 1
1 1 31. (θ 2 −4θ +1)−1 sin z = _______. Ans: cos z
Hint: 2 sin 2x = (1 − cos 2x) 4
D −4 2(D − 4)
2
1 1 1 1 1 1
= e0.x− = (−2+cos 2x) 32. yp = x3 = _______. Ans: −x3 − 6
2 D −4
2 2−2 −4 18
2 D3 −1
33. The particular solution of (D3 − 1)y = 0
21. If f (D)y = (D − 3) (D − 1)y = e , then
2 3x is ______. Ans: ex
1 3x 1 1 x2
yp = e =_______e3x . Ans: · · 34. The particular solution which is common to
f (D) 2 2 2!
1 e3x both the equations (D3 − 1)y = 0 and
22. 2 e3x = _______. Ans: (D3 − D2 )y = 0 is _______. Ans: ex
D −4 5
1 xe3x 35. The P.I. of (D4 − 1)y = x cos x is _______.
23. 2 e3x = _______. Ans:
D −4 6 −ex cos x
Ans:
1 cosh 2x 5
24. 2 cosh 2x = _______. Ans:
D +1 10 1 1
√ Hint: 4 ex cos x = ex cos x
3x D −1 (D + 1)4 − 1
25. If ex/2 cos is a solution of (D3 +
2 1
1)y = 0 another linearly independent solution = ex 4 Re eix
√ D + 4D + 6D + 4D
3
3x
is _______. Ans: ex/2 sin 1
2 = ex Re eix 4
i + 4i3 + 6i2 + 4i
26. For (D2 + 5)y = x sin 2x P.I. is _______. 1 1
= ex Re eix = − ex cos x
Ans: x sin 2x − 4 cos 2x 1−6 5
dy dy
27. x2 + 3x = 0 has the general solution 36. The boundary value problem y + ky = 0,
dx2 dx y (0) = 0, y (π) = 0 has nontrivial solution for
1
y = _______. Ans: c1 + c2 2 _______. Ans: all k > 0
x
dy dy d 2y dy
28. The P.I. of x 2 + 3 = 1 is _______. 37. The general solution of x + = 1 is
dx dx dx2 dx
1 y = _______. Ans: x + c1 log x + c2
Ans: x
3
Hint: [θ(θ − 1) + 30]y = x = ez 38. The equation for which the method of variation
1 1 1 of parameters is applicable is _______.
⇒ 2 ez = ez = x
θ + 2θ 3 3 Ans: y + ay = tan ax
d 2y dy 39. _______ is a Euler–Cauchy equidimentional
29. To reduce the equation x2 −x +y = d 2y dy
dx2 dx equation. Ans: x2 −x +y = x
2 log x to a linear equation with constant coeffi- dx dx
cients the change of independent variable to be
made from x to z is by putting x = _______. 40. _______ is a Legendre’s linear differential equa-
tion. Ans: (1 + 2x)y + 3(1 + 2x)y = x
Ans: ez
Question Bank A-19
41. _______ is a Bernoulli’s equation. 2. Rolle’s theorem is not applicable to f (x) = |x|
1
in [−1, 1] because it fails to satisfy _______.
Ans: y − y = y2
x Ans: the condition of differentiability in
42. The set of L.I. solutions of (D4 − D2 )y = [−1, 1]
0 is _______. Ans: {1, x, ex , e−x } 3. The example which shows that the conditions
of Rolle’s theorem are only sufficient but not
43. Every solution of y + ay + by = 0 (a, b necessary is f (x) = _______.
constants) approaches to zero as x → ∞ 1
provided _______. Ans: a > 0, b > 0 Ans: sin in [−1, 1]
x
1 1
44. The general solution of (x2 D2 + xD)y = 0 is Hint: f (x) = 2 cos = 0 at x = Cn
x x
y = _______. Ans: C1 + C2 (log x) 2
= ∈ (−1, 1)
(2n + 1)π
45. The general solution of an nth order differential 1
equation contains _______. though condition (i) fails to hold as sin is not
x
Ans: exactly n L.I. solutions continuous at 0 in [−1, 1]
46. By a change of the independent variable x to 4. The point where f (x) = | sin x| is not derivable
z by the transformation x = ez , the equation in (0, 2π) is x = _______. Ans: π
(x2 D2 + xD + 1) = log x sin(log x) becomes
_______. Ans: (θ 3 − 3θ 2 + 3)y = ez 5. The third condition of Rolle’s theorem fails to
hold in the case of function f (x) = _______.
47. By the transformation x = ez , the equation, Ans: sin x in [0, 3π/4]
d 3y 1 1
+ 3 y = 2 becomes _______.
dx3 x x Hint: f (x) = sin x in [0, 3π/4]
Ans: (θ 3 − 3θ 2 + 3)y = e2 3π 1
f (0) = 0 ± f (3π/4) = sin = −√
48. The complete solution of (θ 2 − 4)y = 0 where 4 2
d 1
θ = x is _______. Ans: c1 x2 +c2 2 through f ∈ C[0, 3π/4], ∈ D[0, 3π/4]
dx x
49. After the transformation x = ez , the equation 6. Rolle’s condition of differentiability fails to hold
d in respect of the function f (x) = _______.
has become (θ 2 − 4)y = 0 where θ = x
dx Ans: |x| in [−1, 1]
then the original differential equation in the
independent variable x is _______. 7. The value of c of Rolle’s theorem for f (x) =
Ans: x2 y + xy − 4y = 0 x2 − 1 in [−1, 1] is c = _______. Ans: 1/2
50. The two L.I. solutions of (D2 + 1)y = cosecx 8. The value of c of Rolle’s theorem for f (x) =
are u(x) = cos x, upx = sin x then the Wron- e−x sin x in [0, π] is c = _______. Ans: π/4
strian w((u, v) = _______. Ans: 1
9. The value of c of Rolle’s theorem for f (x) =
Chapter 3 Rolle’s Theorem and x(x + 3)e−x/2 in [−3, 0] is c = _______.
Mean Value Theorem Ans: −2
1. If f be a function defined on [a, b] such that 10. The reason why Rolle’s theorem is not applica-
(i) f ∈ C[a, b], (ii) f ∈ D(a, b) and (iii) f (a) = ble to f (x) = tan x in [0, π] is that at x = π/2
f (b) then C ∈ (a, b) exist such that f (c) = 0. _______.
This is the statement of _______ theorem. Ans: f is not defined, not continuous, not
Ans: Rolle’s derivable
11. The value of c of Rolle’s theorem for φ(x) = 20. The value of c of Rolle’s theorem for f (x) =
1 0 1 ex (sin x − cos x) in [π/4, 5π/4] is c = _______.
eπ 0 −1 in [0, π] is c=_______. Ans: π
ex sin x cos x
Ans: π/2 21. ‘c’ of Cauchy’s mean value theorem for the
1
f (0) g(0) h(0) functions f (x) = lnx and g(x) = in [1, e]
x
Hint: φ(n) = f (π) g(π) h(π) e
is c = _______. Ans:
f (x) g(x) h(x) e−1
Hint: By CMV theorem
= sin x(1 + eπ ) (x) = 0
ln e − ln 1 1 e
⇒ x = π/2 = (−c2 ) = −c ⇒ c =
1
e
− 1 c e − 1
√ √ for f (x) =
12. The value of c of Rolle’s theorem 22. ‘c’ of Cauchy’s Mean Value Theorem for the
2x3 + x2 − 4x − 2 in [− 2, 2] is c = functions f (s) = sin x, g(x) = cos x in (a, b) is
2
_______. Ans: or −1 b+a
3 c = _______. Ans:
13. The value of c of Lagrange’s Mean Value 2
Hint: f (x) = sin x, g(x) = cos x
Theorem for f (x) = x2 − 3x + 2 in [1, 3] is
−(− cos c) sin a − sin b
c = _______. Ans: 2 cot c = =
sin c cos a − cos b
2 − 0
Hint: f (x) = 2x − 3 = =1 2 sin b−a cos b+a
b+a
3−1 = 2 2
= cos
b−a b+a
⇒ x = 2 ← (1, 3) a = 1, b = 3 2 sin 2 sin 2 2
b+a
14. An approximate value of root of f (x) = x4−12x+ ⇒c=
2
7 = 0 near x = 2 is _______. Ans: 2.05
23. Taylor’s series expansion of sin x at x = 0 is
Hint:
f (x) = 4x − 12 by L.M.V.T.
3 x3 x5
sin x = _______. Ans: x − +
f (ath) f (a) + hf (a + θh) 3! 5!
f (a) −1 24. Taylor’s series expansion of cos x at x = 0 is
⇒h=− =− = 0.05 x2 x4
f ()a 20 cos x = _______. Ans: 1 − + −···
2! 4!
Approx. root = 2.05
15. An approx. value of root of f (x) = x3 − 2x − 5 Chapter 4 Functions of Several
in (2, 3) is _______. Ans: 2.1 Variables
dy
1. If x3 + y3 + 3xy = 0, then = _______.
16. The value of c of Rolle’s theorem for f (x) = dx
sin x Ans: (x2 − y)/(y − x2 )
in [0, π] is c = _______. Ans: π/4
ex 2. The degree of the homogeneous function
√ √
17. The value of c √ of Rolle’s theorem for f (x) = x+ y
z = is _______. Ans: –1/2
x3 − 12x in [0, 2 3] is c = _______. Ans: 2 x+y
3. If u = x2 + y2 + z 2 , x = et , y = et sin t,
18. The values of c of Rolle’s theorem for f (x) = du
sin x in [0, 2π] is c = _______. z = et cos t, then = _______. Ans: 4e2 t
dt
Ans: π/2, 3π/2
x ∂u ∂u
19. Rolle’s theorem is not applicable for f (x) = x2 4. If u = xy
2
then x + y = _______.
in [1, 2] because _______. Ans: f (1) ± f (2) y ∂x ∂y
Ans: 3u
Question Bank A-21
ux
5. If u = ex sin y, then = _______. 17. If u = x + y + z, uv = y + z, uvw = z, then
uy ∂(x, y, 2)
Ans: tan y = _______. Ans: u2 v
∂(u, v, w)
6. If u(x, y, z) = ln(x + y + z − 3xyz), then
3 3 3
x+y
ux +uy +uz = _______. Ans: 3(x+y+z)−1 18. If u = , v = tan−1 x + tan−1 y, then u
1 − xy
7. If z = (y2 − x2 )ln(x − y), then yzx + xzy = and v are _______.
_______. Ans: x2 − y2 Ans: functionally dependent
19. If x = r sin θ sin , y = r sin θ sin θ, z = r cos θ,
8. If xz = cos y and x = u2 − v, y = ev , then ∂(x, y, z)
∂z 1 then = _______. Ans: r 2 sin θ
= _______. Ans: 2 (cos y − xy sin y) ∂(u, v, w)
∂v x
x 2
+ y 2 20. If x = v + w, y = w + u, z = u + v,
9. If u = sin−1 then xux + yuy = ∂(x, y, z)
x+4 then = _______. Ans: 2
_______. Ans: tan u ∂(u, v, w)
Hint: f (x) = sin x, g(x) = cos x x2
21. The function f (x, y) = , (x, y) =
−(− cos c) sin a − sin b x2 + y2
cot c = = (0, 0) and f (0, 0) = k becomes continuous at
sin c cos a − cos b
the origin if k = _______. Ans: 0
2 sin b−a cos b+a b+a
= 2 2
= cos
2 sin 2 sin b+a
b−a
2
2 22. If u = en cos y, then ux2 + uy2 = _______.
b+a Ans: e2x
⇒c=
2
10. For polar coordinates x = r cos θ, y = r sin θ, Chapter 5 Radius of Curvature
∂(x, y) 1 1. The radius of curvature at θ = π/6 for the
= _______. Ans:
∂(rθ) r x2 y2 √
curve + = 1 is _______. Ans: 7 7/8
11. For cylindrical coordinates x = r cos θ, 12 4
∂(x, y, z)
y = r sin θ, z = 2, = _______. 2. If y1 and y2 of a curve at the origin are
∂(r, θ, z) Ans: r 2 and 2, then the y-coordinate of the cen-
12. If functions u and v of independent vari- tre of curvature is _______. Ans: 7/2
ables x and y are functionally dependent, then
∂(u, v) 3. If the circle of curvature is x2 +y2 −4x−4y = 0,
= _______. Ans: 0
∂(x, y) then the centre of curvature is _______.
∂(u, v) ∂(x, y) Ans: (1, 1)
13. If u(1 − v) = x, uv = y then · =
∂(x, y) ∂(u, v) 4. The envelope of the family of st. lines x cos y+
_______. Ans: xy
y sin x = p is _______. Ans: y2 = 4px
u, v x, y
14. If u = J then J = _______. 5. The locus of the centre of curvature of
x, y x, v
1 a curve is called _______. Ans: evolute
Ans:
u
15. The stationary points of f (x, y) = x3 y2 (1−x−y) 6. The coordinates of centre of curvature of the
a part from 0 is _______. Ans: (1/2, 1/3) curve y = ex at (0,1) is _______. Ans: (–2,3)
16. The extreme values of f (x, y) = x3 y2 −x4 y2 −

7. The radius of curvature at the origin of the curve
1 1 1
x3 y3 occur at _______. Ans: , y = x4 − 4x3 − 18x2 is _______. Ans:
2 3 46
x
8. The envelope of family of st. lines cos θ+ Ans: imaginary
a
y
sin θ = 1 (θ, parameter a, b constants) is 8. A curve symmetrical about both the axes is sym-
b metrical about the _______. Ans: origin
x2 y2
_______. Ans: 2 + 2 = 1
a b 9. The folium of Descartes with equation x3 +y3 =
2
9. If c1 , and c2 are two curves such that c is the 3axy is symmetrical about the_______.
evolute of c1 then c1 is called _______. Ans: line y = x
Ans: involute
10. The folium of Descartes with equation x3 −y3 =
10. If a single parameter family of curves is 3axy is symmetrical about the _______.
given by am2 + bm + c = 0 then its Ans: the line y = −x
envelope is _______. Ans: b2 = 4ac
11. The curve a2 y = x3 (cubical parabola) is
11. The radius of curvature at (1,1) for the parabola symmetrical about the _______.
√
y = −x2 + x + 1 is ρ =_______. Ans: 2 Ans: origin
12. The curve y2 = kx3 (semicubical parabola) is
12. The radius of curvature of the ellipse x = a cos t,
symmetrical about the _______.
y = b sin t at a point ‘t’ is _______.
Ans: the x-axis
Ans: (a2 sin2 t + b2 cos2 t)3/2 /ab
13. The equation of the curve which passes through
Chapter 6 Curve Tracing the origin is _______.
Ans: y2 (a − x) = x2 (x + a)
1. When x and y are replaced simultaneously by
−x and −y the equation remains unaltered then 14. For the curve x2 y2 = a2 (y2 − x2 ), the
the curve is symmetrical about _______. origin is a _______. Ans: node
Ans: (0,0)
15. For the curve (cissoid of Diocles) y2 (2a − x) =
2. If an interchanging x and y the equation
x3 , the origin is a _______. Ans: cusp
remains the same then the curve is symmet-
rical about the_______. Ans: line y = x
16. The curve y2 (x − a) = x2 (x + a) is symmetri-
cal about _______. Ans: x-axis
3. If the constant term is absent from the equation
of a curve, then the curve_______.
17. The curve x2 (x2 + y2 ) = a2 (x2 − y2 ) is symmet-
Ans: passes through the origin rical about_______.
4. If at a point on a curve two or more tangents Ans: both x- and y -axes
exist it is called a _______ point.
18. The curve which has a node is _______.
Ans: multiple
Ans: x2 (x2 + y2 ) = a2 (x2 − y2 )
5. A double point is called a rode if the tangents
to the curve at that point are _______. 19. The curve which has a cusp is_______.
Ans: real and district Ans: y2 (2a − x) = x3
1
6. A double point is called a cusp if the tangents 20. The curve y = x + is symmetrical
to the curve at that point are _______. x
about the _______. Ans: origin
Ans: real and coincident
7. A point on a curve is called an isolated point if 21. The equation of the asymptote to the curve x3 +
the tangents to the curve are _______. y3 = 3axy is _______. Ans: y+x+a = 0
Question Bank A-23
22. The asymptote parallel to the x-axis of the curve 33. Let x = x(t), y = y(t) be parametric equa-
x2 y − 4x2 + 5 = 0 is given by _______. tions of a curve. If x(t) is even and y(t) is
Ans: y = 4 an odd function of t, then the curve is sym-
metrical about the _______. Ans: x-axis
23. The asymptote parallel to the y-axis of the curve
xy2 − 2y2 + y + 1 = 0 is given by_______. 34. Let x = x(t), y = y(t) be parametric equa-
Ans: x = 2 tions of a curve. If x(t) is odd and y(t) is
24. If the polar equation of a curve remains unal- an even function of t, then the curve is sym-
tered when θ is replaced by −θ, then the curve metrical about the _______. Ans: y-axis
is symmetrical about the _______.
35. The asymptote of y2 (2a − x) = x3 (cissoid of
Ans: initial line
Diocles) has equation _______. Ans: x = 2a
25. If the polar equation of a curve remains
unaltered when θ is replaced by (π − θ), 36. The asymptote of xy2 = a2 (a − x) is
then the curve is symmetrical about the ray the _______. Ans: y-axis
θ = _______. Ans: π/2
37. The line _______ is a line of symmetry for x3 +
26. If the polar equation of a cure remains unaltered y3 = 3(x + y)xy. Ans: y = x
π
when θ is replaced by − θ then the curve
2 38. The curve y = x3 − 3x2 − 9x + 9 has a
is symmetrical about the ray θ = _______. point of inflexion at _______. Ans: x = 1
Ans: π/4
27. If the polar equation of a
curve remains unal- 39. The curve r 2 = a2 sec 2θ has _______ as
its asymptotes. Ans: xy = 0
3π
tered when θ is replace by − θ , then the
2 40. The equation of the curve which is bounded by
curve is symmetrical about the ray θ = _______.
the rhombus with vertices x = = a, y = = b is
3π _______.
Ans: x 2/3 y 2/3
2 Ans: + (Hypocycloid)
a b
28. The curve r 2
= a cos 2θ (Lemniscate
2
41. The slope of the curve x3 + y3 = 3axy where
of Bernoulli) is symmetrical about the ray the line y = x meets it is _______. Ans: –1
θ = _______. Ans: 0
42. The x-axis is a double tangent to the curve
29. The curve r = cos 3θ (three-cusped rose) is _______. Ans: y2 (a − x) = x3
symmetrical about the ray θ = _______.
Ans: 0
Chapter 7 Applications of
30. The curve r = a sin 3θ (three-leafed rose ) is Integration
symmetrical about the ray θ = _______.
1. The integral formula for finding the length of a
Ans: π/1 plane curve between t1 and t2 whose equations
31. The cardioid whose equation is r = a(1+cos θ) are x = x(t), y = y(t) is given
by s =_______.
is symmetrical about the ray θ = _______. t2 2 2
dx dy
Ans: 0
Ans: +
t1 dt dt
32. The cycloid whose equations x = a(t+sin t), y = 2. The integral formula for finding the length of a
a(1−cos t) is symmetrical about the _______. plane curve r = f (θ) in polar coordinates from
Ans: y-axis θ1 to θ2 is given by s =_______.
2
t2 1 3/2 dx 3
dr Hint: x= y ⇒ = y1/2
Ans: r2 + dθ 2 dy 4
dθ
t1
1 2 1/2
dx
3. The integral formula for finding the length of a s= 1+ dy
dy
plane curve θ = f (r) in polar coordinates from 0
r1 to r2 is given by s = _______. 3/2 1

16 2 9
r2 2 = · 1+ y
dθ 9 3 16
Ans: 1 + r2 dr 0
dr 32 125 61
r1
= −1 =
4. The integral formula for finding the length of a 27 64 54
plane curve y = f (x) in cartesian coordinates 12. The length of the curve r = a(cos θ + √
sin θ),
from x1 , to x2 is given by s =_______. (0 ≥ 0 ≥ π) is _______. Ans: 2aπ
2
x2
dy 13. The length of the spiral r = eαθ from the pole
Ans: 1+ dx
x1 dx to the point (r, θ) is _______.
r
5. The integral formula for finding the length of a Ans: 1 + α2
α
plane curve x = f (y) is cartesian coordinates t3
from y1 to y2 is given by s=_______. 14. The length of a loop of the curve x = t 2 , y = t−
√ √ √3
y2 2 from t = − 3 to 3 is _______. Ans: 4 3
dx
Ans: 1+ dx √3 2 2
dy dx dy
y1
Hint: √ + dt
6. The element of length of a curve in parametric − 3 dt dt
form is ds =_______. √3
=2 (2t)2 + (1 − t 2 )2 dt
2 2 0
dx dy √ √3
Ans: + dt 3
1 √
dt dt =2 (1 + t 2 )dt = 2 t + t 3 =4 3
0 3 0
7. The element of length of a curve in polar coor- 15. The ratio in which the ray θ = 2π/3 divides
dinates is ds =_______. Ans: (dr)2 +(rdθ)2 the curve x = a(θ − sin θ), y = a(1 − cos θ)
is _______. Ans: 1:3
8. The process of determining the length of arc of 2
2π
a curve is called _______. Ans: rectification dx 2 dy
Hint: C1 = + dθ
0 dθ dθ
9. The process of determining a square equal in 2π
area to the area of a surface is called _______. =a (1 − cos θ)2 + (sin θ)2 dθ
Ans: quadrature 0
2π
θ
10. The perimeter of the semi-circle x2 + y2 = a2 , = 2a sin dθ
2
y ≤ 0 bounded by the diameter y = 0 is 0
2π
_______. Ans: (π + 2)a θ
= −4a cos = 8a;
2 0
11. The length of the arc of the
semi-critical para-
θ 2π/3
1 l2 = −4a cos
bola y3 = 4x2 from 0 to p = , 1 is _______. 2 0
2
61 1
Ans: = −4a − 1 = 2a
54 2
Question Bank A-25

l1 = 4l2 , so ray θ = 2π/3 divides l, in the ratio 2π π 3
Hint: v= r sin θdθ
1:3. 3 0

2πa3 π
16. The length of the curve x = t 3 − 3t, y = 3t 2 =− (1 + cos θ)3 d(cos θ)
from t = 0 to 1 is _______. Ans: 4 3 0
π
−2πa3 (1 + cos θ)4 8
1 2 = · = πa3
dx 2 dy 3 4 0 3
Hint: + dt
0 dt dt
1 21. The volume of the solid generated by revolving
= (3t 2 − 3)2 + (6t)2 dt the astroid x2/3 + y2/3 = a2/3 about the x-axis
0 is _______ πa3 . Ans: 32/105
1
1
t3 π2
=3 (t 2 + 1)dt = 3 +t =4 dx
0 3 0 Hint:V = 2π y2 dθ
0 dθ
π/2
17. The length of the cardioid r = a(1 + cos θ) is 32πa3
= 6πa3 sin7 θ cos2 θdθ =
_______. Ans: 8a 0 105
2 (∵ x = a cos3 θ, y = a sin3 θ; θ = 0 sin π/2)
θ2
dz
Hint: r2 + dθ
θ1 dθ 22. The volume of the solid of revolution of the
π plane region included between the curves y2
= 2a [(1 + cos θ)2 + sin2 θ]1/2 dθ = x3 and x2 = y3 about the x-axis is _______ π.
π
0
= 2a [2(1 + cos θ)]1/2 dθ Ans: 5/28

π
0 1
θ θ π Hint: v=π (x4/3 − x3 )dx
= 4a cos dθ8a sin 0 = 8a x=0
0 2 2 1
3 x4 5π
18. The arc of the upper half of the cardioid r = = π x7/3 − =
7 4 0 28
a(1 + cos θ) is bisected at θ = _______.
Ans: π/3 1
23. The area of the surface of a sphere of radius is
π/3 2
θ _______. Ans: π
Hint: Arc (up to θ = π/3) = 2a cos dθ
0 2
π/3 24. The surface area of the solid of the astroid
θ 1
= 4a sin = 4a · = 2a a = a cos3 t, y = a sin3 t about the x-axis is
2 0 2 12πa2
_______. Ans:
5
19. The volume of the solid of revolution of
the region bounded by y = x2 and x = a is 25. The surface area generated by the region of the
π x
_______. Ans: a5 curve y = c cos h (catenary) between (0, c)
5 c
a a and (x, y) about the x-axis is πc _______.
π 5
Hint: v = π y dx = π
2
x dx = a
4
5 c 2x
0 0
Ans: x +
sin h
2 c
20. The volume of the solid of revolution of the 26. The length of the curve x = a(t + sin t), y =
cardioid r = a(1 + cos θ) about the initial line a(1 − cos t) from t=0 to π is _______.
θ = 0 is _______. Ans: 8πa3 /3
Ans: 4a
27. The length of the curve y = s3/2 from x = 0 to 5 4 25

log − log = log
56 6 5 24
x = 4/3 is _______. Ans:
27
a x
x dy dx
28. The volume of the solid of revolution of area 5. = _______.
bounded by y = x3 , y = 0 and x = 2 about the x2 + y2
0 x/a

x-axis is _______. Ans: 123π/7 π 1
Ans: a − tan−1
4 a
√
Chapter 8 Multiple Integrals 1 1+x2
2 Iny 1 6. (1 + x2 + y2 )−1 dxdy = _______.
1. 1 0 ex dxdy =_______. Ans: 0 0
2 2 2 2 π √
y 1 Ans: log(1 + 2)
Hint: (y − 1)dy = −y = 4
1 2 1 2
2 7. Change the order of integration and evaluate
2. D (x + y )dxdy =_____, where D y = x,
2 a √x/a
44 (x2 + y2 )dydx = _______.
y2 = x. Ans: 0 x/a
√ 105 a3 a
1 x
Ans: +
Hint: (x + y )dydx
2 2
28 20
0 0
1 √ Hint: Limits for x : ay2 to ay; y : 0 to 1
1 3 x
= x y+ y2
dx
3 8. Change
0
1
0
√ the order of integration and evaluate
1 a a2 −x2
= x5/2 + x3/2 dx (a2 − x2 − y2 )1/2 dydx = _______.
0 3 0 0
2 1 2 44 Ans: πa3 /6
= + · =
7 3 5 105
9. The volume of tetrahedron bounded by
3. D
x3 ydx dy =_____ , where D: x2 + y2 = 1 the coordinated planes and the plane
1 1
x ≥ 0, y ≥ 0. Ans: x + y + z = 1 in _______. Ans:
√ 24 6
1 1−x2
Hint: x3 (ydy)dx√ 10. The volume of tetrahedron bounded by the
1 1−x2 x y z
0
y 2 coordinate planes and the plane + + =1
= x3 dx a b c
0 2 0 abc
is _______. Ans:
1 1 3 6
= (x − x5 )dx
2 0
π x
1 1 1 1 11. sin ydydx = _______. Ans: 2 + π2/2
= − = 0 0
2 4 6 24
4 2
dy dx 12. The area bounded by the curves y = x and
4. = _______.
3 1 (x + y)2 y = x2 is _______. Ans: 1/6
Ans: log(25/24)
4 4
1
2
1 1 13. The area bounded by the curves y = x2 and
Hint: dx = − + dx y = x + 2 is _______. Ans: 9/2
3 (x+y) y=1 3 x+2 x+1
4
x+1
= log Hint: Limits for y: x2 to x + 2 ; x : –1 to 2
x+2 3
Question Bank A-27
Chapter 9 Sequences and Series x2 x3 x4

√ √ 10. The series x − √ + √ − √ + · · · is
= (−1)n−1 ( n + 1 − n)−1 , then the
1. If un $ 2 3 4
series un is _______. absolutely convergent if _______.
Ans: −1 < x < 1
Ans: conditionally convergent
2+1 2+1 3+1 4+1
2. If un = 11. The series 2 − 2 + 2 − 2 +· · ·
$ (−1) [(n + 1) − n ], then the
n−1 3/2 3/2
1 2 3 4
series un is _______. is _______. Ans: conditionally convergent
Ans: absolutely convergent sin n
12. The series whose nth term is un = (−1)n−1
3. If un = (−1) un for n ≥ 1 such that u1 >
n−1
3n
u2 > · · · > 0 then the condition lim un = 0 is is _______. Ans: absolutely convergent
_______ of the series. % 1
13. The series is _______ and the series
Ans: necessary and sufficient for convergence n log n
% 1
n! 2n $ is _______. Ans: dgt, cgt
4. If un = n and vn = 2 then un is _____ and n2
$ n n
vn is _______. Ans: divergent, convergent %∞
14. The series (−1)P where p is a prime
n=1
Hint:
un+1 1 1 % _______. Ans: diverges to +α
(1) = → < 1 ⇒ un is cgt.
un (1+ 1n )n e log n
vn+1 1 % 15. The series whose nth term un = (−1)n−1
(2) =2 →2>1⇒ vn is dgt. n
vn (1+ n )
1 2 _______. Ans: converges conditionally
% 1
262x2 33 x 3 44 x4 16. The series (−1)n−1 is _______.
5. The series x + + + + ··· n2
2! 3! 4! Ans: absolutely convergent
(x > 0) divergent +α ing _______.
1 %
∞
1
Ans: x ≥ 17. The series n is convergent
e
n=2
(log n)k 2 −5k+1
−n2 −n3/2 if k_______. Ans: ∈
/ [2, 3]
1 1
6. If un = 1 + and vn = 1 + √ ,
$ n $ n %
∞
1
then un and vn are _______. 18. The series n is conver-
n=2
(log n)5k 2 −2k+1
Ans: cgt, cgt
gent if k_______. Ans: ∈
/ [1, 3]
n+1 −n
n+1 n+1 %
∞
7. If un = − and vn = 1
n n 19. The series n converges if k
2 n=2
(log n)(k−a)2
n − 1 −n $ $ _______. Ans: ∈/ [a − 1, a + 1]
then un is _______ and vn
n
is _______. Ans: cgt, dgt %∞
1
20. The series sec α
is _______.
n
1 1 $ n=2
8. If un = and vn = 2 lies un is Ans: dgt. for all α
n + 1$ n +1
_______ and vn is _______. Ans: dgt, cgt 1 1 1
21. The series 1 − + − + · · · is _______.
% 2! 3! 4!
9. The series (1)n−1 n−p (p > 0) is absolutely Ans: absolutely convergent
convergent if _______ and conditionally con- 22. One of the following series is convergent. It is
vergent if _______. Ans: p > 1, p ≤ 1 _______. Ans: x + x2 + x3 + · · · |x| < 1
$
23. un is a series of positive terms such that Hint: ∇ p = (2xy, x2 , 0) (1,−1,0)
un $
lim = l if l > 1 ⇒ un is cgt. = (2(1)(−1), 12 , 0) = (−2, 1, 0)
n→∞ un+1
$
l < 1 ⇒ un dgt and l = 1 the test fails. ∇p (−2, 1, 0)
Unit normal = =
This test is known as _______-test. |∇p | (−2)2 + 12 + 0
Ans: D’Alembert’s Ratio Test √
= (−2, 1, 0)/ 5
Chapters 10 Vector Differential 10. The unit normal to the surface

√ x3 + y2 +2z 2
Calculus =26 at (2,2,3) is _______/ 11.
1. If f (x, y, z) = x + y + z, g(x, y, z) = x2 + Ans: ī + j̄ + 3k̄
y2 + z 2 and h(x, y, z) = yz + zx + xy then
[∇f , ∇g, ∇h] = _______. Ans: 0 11. The unit normal to the surface f = x2 y + 2z −
3 = 0 at p (1, 1, 1) is _______.
2. ∇(r n ) = _______. Ans: nr n−2 r̄ 2ī + j̄ + 2k̄
Ans:
3
1
3. ∇ = _______. Ans: r −3 r̄ 12. The angel between normals to the surface =
r xy − z 2 at P = (1, 1, 1) and Q = (2, 2, −1)
4. ∇(r) = _______. Ans: r −1 r̄ is _______. Ans: π/2
5. If = f (u) where u = u(x, y, z), then ∇ = Hint: n̄1 = ∇ P

= (y, x, −2z)P = (1, 1, −2)
_______. Ans: f (u)∇u n̄2 = ∇ = (y, x, −2z)Q = (2, 2, +2)
Q
6. The directional derivative of (x, y, z) = 2xy + = n̄1 · n̄2 = 0 ⇒ θ = π/2

z 2 at the point p(1, −1, 3) in the direction
of the vector ī + 2j̄ + 2k̄ is _______. 13. If ā is a constant vector, then ∇(ā · r̄) =
Ans: 14/3 _______. Ans: ā
7. The directional derivative of f (x, y, z) = xy + 14. The maximum value of the directional deriva-
yz + zx in the direction of ī + 2j̄ + 2k̄ at the tive of = x2 yz at (1,4,1) is _______.
point (1,2,0) is _______. Ans: 10/3 Ans: 9
8. The directional derivative of (x,y, z) = xyz 2 +

¯ = (x + 3y)ī + (y − 22)j̄ + (x + pz)k̄ is
15. If
1 1 solenoidal, then p =_______. Ans: −2
c at (0,1,1) in the direction of 0, √ , √
2 √ 2
is _______. Ans: 3/ 2 16. If (2x + 3y + az)ī + (bx + 2y + 3z)j̄ + (2x + by +
3z)k̄ is irrotational, then (a, b, c) =_______.
Hint: ∇f · c̄ = (yz 2 , xz 2 , 2xy2 )0,1,1
Ans: (2,3,3)
1 1
· 0, √ , √ 17. If F̄ = x (y + z) ī + y (z + x) j̄ + z (x + y) k̄,
2 2 then lim ∇ · F̄ = _______.

1 1 Ans: 2(x + y + z)
= (0, 1, 2) · 0, √ , √
2 2
√ 18. If ā is a constant vector, then curl
= 3/ 2 (r̄ × ā) = _______. Ans: −2ā
9. The unit normal to the surface

√ (x, y, z) = x2 y 19. If r̄ = xī + yj̄ + 2k̄ and (r n+2 r̄) is solenoidal,
at (1, −1, 0) is _______/ 5. Ans: (−2, 1, 0) then n =_______. Ans: −5
Question Bank A-29
1
20. Curl [(r̄ × ā) × b̄] = _______. Ans: b̄ × ā 1
1 4
Hint: (t + 2t)2t dt = t 4 + t 3
2
0 2 3 0
21. If f = 4x + 9y + z , then ∇ f = _______.
2 2 2 2
11
Ans: 28 =
6
22. If f = xy/z, then the value of ∇ 2 f
at (2,3,1) is _______. Ans: 24 5. The work done by F̄ = −xyī + y2 j̄ in moving a
particle over the circular path x2 + y2 = 1 from
23. If r̄ = xī + yj̄ + z k̄, then ∇ · (r̄/r 2 ) = _______. 2
(1,0) to (0,1) is _______. Ans:
1 3
Ans: 2
r Hint: F̄d r̄ = −xydx + y2 dy
24. If r 2 = x2 + y2 + z 2 , then ∇ 2 f (r) = _______. C C
2 π/2
Ans: f + f = − cos t sin t(− sin tdt)
r 0
Chapters 11 Vector Integral + sin2 t · cos tdt

Calculus 2
=
1. If C is the curve x = cos t, y = sin t (0 ≤ t ≤ 5
1
π/2), then C x2 yds = _______. Ans: 6. The work done by F̄ = −xyī + y2 j̄ in moving
3
a particle from (0, 0) to (1, 1) along C : x =
Hint: ds = x2 + y2 dt ⇒ x2 yds 11
C
t, y = t 2 is _______. Ans:
π/2 5
1
= cos2 t sin tdt Hint: F̄d r̄ = (3t 2 , 3t 3 ) · (1, 2t)dt
0
1 3 π/2
C 0
1 1
=− cos t = = (3t 2 + 6t 4 )dt =
11
3 0 3 5
0
2. If C is the curve x = t 2 , y = 2t, then t = 0 7. If C is the boundary of the region in the first
ds
to 1 = _______. Ans: 1 quadrant
bounded by y = x and x2 = y, then
C 4 + y2 −1
(x2 + y2 )dx + ydy = _______. Ans:
Hint: ds = x2 + y2 dt = 2 1 + t 2 dt, C 6
√ Hint: By Green’s theorem
1
2 1 + t2
√ =1
0 2 1 + t2 Mdx + Ndy = (Nx + My )dxdy
3. If C is the curve x = t 2 , y = 2t (0 ≤ t ≤ 1) and C

v̄ = xyī + y2 j̄ then C v̄d r̄ = _______. (x2 + y)dx + ydy = (0 − 1)dxdy
C R
52 1 x
Ans:
1
5 = (−1)dy dx
0 x2
Hint: v̄d r̄ = (2t 3 , 4t 4 ) · (2t, 2)dt 1
C 01 =−
52 6
= (4t 4 + 8t 2 )dt =
0 15 8. (3xyī − y2 j̄) · (īdx + j̄dy) = _______, if C is
C
4. If C is the curve x = t 2 , y = 2t (0 ≤ t ≤ 1) the curve y = 2x2 from (0, 0) to (1, 2).
11 −7
then C (x2 + y)dx = _______. Ans: Ans:
6 6
9. The circulation of F̄ = (2x −y +2z)ī +(x +y − Ans: −6et + 7e2t

/ ,
z)j̄ +(3x+2y−5z)k̄ along the circle x2 +y2 = 4 1
in the xy plane is _______. Ans: 8π 8. L−1 = _______.
s3 (s − 1)
1
10. If f¯ = x2 ī + xyj̄ and C is the curve y2 = x from Ans: − t 2 − t − 1 + et
2/ ,
(0, 0) to (1, 1), then C F̄ · d r̄ = _______. dn ¯
7 9. If L{f¯ (s)} = f (t), then L−1 f (s) =
Ans: dsn
12 _______. Ans: (−1)n t n f (t)
11. If C is the circle x2 + y2 = 1, z = 0 then / ,
2 −4 1 1
C
(y dx − x2 dy) = _______. Ans: 10. L−1 √ = _______. Ans: √
3 s πt
¯
11. If f (s) = L{f (t)}, then L{f (t − a)U (t − a)} =
Hint: By Green’s theorem
_______. Ans: e−as f¯ (s)
(y2 dx − x2 dy) = (−2x − 2y)dxdy
12. If H (t −a) is the unit step function of Heaviside,
C
⎡ √ R
⎤ then L{H (t − a)} = _______. Ans: e−as /s
1 1−x2
= ⎣ (−2x − 2y)dy⎦
0 y=0 13. The function defined by
1 k, 0≤t<a
= −2x 1 − x2 − (1 − x2 ) dx f (t) =
0 −k, a ≤ t < 2a
1
2 1
= (1 − x2 )3/2 − x + x3 and f (t + 2a) = f (t) is a _______ function.
3 3 Ans: square wave
0
1 2 14. The function defined by
= 0−1+ −
3 3
t/a, 0≤t<a
f (t) =
Chapter 12 Laplace Transforms (2a − t)/a, a ≤ t ≤ 2a
1
1. L{sin t cos t} = _______. Ans: and f (t + 2a) = f (t) is a _______ function.
s2 +4 Ans: triangular
2. The first shifting theorem states that if 3
L{f (t)} = f¯ (s) then L{eat f (t)} = _______. 15. t 3 δ(t − 4)dt = _______. Ans: 0
Ans: f¯ (s − a) 0 ∞
√ √ √ 1
3. L{e2t t} = _______. Ans: π/ s − 2 16. cos tδ(t − π/4)dt = _______. Ans: √
0 2
√ b 17. The convolution of | ∗ | = _______.
4. L{sin2 3t} = _______. Ans: Ans: t
s(s2 + 12)
/ t ,
18. The convolution of | ∗ |∗ = _______.
5. If L{f (t)} = f¯ (s), then L f (u)du = 1
0 Ans: t 2
f¯ (s) 2
_______. Ans: 19. The convolution of sin wt ∗ cos wt = _______.
s 1
/ , Ans: t sin wt
1 − e−t s+1 2
6. L = _______. Ans: log
/
t
,
s 20. The convolution of t ∗ eat = _______.
s+5 1
7. L 2 = _______. Ans: 2 (eat − 1 − at)
s − 3s + 2 a
Match the Following
Chapter 1 Ordinary Differential Equations of I order and I degree
1) A (Differential equations) B (Curves)
dy
1. y = x a. y = A cos x + B sin x
dx
d 2y
2. =0 b. x + y = c(1 − xy)
dx2
d 2y
3. +y =0 c. Straight lines through the origin
dx2
dy 1 + y2
4. + =0 d. Circles with centre at 0
dx 1 + x2
dy x
5. + =0 e. Straight lines in xy-plane
dx y
f. Circles with centre on
Ans: 1. c 2. e 3. a 4. b 5. d
2) A (Families of curves) B (Orthogonal trajectories)
1. y2 = 4ax a. r n = cn sin nθ
2. x2/3 + y2/3 = a2/3 b. r n sin nθ = cr
3. x2 − y2 = c c. r n cos nθ = an
4. r n = an cos nθ d. 2x2 + y2 = C
5. r n cos nθ = an e. xy = k
f. x4/3 + y4/3 = a4/3
Ans: 1. d 2. f 3. e 4. a 5. b
3) A (Differential equations) B (Integrating factor)
1. Mdx + Ndy = 0 is homogeneous a. (Mx − Ny )−1

and Mx + Ny = 0

2. f1 (x, y)ydx + f2 (xy)xdy = 0 b. exp p(x)dx
and Mx − Ny = 0
3. (My − Nx )/N = p(x) c. (Mx + Ny )−1
4. (Nx − My )/M = q(y) d. x−3 y−3

5. xy(ydx + xdy) + x2 y2 (2ydx − xdy) = 0 e. exp q(y)dy
f. (Nx + My )−1
Ans: 1. c 2. d 3. b 4. e 5. d
4) A (Differential equations) B (Integrating factor)
1. ydx − xdy + y2 xex dx = 0 a. (xy)−1

2. ydx − xdy + 2x2 y sin x2 dx = 0 b. (x2 + y2 )−1
3. xdx + ydy + (x2 + y2 ) tan xdx = 0 c. x−2
2
4. ydx − xdy + 2x2 yey dy = 0 d. y−2
5. xdy + ydx + e−xy dx = 0 e. x−2 y−2
f. exy
Ans: 1. d 2. a 3. b 4. c 5. f
Chapter 2 Linear Differential Equations of Second Order and Higher

Degree
5) A (Differential equations) B (Curves)
1. (D2 − 1)y = −1 a. y = Ae2x + xex

2. (D2 − 3D)y = ex b. y = Ae2x + (x + 1)ex
3. D2 y = sin x c. y = A + Be3x − 22 ex
4. Dy = x d. y = Aex + Be−x − 1
5. (2 − D)y = xex e. y = A + 12 x2
f. y = (A + Bx) − sin x
Ans: 1. d 2. c 3. f 4. e 5. b
Question Bank A-33
Chapter 3 Linear Differential Equations of Second Order and Higher

Degree
6) A (Series expansion) B (Function)
x2 x3 x4
1. x − + − + ··· a. cos2 x
2 3 4
x2 x4
2. 1 + x + − + ··· b. sin−1 x
2 8
x4 2x6
3. 1 − x2 + − + ··· c. tan x
3 45
1 x3 1 · 3 x5
4. x + + + ··· d. esin x
2 3 2·4 5
1 2
5. x + x3 + x5 + · · · e. log(1 + x)
3 15
f. sin x
Ans: 1. e 2. d 3. a 4. b 5. c
Chapter 4 Functions of Several Variables and Partial Differentiation
7) A B
1. If u = 2x − 3y, v = 5x + 4y then a. p is a saddle point
∂(u, v)
=
∂(x, y)
2. If u = x + y, v = y then b. 4v + 1
∂(u, v)
=
∂(x, y)
x+y xy
3. If u = v= then u2 = c. 1
x−y (x − y)2
4. f (a, b) is a stationary value of f (x, y) d. 12
5. If for f (x, y), rt − s2 < 0 at P(a, b) e. 23
f. fx = 0 = fy at P
Ans: 1. e 2. c 3. b 4. f 5. a
Chapter 5 Radius of Curvature
8) A (Curve point) B (Radius of curvature)

√ √ √
1. x3 + y3 = 3 2xy (3/ 2, 3/ 2) a. 2
√ √ √
2. x2 + y2 = 1 (1/ 2, 1/ 2) b. 2 2/3
3. (y2 = x3 + 8) (−2, 0) c. 1
4. r = a(1 − cos θ) (a, π/2) d. 6
5. r 2 = 9 cos 2θ (r, θ) e. 3/8
f. 3/r
Ans: 1. 2. c 3. d 4. b 5. f
9) A (Family of curves) B (Envelope)
x2 y2
1. y = mx + a
m
a. 2
+ 2 =1
a b
√
2. y = mx + a 2 m2 + b 2 b. x + y2 = a2
2
√ √ √
3. y = mx + am2 c. x + y = c
4. x cos θ + y sin θ = a d. x2/3 + y2/3 = e2/3

x y
5. + = 1, a+b=c e. x2 + 4ay = 0
a b
f. y2 = 4ax
Ans: 1. f 2. a 3. e 4. b 5. c
Chapter 6 Curve-Tracing
10) A (Curve) B (Asymptote)
1. x3 + y3 = 3axy a. y = 0
2. y3 − x2 y + 2y2 + 4y + x = 0 b. x = 2a
3. (y2 (2a − x) = x3 c. x = a
4. y = (a − x) = x(a + x) d. y + x + 1 = 0
5. y2 (x2 − a2 ) = x e. x + y + a = 0
f. x = 0
Ans: 1. e 2. d 3. b 4. c 5. a
Question Bank A-35
Chapter 7 Applications of Integration
11) A (Curve) B (Length)

1. Arc of the cycloid x = a(θ − sin θ) a. 3a/2
y = a(1 − cos θ)
2. Arc of x2/3 + y2/3 = a2/3 in the b. 2aπ
first quadrant
3. Arc of parabola y2 = 4x from c. 8a
y = 0 to y = 3/2
4. Semi-perimeter of the cardioid d. 4aπ
r = a(1 − cos θ)
15
5. Perimeter of r = 2a cos θ e. + log 2
6
f. 4a
Ans: 1. c 2. a 3. e 4. f 5. b
Chapter 8 Multiple Integrals
12) A (Integral) B (Value)

√
1 1−x2
2 2
1. e−x −y dydx a. π/2
0 0
1 1
1
2. dydx b. 14/3
1 + x 2 + y2
0 a 0 a
x dx dy
3. c. 1
x2 + y2
0 ∞ y ∞ −y
e
4. dxdy d. π/4 log 2
y
0 a
π
5. x + ydA where R is the region e. a
R 4
bounded by x = 0, y = 1, x + y = 1
f. π/4(1 − e−1 )
Ans: 1. f 2. d 3. e 4. c 5. b
Chapter 9 Sequences and Series
13) A (Series) B (Nature)

1 1 1
1. 1 − 2 + 2 − 2 + · · · a. Diverges to +∞
2 3 4
2. tan 1 + tan 12 + tan 13 b. Converges

$√ √
3. ( n3 + 1 − n3 − 1) c. Oscillates infinitely
1 1 1
4. 1 − + − + · · · d. Converges absolutely
2 3 4
$∞
5. n=1 (−1)n−1 e. Diverges to −∞
f. Oscillates finitely
Ans: 1. d 2. a 3. b 4. c 5. f
Chapter 10 Vector Differential Calculus
14) A B
1. F̄ is irrotational a. 3
2. ∇ · r̄ b. div F̄ = 0
3. F̄ is irrotational c. 0̄ oscillates infinitely
4. ∇ × r̄ d. curl F̄ = 0̄ converges absolutely
5. ∇r e. 0
f. r̄/r
Ans: 1. d 2. a 3. b 4. c 5. f
15) A (Differential operators) B (Result)
1. curl grad r n is irrotational a. Irrotational

2. If f¯ , ḡ are irrotational then f¯ × ḡ b. Harmonic
3. ∇ 2 (xyī + yz j̄ + zxk̄) c. 6
4. ∇ · [∇(x2 + y2 + z 2 − 3xyz)] d. 0
5. φ∇φ e. 0̄
f. Solenoidal
Ans: 1. e 2. f 3. d 4. e 5. a
Chapter 11 Vector Integral Calculus
16) A B

1. r̄d r̄ a. 4πa3
Question Bank A-37

2. S (x +z)dydz +(y +z)dzdx +(x +y) b. 64/3π
dxdy where S : x2 + y2 + z 2 − 4 = 0

3. S (xī + yj̄ + z k̄) · n̄ds c. 4/3πa3
where S : x2 + y2 + z 2 − 1 = 0

4. S (xdydz + ydzdx + zdxdy) d. 0
where S is the sphere x2 + y2 + z 2 = a2

5. S (xī + 2yj̄ + 3z k̄) · nds e. 4π
where S is the surface enclosing volume
f. 6V
Ans: 1. d 2. b 3. e 4. a 5. f
Chapter 12 Laplace Transforms
17) A (Laplace transform) B (Property)

1¯
1. L{f (at)} = f (s) a > 0 a. Laplace transform of periodic function
a
2. L{e−at f (t)} = f¯ (s + a) b. Initial value theorem
/ t ,
1
3. L f (t)dt = f¯ (s) c. Transform of integral
0 s
4. L{f (t − a)U (t − a)} = e−as f¯ (s) d. I Shifting theorem
p
1
5. L{f (t)} = e−st f (t)dts > 0 e. II Shifting theorem
1−e−sp 0
f. Change of scale
Ans: 1. f 2. d 3. c 4. e 5. a
18) A (Inverse Laplace transform) B (Functions)

- 1 .
1. L−1 a. (e−2t − e−3t )/t
- s(s + 11)
2
.
−1
2. L b. (et − e−2t )
- (s +s3)(s − 1)
+ 3. 1
−1
3. L log c. sin at
- ses−as + 2. 2a
−1
4. L d. 1 − cos t
- (s −s b ) .
2 2
−1 1 t
5. L e. (e − e−3t )
(s2 + a2 )2 4
cosh b(t − a), t>a
f.
0, t<a
Ans: 1. f 2. d 3. c 4. e 5. a
True or False Statements
1. The rule that always gives the generalsolution of 13. F(x, y, −1/y ) = 0 is the differential equation
an exact equation Mdx +Ndy = 0 is y const Mdx of the orthogonal trajectories of the system with

+ (terms independent of x in N )dy = c. F differential equation F(x, y, y ) = 0. T

2. The rule that always gives the general solu- dθ
14. F r, θ,−r = 0 is the differential equa-
tion of an exact equation
Mdx + Ndy = 0 dr
∂v tion of the orthogonal trajectories of the
is v + N− dy = c where v = system of curves with differential equa-
∂y
Mdx (see Ex. 7.3 pp. 237–238 Engg. dr
y const tion F r, θ, = 0. F
Maths I (I Edition)). T dθ
15. The differential equation governing the natural
3. The system of parabolas y2 = 4a(x + a) is self- dx
orthogonal. T growth substance is = kx (k > 0). T
dt
4. The equation (D2 +a2 )y = tan ax may be solved 16. The rule to find the particular integral
by the method of variation of parameters. T yp = [1/f (D)]eax is to replace always D in
f (D) by ‘a.’ F
5. The method of variation of parameters can be
cos ax x
applied to find an approximate nth root of a 17. yp = = sin ax. T
positive number. F D 2 + a2 2a
1
18. yp = 2 ex = xex . F
6. Lagrange’s mean value theorem may be applied D − 2D + 2
to find an approximate nth root of a pos- 1 1
itive number. 19. yp = 4 ex cos x = ex cos x. F
T D −1 5
7. Maclaurin’s theorem is a special case of 20. Particular integral of y +

Py + Qy = R is yp =
vR
Taylor’s theorem. T Au + Bv where A = − dx. T
uv − u v
dy
8. An integrating factor of − y = x2 is e−x . 21. The ‘c’ of Rolle’s theorem for f (x) = (x − a)m
dx mb + na
T (x − b)n in [a, b] is c = . T
1 m+n
9. is an integrating factor of xdy−ydx = 0. x2 + ab
x2 +y2 22. If f (x) = log in [a, b] then ‘c’ of
T x(a + b)
Rolle’s theorem is the G.M of a and b. T
dy 2
10. The general solution of x + y = ex is
x2 dx 23. Rolle’s theorem may be applied to
xy + e dx = c. F 1
f (x) = in [−1, 1]. F
dy |x|
11. = ex−y (ex − ey ) is a first order linear
dx 24. The value of Cauchy’s mean value theorem for
differential equation in ey . T √ 1
f (x) = x and g(x) = √ in [1, 4] is 2. T
dy x
12. The general solution of the equation +
dx 25. If u = y/z + z/x, then xux + yuy + zu + z = 1.
Py = Q is y(I.F.) = c + Qdx. F
F
Question Bank A-39
x2 + y2 43. For the curve y2 (a + x) = x2 (b − x), the

26. z = sin−1 ⇒ xzx + yzy = sin z. F
x+y origin is a cusp. F
27. Curvature is the reciprocal of the radius of cur- 44. The curve y2 (a − x) = a + x is symmetri-
vature for a curve y = f (x). T cal about the x-axis. T
28. The evolute of a curve is the envelope of 45. The curve r 2 = a2 cos 2θ is symmetrical
its normal. T about the line θ = π/2. T
29. y = (x2 + 1)/(x2 − 1) has no asymptotes. F 46. The length of the arc of the curve y =
log(ex − 1)/(ex + 1) from x = 1 to x =
30. The curve 9ay2 = (x − 2a)(x − 5a)2 has two 2 is log(e + e−1 ). T
asymptotes. F
47. The surface area of the solid of revolution of the
31. Astroid is a special case of hypocycloid. T circle x2 + y2 = 1 is 1 sq.r. F
xy
32. f (x, y) = , (x, y) = (0, 0) and 48. The volume of the solid of revolution
x2+ y2 x2 y2
f (0, 0) = 0 is continuous at the origin. F of the ellipse 2 + 2 = 1 about
a b
its major axis is 3 πab .
4 2
T
33. u = u(x, y), v = v(x, y) are functionally inde-
u, v 49. The perimeter of the curve r = a sin θ is
pendent if J = 0. F
x, y πa. T
34. u = ey − x, v = ex − y are functionally
independent. T 50. The surface area of the sphere generated by
revolving the circle about its diameter is 32π.
35. u = r cos θ, v = r sin θ are functionally depen- F
dent. F 51. The area bounded by y = cos x between x =
−π/2, x = π/2 and the x-axis is 2 sq.r. T
36. If u(x, y) = ex/y , then xux + yuy = 0. T
52. By changing the order of integration in I =
63
37. The curve r = a sin 3θ (α > 0) has 6 dxdy where x = 2u and y = 3v, we
loops. F
0 0 31
obtain I = 6 0 0 dudv. T
38. The least value √
of the radius of curvature of D + 1 4x 6 4x
53. e = e . F
y = log x is 3 3/2. T D2 − 5 11
1
54. 2 3e2x = xe2x . T
39. The locus of the centre of curvature of a curve D −D−2
is called an envelope of the curve. F D 1
55. 2 sin 4x = − cos 2x. F
D −4 8
40. y2 + 4ax = 0 is the envelope of the family of
a 56. The radius of curvature of the parabola y2 =
curves y = mx + . F 4ax at (at 2 , 2at) is 2a(t 2 + 1)3/2 . T
m
41. The radius of curvature of the curve r = a cos θ 57. The radius of curvature of y = x
√ e at the point
at (r, θ) is a. F where it crosses the y-axis is 2 2. T
42. The curve y = x3 is symmetrical about 58. The radius of curvature of the curve r = aθ
the origin. F at (r, θ) is (r 2 + a2 )3/2 /(r 2 + 2a2 ). T
59. The centre of curvature of the rectangular γ(n)

65. L t n eat = . T
hyperbola xy = 1 at (1, 1) is (–2, 2). F (s − a)n
/ ,
1 − et s−1
60. The curve y = f (x) is rising in [a, b] if 66. L = log . T
t s
dy / ,
> 0 in [a, b]. T 1
dx 67. L−1 3/2 = 2 t/π. F
s
61. The curve y = f (x) is concave upwards if
d 2y 68. The circulation of F̄ = yī + z j̄ + xk̄ round the
> 0. T circle c : x2 + y2 = 1, z = 0 is −π. T
dx2
1
62. L eat = . F 69. The work done by F̄ = 3x2 ī + (2xz − y)j̄ +
s+a z k̄ in moving a particle from (0, 0, 0) to
1 (2, 1, 3). F
63. L eat = . T (a,0)
s+a
a 70. The value of xdy + ydx along the upper
64. L {sin at} = 2 . F (−a,0)
s + a2 half circle x2 + y2 = a2 is 0. T
Question Papers
Code No. RR10102 Set No.1
I year B. Tech. Regular Examinations, May/June, 2006
Engineering Mathematics-I
Time: 3 hours Max Marks: 80
Answer any FIVE Questions
All questions carry equal marks
1. (a) Test the convergence of the following series 1 1

When x = 1, un = √ ; vn = 3/2 ;
% ∞
x2n (n + 1) n n
√ .
(n + 1) n un 1
n=1 ∴ = √ · n3/2
vn (n + 1) n
Solution Let ux denote the xth term of the given n
=
series. Then n+1
1 1
un = √ x2n ; = → 1 = 0 as n → ∞
(n + 1) n 1 + 1n
1 $ $
∴ un+1 = √ x2n+2 ∴ By comparison test both un and vn
(n + 2) n + 1 $ $ 1
√ converge or diverge together. But vn =
un+1 x2n+2 (n + 1) n n3/2
= √ 3
un (n + 2) n + 1 x2n is p-series with p = > 1 and hence convergent.
√ √ 2
n n+1 2 % x2n
= x Therefore, the given series √ con-
n+2 (n + 1) n

1 + 1/n 2 verges if x ≤ 1 and diverges if x > 1.
= 1+ x → x2 as n → ∞
1 + 2/n
$ 1. (b) Find the interval of convergence of the series
By D’Alembert’s ratio test, un converges for x2 x3 x4
+ + + · · · ∞.
x2 < 1 (i.e., x < 1) and diverges for x2 > 1 (i.e., 2 3 4
x > 1) and the test fails for x = 1. We use comparison
$ $ 1 Solution Let un denote the nth term of the given
test now with vn = as auxiliary series. series.
n3/2
xn+1 xn+2 Clearly f (x) is continuous in [0,1] and f (x) is

un = ; un+1 =
(n + 1) (n + 2) derivable in (0,1). ∴ f (x) satisfies the condition of
un+1 xn+2 n + 1 Taylor’s theorem. Hence,
= ·
un n + 2 xn+1 x2 x3
n+1 f (x) = f (0) + xf (0) + f (0) + f (θx) (1)
= |x| 2! 3!
n+2
1 + 1/n where 0 < θ < 1.
= |x| → |x| as n → ∞ Given interval [0, x] = [0, 1] ⇒ x = 1. We have
1 + 2/n
% xn+1 f (0) = (1 − 0)5/2 = 1,
By D’Alembert’s ratio test, the series 5 5
n+1 f (0) = − (1 − 0)3/2 = −
is convergent for |x| < 1 and divergent for |x| > 1 2 2
and the test fails for |x| = 1. In this case, we use 5 3 15
$ $1 f (0) = (−1)2 · · (1 − 0)1/2 = ;
2 2 4
comparison test with vn = . When |x| = 1
n substituting these values in Eq. (1)
un 1 1
= ·n= → 1 = 0 as n → ∞. 12 15 13 15
vn n+1 1+ 1 f (1) = 1 + 1 · (−5/2) + · + −
n 2! 4 3! 8
$ $
∴ By comparison test, both un and vn con- 5 15 15
=1− + − .
$ $ 1 2 8 6
verge or diverge together. But vn = is
n 2. (a) If x + y + z = u, y + z = uv, z = uvw, then
p-series with p = 1 and is divergent if x = 1.
1 1 1 ∂(x, y, z)
When x = −1, the series becomes − + − evaluate .
2 3 4 ∂(u, v, w)
1
+ · · · , which is an alternating series with decreas- Solution From the first two equations x = u−uv,
5
ing terms (un → 0 as x → ∞) and is convergent by on subtraction from the last two equations y = uv −
Leibnitz’s test. Therefore uvw, on subtraction. Now
⎧ ∂x ∂
⎪
⎪ cgs if |x| < 1 and x = −1, = (u − uv) = 1 − v
⎪
⎨ ∂u ∂u
x 2
x 3
x 4
i.e., cgs if − 1 ≤ x < 1 ∂x ∂x
+ + + ··· +
2 3 4 ⎪
⎪ dgs if |x| > 1 and x = 1, = −u and =0
⎪
⎩ ∂v ∂w
i.e., dgs if x < −1, x ≤ 1 ∂y ∂y ∂y
= v − vw, = u − uw and = −uv
∂u ∂v ∂w
Hence the interval of convergence is [−1, 1). ∂z ∂z ∂z
= vw, = uw and = uv
∂u ∂v ∂w
1. (c) Write Taylor’s series for f (x) = (1 − x)5/2
with Lagrange’s form of remainder up to 3 ∂x ∂x ∂x
terms in the interval [0,1]. ∂u ∂v ∂w
∂(x, y, z) ∂y ∂y ∂y
Solution =
∂(u, v, w) ∂u ∂v ∂w
f (x) = (1 − x)5/2 , ∂z ∂z ∂z
5 ∂u ∂v ∂w
f (x) = (−1) (1 − x)3/2 , 1−v 1−v 0
2
5 3 1 = v − vw u − uw −uv
f (x) = (−1)2 · · (1 − x)−1/2 vw uw uv
2 2 2
Question Papers A-43
1−v −u 0 y
= v u 0
vw uw uv
B
= uv[(1 − v)u + uv] = u2 v.
S
2. (b) If ρ1 and ρ2 are radii of curvatures of any o A x
chord of the cardioids r = a(1 − cos θ)
which passes through the pole, then show C
16a2
that ρ12 + ρ22 = .
9
Solution Let A = (OA, α) and B = (OB, π + α) Figure 3a
y
1 1
We have y1 = , y2 = − 3 , on differentiating.
t 2at
A coordinates of the centre of curvature
α
y1 (1 + y12 )
x̄ = x − = 2a + 3at 2 or 2a + 3x (2)
o x y2
1 + y12 −y3
B
ȳ = y + = −2at 3
or (3)
y2 4a2
Figure 2b Eliminating t between Eqs. (2) and (3), the evolute
is the semicubical parabola
be the extremities of a chord AB through the pole O
where AOX = α. At any point P = (r, θ) on the 27ay2 = 4(x − 2a)3 (4)
cardioid
4
whose cube √ is as A = (2a, 0).√We observe that
ρ = a sin θ/2 (1) B = (8a, 4 2a) and C = (8a, −4 2a) are points of
3 intersection of the parabola (1) and the evolute (4).
The radii of curvatures at A and B are We know that the length of the arc of the evolute
between two points ρ1 and ρ2 equals the difference
4a (ρ1 − ρ2 ) of the radii of curvature of the given curve
ρ1 = sin α/2; (2)
3 (evolute) at the two points corresponding to ρ1 and
4a 4a 2
ρ2 = sin(π + α/2) = a cos α/2 (3) ρ2 . Also ρ = √ (x + a)3/2
3 3 a
squaring and adding
∴ Length of evolute = 2(ρx=3a − ρx=0 )
2 √
16a = 2(6 3a − 2a)
ρ12 + ρ22 = ∵ sin2 α/2 + cos2 α/2 = 1. √
9
= 4(3 3a − a)
3. (a) In the evolute of the parabola y2 = 4ax show √
= 4a(3 3 − 1)
that the length of the curve from its cusp
x = 2a to the point where √ it meets the Here length of√ the curve from the cusp
parabola y2 = 4ax is 2a(3 3 − 1). A(x = 2a) = 2a(3 3 − 1).
Solution Equation of parabola (Involute): 3. (b) Find the length of the arc of the curve
ex − 1
y = 4ax or x = at , y = 2at
2 2
(1) y = log x from x = 1 to x = 2.
e +1
Solution 4. (c) A body kept in air with temperature 25◦ C

y = log(ex − 1) − log(ex + 1) cools from 140◦ C to 80◦ C in 20 min. Find
dy ex ex when the body cools down to 35◦ C.
⇒ = x − x
dx e −1 e +1
2ex 2 Solution By Newton’s law of cooling, we have
= 2x = x = cosechx
e −1 e − e−x dθ
2+ = −k(θ − θ ◦ ) (1)
dt
Length of the arc = 1 + y12 dx
1
2 where θ ◦ is the temperature of the surrounding
medium, i.e., air.
= 1 + cosech2 x dx
1
2 dθ
2 ∴ = −kdt (2)
= coth xdx = log sinh x θ − 25◦
1 1
Integrating, we obtain log(θ − 25◦ ) = −kt + c
e2 − e−2 At t = 0, θ = 140◦ C
= log = log(e + e−1 ) = log(e + 1/e).
e − e−1
⇒ log(140◦ − 25◦ ) = −k(0) + c
4. (a) Form the differential equation by eliminating
the arbitrary constants from y = ex (a cos x ⇒ c = log 115◦ (3)
+b sin x).
Now from Eq. (2)
Solution
kt = log 115◦ − log(θ − 25) (4)
y = ex (a cos x + b sin x) or
−x
e y = a cos x + b sin x (1) At t = 20, θ = 80◦ C
Differentiating Eq. (1) w.r.t x twice (by Leibnitz ⇒ 20k = log 115 − log(80 − 25)
theorem)
⇒ 20k = log 115 − log 55 (5)
e−x y2 − 2e−x y1 + e−x y = −a cos x − b sin x
= −e−x y (2) t kt log 115 − log(θ − 25)
= =
y2 − 2y1 + 2y = 0 or 20 20k log 115 − log 55
d 2y dy log 115 − log 10
− 2 + 2y = 0 cancelling e−x (3) = when θ = 35◦ C
dx2 dx log 115 − log 55
4. (b) Solve the differential equation log 11.5
= = 3.31 or t = 66.2 min
y(2xy + ex )dx − ex dy = 0. log(23/11)
Solution The equation y(ex y + ex )dx − ex dy = 0 The body cools to 35◦ C after t = 66.2 min.
can be put in the form
5. (a) Solve the differential equation
2xy2 dx + yex dx − ex dy = 0 or (D2 + 2D − 3)y = x2 e−3x .
uex dx − ex dy
2xdx + = 0 or Solution The given equation (D2 + 2D − 3)y =
y2
x 2 −3x
x e is a non-homogeneous linear differential equa-
e
d(x2 ) + d = 0 which is exact tion. Its complete solution consists of two parts.
y
(a) the complementary function (C.F) obtained by
ex solving the equation (D3 − 4D2 − D + 4)y = 0
Integrating x2 + = c.
y (b) the particular integral (P.I)
C.F
Auxiliary equation is yP.I = A cos 2x + B sin 2x
m2 + 2m − 3 = 0 ⇒ (m − 1)(m + 3) = 0
Here u = cos 2x, v = sin 2x, R = sec 2x
yC.F = c1 ex + c2 e−3x
P.I cos 2x sin 2x
W (u, v)(x) =
1 1 −2 sin 2x 2 cos 2x
yP.I = e−3x x2 = e−3x x2
(D+3)(D−1) D(D−3−1) = 2 = 0

1 D −1 2 vR sin 2x · sec 2x
= e−3x 1− x A=− dx = − dx
−4D 4 W

2
1 sin 2x
−3x 1 D D2 D3 =− dx
=e 1 + + 2 + 3 + · · · x2 2 cos 2x
−4D 4 4 4
−3x
1
e D 1 D D 2
= log | cos 2x|
= + + + · · · x2 4
−4 4 4 16 64 uR cos 2x · sec 2x 1
B= dx = dx = x
e−3x x3 2x 2 W 2 2
= + x2 + + ···
−4 3 16 64
−3x
3 The complete solution is
e x x 1
= + x2 + +
−4 4 8 32
∴ y = yC.F + yP.I
Now the complete solution is 1 1
= c1 cos 2x + c2 sin 2x + log | cos 2x| + x.
y = yC.F + yP.I or 4 2
3
e−3x x 2 x
x −3x
y = c1 e + c2 e − − x + 9s2 − 12s + 5
4 3 8 6. (a) If L{f (t)} = , then find L{f (3t)}
(s − 1)3
the last term merges with c2 . using change of scale property.
5. (b) Solve the differential equation (D2 + 4)y = 9s2 − 12s + 5

Solution Let L{f (t)} = f¯ (s) =
sec 2x by the method of variation of (s − 1)3
parameters. By change of scale property, we have
Solution The given differential equation is 1

L{f (t)} = f¯ (s) ⇒ L{f (t)} = f¯ (s/k)
(D + 4)y = sec 2x which is a non-homogeneous
2
k
linear differential equation. Its complete solution 1¯ 1 9( 3s )2 − 12( 3s ) + 5
consists of two parts. ∴ L{f (3t)} = f (s/3) =
3 3 ( 3s − 1)3
(a) the complementary function (C.F) and
(b) the particular integral (P.I) 9(s2 − 4s + 5)
= .
C.F (s − 3)3
Auxiliary equation is / ,
−1 s
6. (b) Find L .
m2 + 4 = 0 ⇒ m = ±2i (s2 + 4)2
yC.F = c1 cos 2x + c2 sin 2x / ,
1
−1
P. I Solution We know that L =
(s2 + a2 )
P.I is found by the method of variation of parame- 1 1 1
ters. We write, sin at where f¯ (s) = 2 and f (t) = sin at.
a s + a2 a
By the property of derivatives 7. (a) Find the work done by F̄ = (2x − y − z)ī+
- . (x + y − z)j̄ + (3x − 2y − 5z)k̄ along a curve
−1
L−1 f¯ (s) = f (t) ⇒ L−1 f¯ (s) C in the xy-plane given by x2 + y2 = 9, z = 0.
= −tf (t) Solution
- /. ,
−1 d 1
∴ L−1 f¯ (s) = L−1 F̄ = (2x−y−z)ī+(x+y−z)j̄+(3x−2y−5z)k̄
ds s2 + a2
/ ,
−2s r̄ = xī + yj̄ + k̄ ⇒ d r̄ = dxī + dyj̄ + dz ī
= L−1
(s2 + a2 )2
In the xy-plane, z = 0 ⇒ dz = 0 so that
1
= −t sin at
a F̄d r̄ = (2x − y)dx + (x + y)dy
/ ,
s t
⇒ L−1 = sin at
(s2 + a2 )2 2a Work done by force F̄ is
Taking a = 2, we have
/ , W = F̄ · d r̄ where C : x2 + y2 = 9
s t C
L−1 = sin 2t.
(s + 4)
2 2 4 Writing x = 3 cos t, y = 3 sin t
dx = −3 sin t dt, dy = 3 cos t dt (0 ≤ t ≤ 2π)
6. (c) Evaluate (x2 + y2 )dxdy in the positive 2π
quadrant for which x + y ≤ 1. W = F̄d r̄ = (6 cos t − 3 sin t) [−3 sin t dt
C 0
Solution +(3 cos t + 3 sin t)3 cos t] dt

x=1 y=1−x
2π
(x2 +y2 )dxdy = dx (x2 + y2 )dy = (−18 cos t sin t + 9 sin2 t
R x=0 y=0 0
x=1 y=1−x + 9 cos2 t + 9 cos t sin t)dt
1
= x2 y + y3 dx 2π
3

x=0

y=0 = (9 − 9 sin t cos t)dt
x=1
1
0
2π
= x2 − x3 + (1 − x)3 dx 9
x=0 3 = 9t − sin2 t = 18π
3 4
x=1 2
x x 1 0
= − − (1 − x) 4
3 4 12 x=0 7. (b) If the scalar fields, u and v and a vector field
1 1 1 1 F̄ are such that uF̄ = v then prove that
= − −0+ =
3 4 12 6 curl F̄ = 0̄.
y Solution The given relation does not give the

B(0, 1) desired answer Curl F̄ = 0̄
8. (a) Find the area of the Folium of Descartes

y = 1– x
x=0 x3 + y3 = 3axy(a > 0) using Green’s
Theorem.

1 1
x
x Solution A= xdy = x · xdt
O y=0 A(0, 1) 2 2
Put y = t x, t varies from 0 to ∞
Figure 6c
3at 2 8. (b) Use

x(1 + t 3 ) = 3at 2 ⇒ Divergence theorem to evaluate
4t 3 (xi + yj + z 2 k) · n dS where S in the

9a2 t 2 3a2 ∞ 3t 2 S
⇒x = 2
= dt surface bounded by the curve x2 + y2 = z 2 in
(1 + t 3 )2 2 0 (1 + t 3 )2
the plane z = 4.
3a2 ∞ du 3a2 1 ∞
= =
2 0 u2 2 u 1
2
Solution The question is incorrect. Surface is
3a 3a2 not clearly mentioned. Hence solution is not possible.
= (1 − 0) = .
2 2
1. (a) Test the convergence of the following series By Leibnitz test, the series is convergent. Further
% ∞
1 · 3 · 5 · · · (2n + 1) % ∞ %∞
1
. |un | = .
n=1
2 · 5 · 8 · · · (3n + 2) n=2 n=2
n(log n3 )
$ By the Cauchy Maclaurin’s integral test, since
Solution should range from n = 0 to ∞ and ∞ ∞
dx 1 1
not from n = 1 to ∞ =− =
x log x 3 2(log x) 2 2(log 2)2
1 · 3 · 5 · · · (2n + 1) 2 2
Let un = (a finite quantity) the series is absolutely convergent.
2 · 5 · 8 · · · (3n + 2)
1 · 3 · 5 · · · (2n + 1)(2n + 3) 1. (c) Show that
Then un+1 =
2 · 5 · 8 · · · (3n + 2)(3n + 5) sin−1 x 12 12 ·32 5 12 ·32 ·52 7
√ = x+ x3+ x+ x +· · · .
un+1 2n + 3 2 1−x2 3! 5! 7!
∴ = → < 1 as n → ∞
un 3n + 5 3
Solution The question is not correct. It should
By D’Alembert’s ratio Test, the series is conver- be show that
gent. 12 3 12 · 32 5 12 · 32 · 52 7
sin−1 x = x + x + x + x +· · ·
% (−1)n 3! 5! 7!
1. (b) Prove that the series converges
n(log n)3 Let f (x) = sin−1 x
absolutely.
= a0 +a1 x+a2 x2 + · · · + an xn + · · · (1)
$
Solution should range from n = 2 to ∞ (For
n = 1, log n = 0 and ∴ u is not defined.)
1 Differentiating Eq. (1) w.r.t x, we get
Let un = for n ≥ 2.
(log n)3
%∞ %∞
(−1)n 1
The given series (−1)n un = is √ = a1 +2a2 x+3a3 x2 +· · ·+nan xn−1 +· · · (2)
n=2 n=2
n(log n)3 1−x2
an alternating series with u2 > u3 > u4 > · · · , i.e., 1
each successive term is numerically decreasing. But √ = (1 − x2 )−1/2
1 − x2
1
Also lim un = lim =0 = 1+a1 +2a2 x+3a3 x2 +· · ·+nan xn−1 + · · · (3)
n→∞ n→∞ (log n)3
Comparing Eqs. (2) and (3), we have = 3a2 (1 + cos θ) = 6a2 cos2 θ/2 (5)
a1 = 1, 2a2 = 0, 3a3 = 1/2, The radius of curvature
1·3 (r 2 + r12 )3/2 (4a2 cos2 θ/2)3/2
4a4 = 0, 5a5 = ,··· (4) ρ= =
2·4 r 2 + 2r1 − rr1
2
6a2 cos2 θ/2
Also a0 = sin−1 0 = 0 from Eq. (1) if we take the 8a cos θ/2
3 3
4a
smallest positive value of the inverse function. Hence = 2 = cos θ/2
6a cos2 θ/2 3
2
1 x3 1·3 x5 1·3·5 x7 ρ2 16a
cos2 θ/2 8a
sin−1 x = x+ + + +· · · (5) = 9 = ( constant )
2 3 2·4 5 2·4·6 7 r 2a cos2 θ/2 9
which can be written as
3. (a) Trace the curve y2 = (x − 2)(x − 3)2 .
12 12 ·32 5 12 ·32 ·52 7
sin−1 x = x+ x3 + x + x +· · · (6)
3! 5! 7! Solution
Symmetry:
2. (a) If u = x2 − y2 , v = 2xy where x = r cos θ,
∂(u, v) The curve is symmetric about the x-axis since the
y = r sin θ, show that = 4r 3 . equation is not attended when y is replaced by −y.
∂(r, θ)
The curve is not symmetric about the y-axis nor
Solution We have about any other line or origin. The curve does not
pass through (0,0).
u = x − y2 = r 2 (cos2 θ − sin2 θ) = r 2 cos 2θ
2
v = 2xy2 = 2r 2 cos θ sin θ) = r 2 sin 2θ y
∂(u, v) ur uθ 2r cos 2θ −2r sin 2θ

2 Tangent
= =
∂(r, θ) vr vθ 2r sin 2θ 2r 2 cos 2θ
= 4r 3 (cos 2θ + sin2 2θ) = 4r 3 · 1 = 4r 3 o A x
(2, 0) B
2. (b) For the cardioid r = a(1 + cos θ) prove that (3, 0)
ρ2 /r is constant where ρ is the radius of Tangent
curvature. Tangent
Solution Figure 3a
r = a(1 + cos θ) (1) Intersection of the curve with the axes:
Differentiating w.r.t θ twice The curve meets the x-axis at x = 2 and x = 3. For
x < 2, y is imaginary. So, the curve does not meet
r1 = −a sin θ, (2) the y-axis nor does it extend to the left of x = z.
r2 = −a cos θ (3) Tangents at A and B:
Shifting the origin to B = (3, 0), the equation
Now, r + 2
r12 = a (1 + cos θ) + (−a sin θ)
2 2 2
becomes y2 = (x + 1)x2 . Equating to zero the lowest
= a2 [1 + 2 cos θ + cos2 θ + sin2 θ] degree terms, we get y = ±x as the tangents at
= a2 (2 + 2 cos θ) B(3, 0) which are inclined at ±45◦ to the x-axis.
There are no asymptotes to the curve. The shape
= 4a2 cos2 θ/2 (4) of the curve is as shown in the figure.
r +
2
2r12 − rr2 = a (1 + cos θ) + 2a sin θ
2 2 2 2
+ a2 (1 + cos θ) cos θ 3. (b) Prove that the volume of revolution of

r 2 = a2 cos 2θ about the initial line is
= a2 [1+2 cos θ+cos2 θ+2 sin2 θ πa2 √ √
√ [3 log( 2 + 1 − 2)].
+ cos θ + cos2 θ] 6 2
Solution The curve forms two loops, symmet- 4. (a) Form the differential equation by eliminating
√
rical about the origin (Fig. 4.31) (Lemmiscale of the arbitrary constant from y = 1+c 1 − x2 .
Bernoulli). For the upper half of right-hand loop, θ
varies from 0 to π/4. Solution The given relation is
∴ Volume of solid of resolution=2 (Vol. generated by
the upper half of right-hand loop, in the first quadrant) y = 1 + c 1 − x2 (1)
π/4 Differentiating Eq. (1) w.r.t x, we get
2π 3
=2 r sin θdθ
0 3 dy x x
= −c √ = −c 1 − x2 (2)
4πa3 π/4 dx 1 − x2 1 − x2
= (cos 2θ)3/2 sin θdθ ∵ r 2 = a2 cos 2θ
3 0
Eliminating c between Eqs. (1) and (2)
4πa3 π/4
= (2 cos2 θ − 1)3/2 sin θdθ dy x
3 0 = (1 − y) (3)
√ √ dx 1 − x2
Put 2 cos θ = t ⇒ − 2 sin θdθ = dt
√ which is the required differential equation.
θ = 0 ⇒ t = 2 θ = π/4 ⇒ t = 1
4. (b) Solve the differential equation
4πa3 1
= − √ √ (t 2 − 1)3/2 dt (1 − ex/y )dx + (1 − x/y)ex/y dy = 0.
3 2 2
√ √2
2 2πa3 Solution Please refer to Ex. 1.21 in Ch. 1 on p.
= (t 2 − 1)3/2 dt
3 1
8 (Revised Edition).
We evaluate
4. (c) In a certain chemical reaction, the rate of
I = (t 2 − 1)3 /2dt by integration by parts conversion of a substance at time t is
proportional to the quantity of the substance
√ √
2
2 still untransformed at that instant. At the end
= (t −1) t
2 3/2
−3/2 (t 2 −1)1/2 ·2t ·tdt of one hour 60 grams remain and at the end
1 1
√ √
2 of four hours 21 grams. How many grams of
= 2−3 (t 2 − 1)1/2 [t 2 − 1 + 1]dt the first substance was there initially?
1
√ √
2 Solution By the law of natural decay
= 2 − 3I − 3 (t 2 − 1)1/2 dt
1 dy
√2 = ay ⇒ y = ceat
√ t 2 1 dt
4I = 2−3 t − 1 − log(t + t 2 − 1) Let y = y0 at t = 0 (initially). Then y = y0 eat
2 2 1
√
√ 3 2 3 √ When t = 1 hr y = 60 grams ⇒ y0 ea = 60
= 2− + log( 2 + 1)
√ 2 2 When t = 4 hr y = 21 grams ⇒ y0 e4a = 21
2 3 √ 4 1/3
=− + log( 2 + 1) 60
2 2 y0 = = 85.13 gm
1 √
√ 21
= 3 log( 2 + 1) − 2 5. (a) Solve the differential equation
2√
2 2πa3 1 √ √ (D2 + 1)y = x2 e3x .
= · 3 log( 2 + 1) − 2
3 8
√ 3
Solution The given differential equation
2πa √ √
= 3 log( 2 + 1) − 2 . (D2 + 1)y = x2 e3x (1)
12
is a non-homogeneous linear differential equation. d d

where D = and θ = . The given equation
Its complete solution consists of two parts: dx dz
(a) complementary function (C.F) (x D −4xD+6)y = (log x) is transformed to [θ(θ−
2 2 2
(b) particular integral (P.I). 1) − 4θ + 6] = z 2 or (θ − 2)(θ − 3)y = z 2 which is

C.F a non-homogeneous linear differential equation with
Auxiliary equation m2 + 1 = 0 ⇒ m = ±i the independent variable z. The complete solution is
y = yC.F + yP.I
yC.F = c1 cos x + c2 sin x (2)
C.F : yC.F = c1 e2z + c2 e3z = c1 x2 + c2 x3
P.I
since m = 2, 3 are the roots of the A.E. P.I
1 1
yP.I = e3x · x2 = e3x x2 1
D +1
2 (D + 3)2 + 1 yP.I = z2
1 (θ − 1) (θ − 3)
= e3x 2 x2
D + 6D + 10 1 θ −1 θ −1 2
= 1− ) 1− z
e3x 3 1 9 6 2 3
= 1 − D − D2 + D2 + · · · x 2
10 5 10 25 1 1 1 1 1
= 1+ θ+ θ 2 +· · · 1− θ+ θ 2 +· · · z 2
6 2 4 3 9
(1 + u)−1 = (1 − u + u2 − · · · )
1 1 1 1 1 1 1
3 1 = 1+ + θ+ + + + θ 2 +· · · z 2
where u = D + D2 6 2 3 2 3 4 9
5 10
1 2 5 19
e3x 2 6 13 = z + · 2z + ·2
= x − x+ ·2 6 6 36
10

5 50
1 2
e 3x
6 13 = 36 log x + 60 log x + 38
= x2 − x + 216
10 5 25 The complete solution is
The complete solution is y = yC.F + yP.I
y = yC.F + yP.I = c1 x2 +c2 x3 +[36(log x)2 +60(log x)+38]/216
3x
/ , ∞
e 6 13 1
= c1 cos x + c2 sin x + x2 − x + 6. (a) Prove that L f (t) = f¯ (s)ds
10 5 25 t
/ , s
e3x 1
= c1 cos x + c2 sin x + (250x2 − 30x + 13). where L f (t) = f¯ (s).
250 t
5. (b) Solve the differential equation: Solution Please refer to Theorem 10 in Ch. 12
(x2 D2 − 4xD + 6)y = (log x)2 . on p. 17 (Revised Edition).
Solution This is an Euler–Cauchy equation 3(s2 − 2)2
(Equidimensional Equation). We change the inde- 6. (b) Find the inverse transform of .
2s5
pendent variable to z by putting x = ex or z = log x.
so that Solution
/ , / , / ,
dy dy dz 1 dy 3 3 −1 1 1
= · = or xDy = θy L−1 = L − 6L−1 3
dx dz dx x dz 2s5 2 s s
/ ,
d 2y 1 d 2 y 1 dy 1
2
= · + 6L−1 5 by linearity
dx x dz 2 x2 dz s
2
1 d y dy 3 t2 t4
= 2 − or x2 D2 y = θ(θ − 1)y = ·1−6· +6·
x dz 2 dz 2 2! 4!
3 1 3 1 π 3πa4
= − 3t 2 + t 4 . = a4 · · · = .
22 4 4 2 2 16
6. (c) Evaluate (x + y2 )dx dy over the area
7. (a) Find the angle between the surfaces x2 + y2 +
x2 y2
bounded by the ellipse 2 + 2 = 1. z 2 = 9 and z = x2+y2−3 at the point (2, −1, 2).
a b
Solution Let φ = x2 + y2 + z 2 − 9 = 0 and
x2 y2
Solution Solving 2 + 2 = 1 for y, we get ψ = x + y − z − 3 = 0 be the two given surfaces.
2 2
a b The angle between two intersecting surfaces is equal
b 2
y = ± a − x2 . So the region of integration is to the angle between their normals at an intersecting
a
b 2 b 2 point.
R : −a ≤ x ≤ a, − a − x2 ≤ y ≤ a − x2 Let x¯1 = φ and x¯2 = ψ be the normals at the
a a
since the integrand is even in x and y common point (2, –1, 2). Then

I= (x2 + y2 ) dx dy n¯1 = [ (x2 + y2 + z 2 − 9)](2,−1,2)
R
⎡ √ ⎤ = (2x, 2y, 2z)(2,−1,2) = (4, −2, 4)
a ab a2 −x2
⎣ n¯2 = [ (x2 + y2 − z − 3)](2,−1,2)
=4 (x2 + y2 )dy⎦ dx
x=0 y=0 = (2x, 2y, −1)(2,−1,2) = (4, −2, −1)
√ n¯1 · n¯2
a y= ba a2 −x2 cos θ =
1 |n¯1 ||n¯1 |
=4 x y + y3
2
dx
x=0 3 y=0 (4, −2, 4) · (4, −2, −1)
a =
b 2 4 + (−2)2 + 42 42 + (−2)2 + (−1)2
2
=4 x2 · a − x2
x=0 a 16 8
=√ √ = √
1 b3 36 21 3 21
+ (a2 − x 2 )3/2 dx

3 a3 8
⇒ θ = cos−1 √ .
b 4 b3 3 21
= 4 I1 + · 3 I2
a 3 a
a 7. (b) Evaluate (r̄/r 3 ) where r̄ = xī + j̄ + z k̄ and
where I1 = x2 a2 − x2 dx, |r̄| = r.
a
0
Solution
I2 = x2 (a2 − x2 )3/2 dx r̄ = xī + j̄ + z k̄ and r 2 = x2 + y2 + z 2
0
∂r ∂r x
b πa4 4 b3 3πa4 ⇒ 2r = 2x or =
=4 + · 3 ∂x ∂x r
a 16 3 a 16 ∂r y ∂r z
πa3 b πb3 a πab 2 Similarly = , = .
= + = (a + b2 ) ∂y r ∂z r
4 4 4
r̄
x = a sin t, dx = a cos tdt Also 3 = (r −3 x)ī + (r −3 y)j̄ + (r −3 z)k̄
r
when x = 0, t = 0; when x = a, t = π/2 % ∂
π/2 (r̄/r 3 ) = (r −3 x)
∂y
I1 = a4 sin2 t cos2 tdt %
∂r
r −3 · 1 + x · (−3)r −4
0
=
11 π πa4 ∂x
=a · 4
· = %
42 2 16 = −3
(r − 3x r ) 2 −5
π/2
%
I2 = a4 cos4 t dt = 3r −3 − 3r −5 x2
0

= 3r −3 − 3r 2 r −5 = 0. = (x3 + x4 + 2x3 )dx
C1

8. Verify Green’s theorem for (xy + y2 )dx + x2 dy 3 4 1 5
C = x + x
where C is bounded by y = x and y = x2 . 4 5
3 1 19
Solution By Green’s theorem, we have = + = (3)
4 5 20
Along C2 : y = x from A = (1, 1) to O = (0, 0). The
y
line integral is

y = x2
Mdx + Ndy = (x · x + x2 )dx + x2 dx
C2 C2
0 0
= 3x2 dx = x3 = −1 (4)
y =x 1 1
c2
c1 From Eqs. (2) and (4), we get
o x

19 1
Mdx + Ndy = −1=− (5)
20 20
Figure 8 C
∂N ∂M
Again − dx dy
∂x ∂y
S

∂N ∂M
Mdx + Ndy = − dxdy (1) = (2x − x − y) dx dy
C S ∂x ∂y S
1 x
Here M = xy + y2 and N = x2 = (x − 2y) dx dy
y=x2
The st. line y = x and the parabola intersect at the
x=0
origin (0,0) and at A(1, 1). C consists of two parts C1

1 x
= xy − y2 x2
dx
(along parabola) and C2 (along st. line) 0
1
∴ Mdx + Ndy = Mdx + Ndy = [(x2 − x2 ) − (x3 − x4 )]dx

0
C C1
1 1
x5 x4
+ Mdx + Ndy (2) = (x4 − x3 )dx = −
C2 0 5 4 0
1 1 1
Along C1 : y = x2 the line integral is = − =− (6)
5 4 20
The equality of Eqs. (5) and (6) verifies the Green’s
Mdx + Ndy = [x(x2 ) + x4 ]dx + [x2 ]d(x2 ) theorem (1).
C1 C1
1. (a) Test the convergence of the following series 4

% (2n + 1) = =2>1
2
xn (n > 0).
n3 + 1
By Raabe’s test, the series is convergent.
Solution Let Hence the series is convergent if x ≤ 1 and diver-
2n + 1 n 2n + 3 gent if x > 1.
ux = x , ∴ ux+1 = nn+1
n +1
3 (n + 1)3 + 1 1. (b) Test the following series for absolute / condi-
%∞
log n
ux+1 2n + 3 n3 + 1 tional convergence (−1)n 2 .
= x n
ux (n + 1)3 + 1 2n + 1 n=1
2 + 3/n 1 + 1/n3
= · x log n
2 + 1/n (1 + 1/n)3 + 1/n3 Solution Let un = , u1 = 0. The series
n2
→ x as n → ∞ starts with u2 . Clearly u2 > u3 > u4 > · · · . The
terms are decreasing and lim un = 0.
By D’Alembert’s ratio test, the series converges n→∞
for x < 1 and diverges for x > 1. The test fails if By Leibnitz’s test for alternating series, the given
% log n
x = 1. For x = 1, we have to apply Raabe’s test. series (−1)n 2 is convergent.
Consider n
Consider the series
ux (2n + 1)[(n + 1)3 + 1]
−1= −1 %
∞ %
∞
ux+1 (2n + 3)(n3 + 1) log n
|(−1)n un | = (−1)n
4n3 + 9n2 + 5n − 1 n2
= n=2 n=2
(2n + 3)(n3 + 1) %
∞
log n
=
ux n2
Now lim n −1 n=2
n→∞ ux+1
(4n3 + 9n2 + 5n − 1) of absolute values. By Cauchy–Maclaurin ∞integral
= lim n · $
n→∞ (2n + 3)(n3 + 1) test, ∞n=k u(n) and the improper integral k u(x) =
(4 + 9 · 1/n + 5 · 1/n2 − 1/n3 ) dx both converge or diverge.
= lim log x
n→∞ (2 + 3/n)(1 + 1/n3 ) Here u(x) = 2 and k = 2. We evaluate
x
∞
log x ∂u
I= dx = 0 ⇒ x = 0 or x + 2y = π (3)
2 x2 ∂y
1 From Eqs. (2) and (3), we get
Put log x = t, dx = dt
x
x = y = π/3 (4)
x = e ; x = 2 ⇒ t = log 2
t
∞ ∞
= te−t dt = −(t + 1)e−t ∂2 u
Now l = = 2 sin y cos(2x + y)
log 2 log 2 ∂x2
√ √
= (1 + log 2)/2 = a finite quantity = 2 · 3/2 · (−1) = − 3
∴ The series is absolutely convergent. ∂2 u
m= = cos y sin(2x+y)+sin ycos(2x+y)
∂x∂y
1. (c) Verify Cauchy’s mean value theorem for √
= sin 2(x + y) = − 3/2
f (x) = ex and g(x) = e−x in [a, b].
∂2 u
n= = 2 sin x cos(x − y)
Solution f (x) = ex and g(x) = e−x on ∂y2
[a, b] ⊂ R+ are continuous on [a, b] and derivable √ √
= 2 · 3/2 · (−1) = − 3
on (a, b) and g (x) = −e−x = 0 for all x ∈ (a, b)
So, the condition of Cauchy’s mean value theorem 3 9
is satisfied. ∴ We have ln − m2 = 3 − = >0
4 4
Hence there exists c ∈ (a, b) such that ⇒ u is max at (x, y) = (π/3, π/3) and
f (b) − f (a) f (c) eb − ea √ √ √ √
= ⇒ −b 3 3 3 3 3
g(b) − g(a) g (c) e − e−a umax = · · =
e c
a+b 2 2 2 8
= ⇒ =c at the point of stationary value (π/3, π/3).
−e −c 2
a+b
The existence of c = ∈ (a, b) satisfying 2. (b) Find the radius of curvature of the curve
2 x2 y = a(x2 + y2 )at(−2a, 2a).
the above condition verifies the theorem.
Solution Differentiating x2 y = a(x2 + y2 ) w.r.t
2. (a) Find the stationary points of u(x, y) = sin x x, we get
sin y sin(x + y) where 0 < x < π, 0 < y < π
and find the maximum of u. x2 y1 + 2xy = a(2x + 2yy1 )
2ax − 2xy
Solution ⇒ y1 = 2
x − 2ay
u = sin x sin y sin(x + y) 0 < x, y < π (1) −4a2 + 8a2
∂u y1 at (−2a, 2a) = (not defined)
= cos x sin y sin(x + y) + sin x sin y cos(x + y) 4a2 − 4a2
∂x Hence we use the implicit eqn. formula:
= sin y[cos x sin(x + y) + sin x cos(x + y)] (fx2 + fy2 )3/2
= sin y sin(2x + y) ρ=
|fxx fy2 − 2fxy fx fy + fyy fx2 |
∂u
= 0 ⇒ y = 0 or 2x + y = π (2) f (x, y) = x2 y − a(x2 + y2 ) = 0
∂x
∂u ∴ fx = 2xy − 2ax, fy = x2 − 2ay,
= sin x cos y sin(x + y) + sin x sin y cos(x + y)
∂y fxx = 2y − 2a; fyy = −2a, fxy = 2x
= sin x [cos y sin(x + y) + sin y cos(x + y)] At (−2a, 2a), fx = 4a2 , fy = 0,
= sin x sin(x + 2y) fxx = 2a, fxy = 4a, fyy = −2a
((4a2 )2 + (0)3/2 ) 4. (b) Solve the differential equation

∴ρ=
|2a · 0 − 2 · 4a · 4a2 + (−2a)(2a)2 | dy
+ (y − 1) cos x = e− sin x cos2 x.
= 2a. dx
Solution The equation can be written as
3. (a) Trace the curve r = a + b cos θ(a > b).
d(y − 1)
Solution Please refer to Ex. 6.18 on p. 16 of Ch. + (y − 1) cos x = e− sin x cos2 x
6. dx
which is linear in (y − 1);
3. (b) Obtain the surface area of the solid of P = cos x, Q = e− sin x cos2 x.
revolution of the curve r = a(1 + cos θ) about
I.F is e Pdx = e cos xdx
= esin x
the initial line.
The solution is
Solution The required surface area
π y(I.F) = c + Q(I.F)dx ⇒ (y − 1)esin x
ds
= 2π yds = 2π r sin θ · dθ
π 0 dθ = c + cos2 dx
= 2π a(1 + cos θ) sin θ
1
0 cos2 dx = (1 + cos 2x)dx
+ 2
a2 (1 + cos θ)2 + a2 sin2 θ dθ 1 1
* = x + x sin 2x
ds dr 2 4
∵ = r 2 + ( )2 1
dθ dθ ⇒ y = (x + sin x cos x) + esin x + c.
π 2
= 16πa2 cos4 θ/2 sin θ/2dθ
0 4. (c) The temperature of a cup of coffee in 92◦ C
π
2 − cos θ/2
5
32 2 when freshly poured, the room temperature
= 32πa = πa . being 24◦ C. In one minute, it was cooled to
5 0 5
80◦ C. How long a period must elapse, before
4. (a) Form the differential equation by eliminating the temperature of a cup becomes 65◦ C.
√ arbitrary constant c from y = 1 + x +
2
the
c 1+x . 2 Solution If T denotes the temperature of coffee
then by Newton’s law of cooling
Solution The given equation can be written as
dT
y = −k(T − T0 ), k > 0 where T0 = 24◦ C (1)
√ = 1+ +c
x2 dt
1 + x2
Differentiating w.r.t x, we get Separating the variable and integrating, we get
√
dT
1 + x2 dy − y√x 2 = −k dt
dx 1+x x T − 24
=√
1+x 2
1 + x2 ⇒ log(T − 24) = −kt + log e
Cross-multiplying we get or T − 24 = ce−kt (2)
dy Initially, when t = 0, T = 92 ⇒ c = 68 from

(1 + x2 ) − xy = x(1 + x2 )
dx Eq. (2)
which is the required differential equation.
∴ T − 24 = 68e−kt (3)
Also when t = 1, T = 80◦ C The complete solution is

68 y = yC.F + yP.I
⇒ ek =
56
e3x
⇒ k = log
68
= log
17 = c1 e−x + c2 ex + c3 e4x − (sin 2x + 7 cos 2x).
56 14 200
we have to find t corresponding to T = 65. From 5. (b) Solve the differential equation
Eq. (3) y + 2y − y − 2y = 1 − 4x3 .
65 − 24 = 68e− log(17/14) Solution The given equation is (D3 + 2D2 −

14 41 D + 2)y = 1 − 4x3 and is a non-homogeneous linear
⇒ 41 = 68 × t or t = differential equation. Its complete solution consists
17 56
5. (a) Solve the differential equation of two parts:
(D3 − 4D2 − D + 4)y = e3x cos 2x. (a) C.F (b) P.I
C.F
Solution The equation (D3 − 4D2 − D + 4)y = A.E is m3 + 2m2 − m − 2 = 0 ⇒ m = 1, −1, −2
3x
e cos 2x is a non-homogeneous linear differential
yC.F = c1 ex + c2 e−x + c3 e−2x
equation. Its complete solution consists of two parts
(a) the complementary function (C.F) obtained by P.I
solving the eqn. (D3 − 4D2 − D + 4)y = 0
1
(b) the particular integral (P.I) yP.I = 1 − 4x3
C.F (D − 1)(D + 1)(D + 2)

A.E is m3 − 4m2 − m + 4 = 0 ⇒ m = −1, 1, 4 1 −1 D −1
= 1 − D2 1+ 1 − 4x3
yC.F = c1 e−x + c2 ex + c3 e4x −2 2
1
P.I = 1 + D2 + D4 + · · ·
−2
1
yP.I = e3x cos 2x D D2 D3
D3 − 4D2 − D + 4 1− + − + · · · 1 − 4x3
2 4 8
1
= e3x cos 2x 1 5 2 5 3 1
(D+3) −4(D+3)2 −(D+3)+4
3 = 1− D+ D − D +· · · 2x − 3
2 4 8 2
1
= e3x 3 cos 2x 15 1
D + 5D − 2D − 8
2 = 2x3 − 3x2 + 15x − −
2 2
Replace D2 by –4 = 2x3 − 3x2 + 15x − 8
cos 2x
= e3x The complete solution is
−4D − 20 + 2D − 8
e3x (D − 14) y = yC.F + yP.I
= cos 2x
−2 (D + 14)(D − 14) = c1 ex + c2 e−x + c3 e−2x + 2x3 − 3x2 + 15x − 8.
e3x (D − 14) cos 2x
= 9s2 − 12s + 5
−2 D2 − 196 6. (a) If L{f (t)} = find L{f (3t)}
(s − 1)3
Replacing D2 by –4 in the denominator using change of scale property.
e3x −2 sin 2x − 14 cos 2x
= Solution
−2 −200
e 3x 9s2 − 12s + 5
= (sin 2x + 7 cos 2x) L{f (t)} = = f¯ (s).
−200 (s − 1)3
%
By change of scale property = ī × [0 + w̄ × ī]
1¯ s ∂w̄ ∂r̄
L{f (3t)} = f ∵ = 0 and = ī
3 3 ∂x ∂x
2 % %
1 9 3s − 12 3s + 5 = (ī · ī)w̄ − (w̄ · ī)ī
= s 3
3 −1 ∵ ā × (b̄ × c̄) = (ā · c̄)b̄ − (ā · b̄)c̄
3
9s2 − 4s + 5 = 3w̄ − w̄ = 2w̄.
=
(s − 3)2
/ , 7. (b) Evaluate F̄ · d r̄ where F̄ = (x − 3y)ī+
s
6. (b) Find L−1 . C
(s2 + 4)2 (y − 2x)j̄ and C is the closed curve in the xy-
/ , plane: x = 2 cos t, y = 3 sin t from
1 t = 0 to t = 2π.
Solution We know that L−1 = sin 2t
(s2 + 4)
and Solution
F̄ = (x − 3y)ī + (y − 2x)j̄,
d 1 −2s
= 2 r̄ = xī + yj̄ ⇒ d r̄ = dxī + dyj̄
ds (s2 + 4) (s + 4)2
/ , / ,
s 1 d 1 F̄d r̄ = (x − 3y)dx + (y − 2x)dy
∴ L−1 = L −1
−
(s2 + 4)2 2 ds (s2 + 4) But C : x = 2 cos t, y = 3 sin t
t
= sin 2t and dx = −2 sin t dt, dy = 3 cos t dt and t varies
4 from 0 to 2π.
2π
6. (c) Evaluate (x2 + y2 ) dx dy in the positive F̄d r̄ = (2 cos t − 3 · 3 sin t)(−2 sin t)dt
t=0
quadrant for which x + y ≤ 1.
+ (3 sin t − 2 · 2 cos t) · 3 cos t dt
2π
Solution 5
= sin 2t +9(1−cos 2t)−6(1+cos 2t)dt
x=1 y=1−x 0 2
(x2 +y2 )dxdy = dx x2 + y2 dy 2π
5
R

x=0

y=0
= (3 + sin 2t − 15 cos 2t)dt
x=1
1 y=1−x 0 2
= x2 y + y3 dx 2π
3 y=0 5 15
x=0

x=1 = 3t − cos 2t − sin 2t
1 2 2 0
= x − x + (1 − x) dx
2 3 3
5 15
x=0 3 = 6π − (1 − 1) − · 0 = 6π.
x 3
x 4
1 x=1 2 2
= − − (1 − x)4
3 4 12 x=0 8. Verify divergence theorem for F̄ = x2 ī +y2 j̄ +z 2 k̄
1 1 1 1 over the surface S of the solid cut off by the
= − −0+ = . plane x + y + z = a in the first octant.
3 4 12 6
7. (a) If w̄ is a constant vector, evaluate curl v̄
Solution By Gauss’s divergence theorem
where v̄ = w̄ × r̄.
F̄ · n ds = divF̄dv
Solution S v
∂ %
curl (w̄ × r̄) = t̄ × (w̄ × r̄) Let φ(x, y, z) = x + y + z − a = 0 be the given
∂x
plane. Then
% ∂w̄ ∂r̄
= ī × × +w̄ × ∂φ ∂φ ∂φ
∂x ∂x = = =1
∂x ∂y ∂z
% ∂φ a
2
and grad φ = ī = ī + j̄ + k̄ is normal to the = ax − x + (a − x) dx
2 33
∂x x=0 3
plane. ∴ Unit normal to the plane S is
a
x3 x4 2 x4
grad φ ī + j̄ + k̄ =a − + ·
n̄ = = √ 3 4 3 4
|grad φ| 3 0
1 1 1
= a4 − +
Let R be the projection of S on the xy-plane. Then 3 4 6
the equation of the plane is x + y = a ⇒ y = a − x. 4−3+2 4 a4
Also when y = 0, x = a = a = (1)
12 4
∂ 2 ∂ ∂
F̄ · ndxdy Now ∇ · F̄ = div F̄ = (x ) + (y2 ) + (z 2 )
∴ F̄ · n ds = ∂x ∂y ∂z
S R n̄ · k̄
= 2(x + y + z)
a 2 2 +z 2
a−x x +y√
3
= 1
dxdy ∴ divF̄dv
x=0 y=0 √
V
a a−x
3
a−x−y
2 2 a a−x
= 2
x + y + (a − x − y) dxdy = (x + y + z)dz dy dx
x=0 y=0 x=0 y=0 z=0
a a−x a−x−y
a a−x
1
= 2x + 2y − 2ax + 2xy
2 2
=2 z(x + y) + z 2 dy dx
x=0 y=0 x=0 y=0 2 0
a a−x
−2ay + a2 dy dx a−x−y
=2 (a−x−y) x+y+ dy dx
a
2 x=0 y=0 2
= 2x2 y + y3 − 2axy + xy2 a a−x
3
x=0
a−x = (a − x − y)(a + x + y)dy dx
x=0 y=0
− ay2 + a2 y dx a a−x
0 = [a2 − (x − y)2 ]dy dx

a
2 x=0 y=0
= 2x2 (a − x) + (a − x)3 − x(a − x)2 a a−x
3
x=0
= [a2 − (x2 − y2 )]dy dx
−2ax(a − x) − a(a − x)2 + a2 (a − x) dx x=0 y=0
a a−x
2 a
1 a4
= x (a − x) + (a − x)(a2 − 2ax + x2 ) = a2 y − x2 y − y3 − xy2 dx = (2)
x=0 3 4
x=0
0
−(a − x)3 − 2/3(a − x) dx
The equality of Eqs. (1) and (2) proves the verifi-
(the middle terms cancel) cation.
1. (a) Test the convergence of the following series When x2 = 1, we have

% 1
. un 1 + 2n1
(log log x)n n =n 2 − 1
un+1 1− 1 2n
Solution Let
1 n 1+ 1
2n
− 1 + 1n − 1
4n2
un = ; = 2
(log log x)n 1 − 2n1
1
un1/n = → 0 < 1 as n → ∞ n 3
− 1
log log n 2n 4n2 3
⇒ → >1
1 2 2
By Cauchy’s nth root test, the series is convergent. 1− 2n
1. (b) Find the interval of convergence of the series By Raabe’s test, the series is convergent when
1x3
1·3x 1·3·5x
5 7 x2 = 1, as n → ∞
x+ + + + ··· . Hence the given series is convergent if x2 ≥ 1 and
2 3 2·4 5 2·4·6 7
divergent if x2 > 1
x ∴ Then the interval of convergence is [−1, 1].
Solution Here u1 = ,
1
1 · 3 · 5 · · · (2n − 3) x2n−1 x x2
un = (n ≥ 2) 1. (c) Show that log(1 + ex ) = log 2 ++
2 · 4 · 6 · · · (2n − 2) 2n − 1 2 8
x4 ex 1 x
1 · 3 · 5 · · · (2n − 3)(2n − 1) x2n+1 − · · · and hence deduce that x = + −
un+1 = 12 e +1 2 4
2 · 4 · 6 · · · (2n − 2)(2n) 2n + 1 x 3
un+1 2x − 1 2 + ··· .
∴ = ·x 48
un 2x
2 − 1n 2 − 1n 2 Solution Maclaurin’s series expansion (Taylor’s
= · ·x series at x = 0) for f (x) is
2 2 + 1n
x2 xn
→ x2 as n → ∞ f (x) = f (0)+xf (0)+ f (0)+· · ·+ f (n) (0)+· · ·
2! n!
(1)
By D’Alembert’s ratio test, the series converges if
x2 < 1 and diverges if x2 > 1. The test fails if x2 = 1. Here f (x) = log 1 + ex f (0) = log 2
ex 1 2. (b) Prove that the evolute of our hyperbola

f (x) = f (0) =
1+e x 2
(1 + ex )ex − ex · ex ex x2 y2

f (x) = = − = 1 is (ax)2/3 −(by)2/3 = (a2 + b2 )2/3
(1 + e )
x 2 (1 + ex )2 a2 b2
1 by considering evolute as the envelope of the
f (0) = 2
2 normal.
(1 + ex )2 · ex − ex · 2(1 + ex )ex
f (x) = Solution Evolute of a curve C is the envelope of
(1 + ex )4
the family of normals to the given curve C.
ex − e2x
= Let P = (a cosh t, b sinh t) be any point on the
(1 + ex )3 given hyperbola
f (0) = 0
x2 y2
− =1 (1)
(1+e ) (e −2e )−(e −e )3·(1+e ) e
x 3 x 2x x 2x x 2 x
a2 b2
f v (x) =
(1+ex )6
(1 + e )(e − 2e2x ) − 3e2x − e3x
x x dx d
= Then = x̄ = (a cosh t) = a sinh t;
(1 + ex )4 dt dt
dy
(1 + 1)(1 − 2) − 3(1 − 1) 1 = ȳ = b cosh t (2)
f v (0) = =− dt
(1 + 1)4 8 dy dy/dt ẏ b cosh t
Substituting these values in Eq. (1), we obtain So, = = =
dx dx/dt ẋ a sinh t
x x2 x4 b
log(1 + ex ) = log 2 + + − ··· = coth t (3)
2 8 12 a
Differentiating Eq. (2) w.r.t x, we get Equation of normal to hyperbola (1) at a point
e x
1 x x 3 P = (a cosh t, b sinh t) is
= + + + ··· .
1+e x 2 4 48 −a
y − b sinh t = (x − a cosh t) (4)
2. (a) If x + y = u, y + z = v, z + x = w
2 2 2 b coth t
∂(x, y, z) by ax
find . or + = a2 + b2 (5)
∂(u, v, w) sinh t cosh t
Differentiating w.r.t t
Solution ∂x ∂x ∂x
cosh t sinh t
∂u ∂v ∂w by · + ax · =0
∂(x, y, z) ∂y ∂y ∂y sinh t cosh2 t
= sinh t cosh t
∂(u, v, w) ∂u ∂v ∂w ⇒ =
∂z ∂z ∂z −(by) 1/3 (ax)1/3
∂u ∂v ∂w 1
= (6)
(ax) − (b − y)2/3
2/3
1
1 0
2x Eliminating ’t’ from Eq. (5) using Eq. (6), we get
1
= 1 0
2y (ax)2/3 − (by)2/3 (ax)2/3 − (by)2/3
1 ax + by
0 1 (ax) 1/3 −(by)1/3
2z
1 =a +b
2 2
=1+
8xyz or (ax)2/3 − (by)2/3 = (a2 + b2 )2/3
which is the required envelope of the normals to the dy dy/dθ cos 2θ

∴ = =
hyperbola, i.e., evolute of the hyperbola. dx dx/dθ sin 2θ
dy
3. (a) Trace the curve x = a(θ − sin θ), = 0 when θ = π((2n + 1)π, in general). At these
dx
y = a(1 − cos θ). dy
vertices tangents are horizontal. → ∞ when θ →
dx
Solution The equations of the curve are 0 or 2π (2nπ in general).
x = a(θ − sin θ), y = a(1 − cos θ) (1) At these points, i.e., θ = 0, and θ = 2π tangents
are vertical. We tabulate the values of the function
where θ is a parameter. dy
x, y, .
Symmetry: dx
Here x is an odd function of θ since 0 A B C D
x(−θ) = a(−θ −sin(−θ)) = −a(θ −sin θ) = −x(θ) θ 0 π/2 π 3π/2
2π
π 3π
and y is an even function of θ since x 0 a −1 aπ a +1 2aπ
2 2
y(−θ) = a(1 − cos(−θ)) = a(1 − cos θ) y 0 a 2a a 0
dy
∞ 1 0 -1 ∞
The curve is symmetrical about the y-axis. dx
Origin:
The curve passes through the origin since
θ = 0 ⇒ x = 0, y = 0. As θ varies from 0 to π, the curve rises from 0 to
Intersection with the axes: aπ and as θ varies from π to 2π the curve falls from
x = 0 ⇒ θ = 0. This gives y = 0. Again y = 0 aπ to 0. At points where θ = π/2 and 3π/2 the slopes
gives θ = 0, ±2π, ±4π, · · · For these values of θ, y of the tangents to the curve are 1 and –1 respectively.
has the least value 0 if θ = 0, 2aπ, 4aπ, · · · . These The curve has no asymptotes. The shape of the
are the points where the curve meets the x-axis. curve is as shown in the figure and it is known as
Thus between each pair of consecutive values of x cycloid.
we have one each of the curve. Cycloid is a curve described by a point on the cir-
Limitations and the extent: cumference of a circle, which rolls on a fixed st. line
Since | cos θ| ≤ 1 for all values of θ, y has the least without sliding.
value 0 if θ = 0, 2π, 4π, · · · and the greatest value
y
2a if θ = ±π, ±3π, · · ·
So, the curve lie, between the horizontal lines
y = 0 (the x-axis) and y = 2a y = 0 is called the y = 2a
base and the points at which y = 2a are called the
vertices.
Periodicity:
The curve repeats itself in x-intervals of length 2aπ o θ=0 3
corresponding to the parametric values repeating 2 2 2
x=0 x = a x = 2a
with periodicity of length 2π; the first arch lying
between x = 0 and x = 2π corresponding to θ = 0 Figure 3a Cycloid
and θ = 2π, respectively. 3. (b) A sphere of radius ‘a’ units is divided into two
Special points: parts by a plane distant a/2 from the centre.
dx θ Show that the ratio of the volumes of the two
We have = a(1 − cos θ) = 2a sin2 ;
dθ 2 parts is 5:11.
dy θ θ
= a sin θ = 2a sin cos Solution When a semi-circle of radius ‘a’ units
dθ 2 2
revolves about its diameter, a sphere of radius ‘a’ Solution The given relation is
units is formed which is divided into two spherical y2 = (x − c)2 (1)
segments major and minor by a plane a/2 units from
Differentiating Eq. (1) w.r.t x, we get
its centre. The equation of the generating circle is
x2 + y2 = a2 . 2yy1 = 2(x − c) or y2 y22 = (x − c)2 (2)
Volume of the minor spherical segment Substituting for (x − c)2 from Eq. (1) and
cancelling y2 we get y12 = 1 which is the required
y differential equation.
B(0, a)
y+x−2 dy
= .
a y−x−4 dx
2
A'(-a, 0) o x
A(a, 0) Solution The given equation is a first order
and first degree differential equation of the form
dy
dx
= f (x, y) where f is a non-homogeneous
B'(0, -a) function in x and y. Separating the variables and
Figure 3b grouping the terms the equation can be written as
(x − 2)dx + ydx + xdy − (y − 4)dy = 0.
a a Multiplying by 2 and nothing that
V1 = πy dx = π
2
(a2 − x2 )dx ydx + xdy = d(x, y) we can integrate and get
a/2 a/2
a the general solution x2 + 2xy − y2 − 4x + 8y + c = 0.
1
= π a2 x − x3
3 a/2 4. (c) If the temperature of air is 20◦ C and tempera-

a 3
1 3 a3 ture of the body drops from 100◦ C to 80◦ C
=π a − 3
− a − in 10 min. what will be its temperature after
2 3 8
3 20 min? when will the temperature be 40◦ C?
5πa
=
24 Solution By Newton’s Law of cooling
Volume of the minor spherical segment dθ
= −k(θ − θ0 ) = −k(θ − 20) (1)
a/2 a/2 dt
V2 = πy dx = π
2
(a2 − x2 )dx Separating the variables and integrating, we get
−a −a
a/2 θ − 20 = ce−kt (2)
1
= π a x − x32
where c is a constant.
3 −a
3
a 1 a3 Initially, when t = 0, θ = 100◦ C ∴ c = 80
=π +a −
3
+a3
2 3 8 Eq. (2) now becomes
27a3 π θ − 20 = 80 e−kt (3)
=
24 ◦ −10k
When t = 10 min θ = 80 C ∴ 60 = 80 e
The ratio of two volumes 3
⇒ e−10k =
4
5πa3 27πa3 When t = 20 min
V1 : V2 = : = 5 : 27.
24 24 3
θ − 20 = 80e−20k = 80(e−10k )2 = e−10k = (4)
4. (a) Form the differential equation by eliminating 4
the arbitrary constant from y2 = (x − c)2 . ⇒ θ = 65 (temperature after 20 min)
When θ = 40 we get from Eq. (3), is an Euler–Cauchy Equidimensional equation.

Change the independent variable x to z by putting
1 t
20 = 80e−kt ⇒ (ek )t = ⇒ (3/4)1/10 (5) x = ez or z = log x so that xDy = θy, x2 D2 y =
4 θ(θ − 1)y where θ = dzd and D = dxd
1 1 3 1 Equation (1) becomes
= ⇒t· log = log or
4 10 4 4
10 log(1/4) [θ(θ − 1) − θ + 1]y = z ⇒ (θ − 1)2 y = z (2)
t= . (6)
log(3/4)
This is a non-homogeneous linear differential
5. (a) Solve the differential equation equation. Its complete solution consists of
(D2 + 4)y = ex + sin 2x. (a) complementary function, yC.F and
(b) particular integral, yP.I
Solution The given differential equation C. F
We have to solve the homogeneous equation
(D2 + 4)y = ex + sin 2x (1)
(θ − 1)2 y = 0
is a non-homogeneous linear differential equation.
Its complete solution consists of (a) complementary A.E. is (m − 1)2 = 0
function, yC.F and (b) particular integral yP.I .
C. F ⇒ yC.F = (c1 + c2 z)ez = [c1 + c2 (log x)]x
To find C.F we have to solve the homogeneous
equation P. I
(D2 + 4)y = 0 (2) To find P.I, we write
1
A.E is m2 + 4 = 0 ⇒ m = ±2i yP.I = z = (1 − θ)−2 z (3)
(θ − 1)2
YC.F = c1 cos 2x + c2 sin 2x (3) = (1 + 2θ + 3θ 2 + · · · )z
= z + 2 = log x + 2 (4)
P. I
To find P.I, we write The complete solution is
1 y = yC.F + yP.I = [c1 + c2 (log x)]x + log x + 2 (5)
yP.I = (ex + sin 2x)
+4
D2 / ,
s2
1 1 6. (a) Find L−1 using convolu-
= 2 ex + 2 sin 2x (s2 + 4)(s2 + 9)
D +4 D +4 tion theorem.
x
e x sin 2x
= − (4)
5 4 Solution Please refer to Ex. 12.29 (iii) in chapter
The complete solution is 12, Page 25.

y = yC.F + yP.I
4a
(x2 − y2 )
y
6. (b) Show that dxdy
y2 /4a x + y
2 2
ex x sin 2x 0
= c1 cos 2x + c2 sin 2x + − . (5) = 8a2 (π/2 − 5/3).
5 4
5. (b) Solve the differential equation Solution The region of integration is that bound-
(x2 D2 − xD + 1)y = log x. ed by the st. line y = x and parabola y2 = 4ax.
To change to polars put x = r cos θ, y =
Solution The given differential equation r sin θ, dx dy = rdr dθ.
The limits for θ are θ = π/4 (along y = x) to
(x2 D2 − xD + 1)y = log x (1) θ = π/2 (y-axis). The limits for r are r = 0 at
4a cos θ F̄d r̄ = 3x2 dx + dy + zdz

0 to r = on y2 = 4ax. Also x2 − y2 =
sin2 θ (2,1,3) (2,1,3)
r 2 (cos2 θ − sin2 θ), x2 + y2 = r 2 ∴ F̄d r̄ = 3x2 dx + dy + zdz
The given integral becomes (0,0,0) (0,0,0)
(2,1,3)
z2
= x3 + y +
y 2 (0,0,0)
a)
9 27
(4a, 4 parabola =8+1+ = .
2 2
y2 = 4a
8. Verify divergence theorem for 2x2 yī − y2 j̄ + 4xz 2 k̄
y=x
o x taken over the region of first octant of the cylinder
y2 + z 2 = 9, x = 2.
Solution Divergence
theorem:

· F̄dv = F̄ · ndx (1)
Figure 6b v s
x
4a y
(x2 − y2 )
I= dxdy
0 y2 /4a x2 + y2 y2 + z2 = 9, x = 2
C
π/2 4a cos θ B'
sin2 θ
= (cos2 θ − sin2 θ)rdrdθ A'
θ=π/4 r=0
π/2 4a cos θ/ sin2 θ
r2
= (cos θ − sin θ) 2 2
dθ
π/4 2 0
π/2 O
z
= 8a2 (cot4 θ − cot2 θ)dθ B
π/4
A y2 + z2 = 9, x = 0
3π − 8 π π 5
= 8a2 + − 1 = 8a2 − .
12 4 2 3 y
7. (a) Find curl (r̄f (r)) where r̄ = xī + yj̄ + z k̄ and Figure 8
|r̄| = r.
Solution F̄ = 2x2 yī − y2 j̄ + 4xz 2 k̄

curl (r̄f (r)) = × (r̄f (r)) ∂ ∂ ∂
· F̄ = (2x2 ) − y2 + 4xz 2
= × r̄f (r) + f (r) × r̄ ∂x ∂y ∂z
r̄ = 4xy − 2y + 8xz
= 0̄ + f (r) .
r
· F̄
7. (b) Find the work done in moving a particle in the V
force field F̄ = 3x2 ī + j̄ +z k̄ along the straight √ 2 3 9−y2
line from (0,0,0) to (2,1,3). = (4xy−2y+8xz)dz dy dx
x=0 y=0 z=0
Solution Work done 2 3 √
9−y2
P2 = [(4xy−2y)z+4xz ]0 2
dydx
W = F̄d r̄
x=0 y=0
P1 2
= (1−2x)(−2y) 9−y2 +4x(9−y2 )dy dx
r̄ = xī + yj̄ + z k̄ ⇒ d r̄ = dxī + dyj̄ + dz k̄ x=0
⎧ 3 ⎫
⎨
2 2 3/2 3
3 ⎬ On S5 : y2 + z 2 = 9,
9−y 9y
= (1−2x) +4x 9y− dx
⎩ 3/2 3 0⎭
x=0
0 (y2 + z 2 ) 2yj̄ + 2z k̄
2 n̄ = =
2 | (y + z )|
2 2
4y2 + 4z 2
= (1 − 2x)(0 − 27) + 4x(27 − 9) dx
0 3 yj̄ + z k̄ yj̄ + z k̄
2 =√ = ;
4×9 3
= [−18(1 − 2x) + 72x] dx
0 −y3 + 4xz 3 z
F̄ · n̄ = ⇒ n̄ · k̄ =
= [−18(2 − 4) + 36(4)] = 180 3 3
1
= 9 − y2
S = S1 ∪ · · · ∪ S5 : 3
dxdy
where S1 : OAB (in x = 0); ∴ F̄ · n̄ds = F̄ · n̄
S5 R |n̄ · k̄|
S2 : OBB C (in y = 0);
S3 : OAA C (in z = 0); where R is the projection of S5 on the xy − plane
S4 : CA B (in x = 2);
4xz 3 − y3
S5 : ABB A (curved Surface) ; = dxdy
R 9 − y2
On S1 : x = 0, n̄ = −ī; 2 3

On S2 : y = 0, n̄ = −j̄ = 4x(9 − y2 ) − y3 (9 − y2 )−1/2 dy dx
x=0 y=0
On S3 : z = 0, n̄ = −k̄ 2 2
= 72xdx − 18 dx
0 0
∴ F̄ · n̄ds = 0 (2) 2
x
S1 , S2 , S3 = 72 − 18[x]20
2
On S4 : x = 2, n̄ = −ī; F̄ · n̄ = 8y
√ = 144 − 36 = 108 (4)
3 9−z2
⇒ F̄ · n̄ds = (8ydy)dz
S2 z=0 y=0 Hence F̄ · n̄ds = 0 + 0 + 0 + 72 + 108
3 3 S
= (4y2 )+
0 9− z 2 dz =4 (9 − z 2 )dz [by Eqs.(3), (4), (5)]
z=0 0
3
z3 Equality of Eqs. (2) and (6) verifies the divergence
= 4 9z − = 72 (3) theorem (1).
3 0
I year B. Tech. Supplementary Examinations, Aug/Sep, 2006
1. (a) Test the convergence of the following series 1

% (n!)2 Solution Let un = .
n(log n)2
x2n (x > 0). We have u2 > u3 > · · · and
(2x)!
1
lim un = lim = 0.
Solution Here n→∞ n→∞ n(log n)2
By Leibnitz’s test, the given alternating series is
(n!) 2n2
[(n + 1)!]2 2n+2 convergent.
un = x ; un+1 = x
(2n)! (2n + 2)! Also by Cauchy–Maclaurin Integral Test,
% % 1
un+1 [(n+1)!]2 (2n)! 2 (n+1)2 |un | = is convergent.
= · ·x = ·x2 n(log n)2
un (2n+2)! (n!)2 (2x+1)(2x+2)
∴ The given series is absolutely convergent.
(1+1/n)2
= 2 x2
2 (1+1/2n )(1+1/n) 1. (c) Expand sin3 x as a power series in x up to the
⎧ 2
⎨x term containing x3 .
< 1 cgt
→ 4
⎩ > 1 dgt Solution
3 1
f (x) = sin3 x = sin x − sin 3x, f (0) = 0
by D’Alembert’s Ratio Test. When x = 4,
2 4 4
3 3
un 2n + 1 f (x) = cos x − cos 3x, f (0) = 0
n −1 =n −1 4 4
un+1 2n + 2 3 32
f (x) = − sin x + sin 3x, f (0) = 0
−n 1 4 4
= → − as n → ∞
2n + 2 2 3 33
f (x) = − cos x + cos 3x,
By Raabe’s test, the series diverges. 4 4
Hence the given series converges for x2 < 4 and −3 3 2
f (0) = + =6
diverges for x2 ≤ 4. 4 4
3 34
f v(x) = sin x − sin 3x, f v (0) = 0
1. (b) Test the following series for absolute/ 4 4
% ∞
(−1)n
conditional convergence . By Maclaurin’s theorem (Taylor’s theorem
n=2
−n(log n)2 for x = 0)
x x2 x3 2. (b) Find the shortest distance from the origin to

f (x) = f (0) + f (0) + f (0) + f (0)
1! 2! 3! the surface xyz 2 = 2.
x4 v
+ f (0) + · · · Solution Let P(x, y, z) be any point on the
4!
Substituting the above values, we have surface
x x2 x3 xyz 2 = 2 (1)
sin3 x = 0 + ·0+ · 0+ · 6 + 0 + · · · = x3 .
1! 2! 3!
2 Then OP = d = x2 + y2 + z2
y
2. (a) If u = tan−1 Show that 2
x ⇒ f (x, y) = d 2 = x2 + y2 + (2)
∂u2
∂u
2
∂2 u xy
x2 2 + 2xy + y2 2 = − sin2 u sin 2u.
∂x ∂x∂y ∂y 2 2
2 Now fx = 2x − 2
; fy = 2y − 2 .
y xy xy
Solution u = tan−1 is not a homogeneous Solving fx = 0 and fy = 0, we get
x 2
y x3 y − 1 xy3 − 1
function. If we write it as tan u = then it is a = 0 and =0
x x2 y xy2
homogeneous function in x and y of degree 1.
By Euler’s theorem, we have ⇒ x3 y = 1, xy3 = 1 ⇒ xy(x2 − y2 ) = 0
∂u ∂u
x (tan u) + y (tan u) = (tan u) (1) The stationary points are P1 = (1, 1) and
∂x ∂y
P2 = (1, −1).
∂u ∂u
or x sec2 u + y sec2 u = tan u
∂x ∂y 4 4
Also, fxx = 2 + , fyy = 2 +
∂u ∂u 1 2 tan u 1 x3 y xy3
or x + y = = sin 2u (2)
∂x ∂y 2 1 + sec u
2 2 2
and fxy =
Differentiating Eq. (2) partially w.r.t x and y x2 y2
respectively At P1 = (1, 1) l = fxx = 6 > 0,
∂2 u ∂u ∂2 u 1 ∂u n = fyy = 6, m = fxy = 2
x +1 +y = · 2 cos 2u
∂x 2 ∂x ∂x∂y 2 ∂x and ln − m = fxx fyy − = 6.6 − 22 = 32 > 0
2
fxy2
∂u
2
∂u
2
∂u 2 √
or x 2 + y = (cos 2u − 1) (3) Now z 2 = = 2 by (1) ⇒ z = ± 2
∂x ∂x∂y ∂x 1
∂u
2
∂u
2
∂u 1 ∂u √
x + y 2 + 1 = · 2 cos 2u So, minimum occurs at (1.1, 2) and the shortest
∂x∂y ∂y ∂y 2 ∂y
√
∂u
2
∂u2
∂u distance from the origin to the surface is 1 + 1 + 2
or x + y 2 = (cos 2u − 1) (4)
∂x∂y ∂y ∂y or 2.
Multiplying Eq. (3) by x and Eq. (4) by y and
adding we obtain 3. (a) Trace the hypocycloid x = cos3 θ, y = sin3 θ.
∂2 u ∂2 u 2∂ u
2
Solution Please refer to Ex. 6.11 on p. 10 of Ch.
x2 + 2xy + y
∂x2 ∂x∂y ∂y2 6 (we have to choose a = b = 1).

∂u ∂u
= (cos 2u − 1) x + y 3. (b) Find the surface area generated by rotating the
∂x ∂y
x
∂u arc of the catenary y = a cosh from x = 0
= (cos 2u − 1) = − sin2 u · sin 2u (5) a
∂x to x = a about the x-axis.
Solution The curve is symmetrical about the y- dy 2x

x x 4. (b) Solve the differential equation + y=
axis since y = a cosh = a cosh . dx 1 + x2
a a 1
The x-axis is called the directrix for the curve given y = 0 when x = 1.
1 + x2
called catenary. The curve meets the y-axis at A =
(0, a); it does not meet the x-axis.
y
P dy 2x 1
s + y= (1)
dx 1 + x 2 1 + x2
A (0, a)
is a first order linear equation with
0 x 2x 1
P= , Q=
1 + x2 1 + x2

Figure 3b 2x dx
I.F = e Pdx = e 1+x2
2
x = elog(1+x ) = (1 + x2 )
y = a cosh (1)
a
dy x Its general solution is
⇒ = sinh (2)
dx a
y(I.F) = y(1 + x2 ) = c + Q(I.F)dx
If s = AP is the arc, then
1
2 * =c+ (1 + x2 )dx
ds dy x x (1 + x2 )2
= 1+ = 1 + sinh2 = cosh = c + tan−1 x (2)
dx dx a a
The required surface area Given condition: y = 0 when x = 1

a
ds π
= 2πy dx ⇒ 0(1 + 1) = c + tan−1 1 = c +
dx 4
0
a
x x ∴ The required solution is
= 2πa cosh cosh dx
a a
a0 y(1 + x2 ) = tan−1 x − π/4.
2x
= πa cosh + 1 dx
0 a 4. (c) Obtain the orthogonal trajectories for the
a
a 2x family r n = an cos nθ.
= πa sinh +x
2 a
0 Solution By logarithmic differentiation of
sinh 2
= πa2 +1 .
2 r n = an cos nθ (1)
4. (a) Form the differential equation by eliminating 1 dr
we get = − tan nθ (2)
the arbitrary constant c from tan x tan y = c. r dθ
Solution
which is the differential equation for the given family
tan x tan y = c (1) of curves. To obtain the differential equation for the
dx dθ
Differentiating Eq. (1) w.r.t x, we get orthogonal system, we have to replace by −r 2
dθ dr
dy in Eq. (2). So, we get
sec2 x tan y + tan x sec2 y =0 (2)
dx dθ
r = tan nθ (3)
which is the required differential equation. dr

Separating the variable and integrating, we get e−3x 3 13 10
= x + 2x2 + x +
12 6 9
dr cos nθ
= dθ The complete solution is
r sin nθ
⇒ n log r = log sin nθ + constant y = yC.F + yP.I

⇒ r n = cn sin nθ (4) −x e−3x 3 13 10
= c1 e +c2 e +
3x
x + 2x + x +
2
12 6 9
(equation of the required orthogonal system).
d2 1 dy 12 log x
(D2 − 2D − 3)y = x3 e−3x . + = .
dy2 x dx x2
Solution The given equation (D2 − 2D − 3)y =
3 −3x Solution The given equation can be written as
x e is a non-homogeneous linear differential equa-
tion. Its complete solution consists of d2 dy
(a) complementary function yC.F and x2 2
+x = 12 log x (1)
dy dx
(b) particular integral yP.I
C.F. which is Euler–Cauchy Equidimensional equation. If
To get C.F, we have to solve the homogeneous d
we put x = ez or z = log x denote = 0 then
equation dz
(D2 − 2D − 3)y = 0 d d2
x = xD = θ, x2 2 = x2 D2 = θ(θ − 1)
A.E is m2 − 2m − 3 = 0 dx dx
(obtained by taking y = emx ) the equation transforms to
⇒ (m − 3)(m + 1) = 0 ⇒ m = 3 or − 1
[θ(θ − 1) + θ]y = 12z or θ 2 = 12z (2)
−x
yC.F = c1 e + c2 e
3x
Integrating twice we obtain the general solution
P.I
To get P.I., we write y = 2z 3 + c1 z + c2 = 2(log x)3 + c1 (log x) + c2
1 Aliter:
yP.I = e−3x · x3
D2 − 2D − 2 The equation can be written as
1
= x3 1 d d 12 log x
(D − 3) − 2(D − 3) − 3
2
x y=
1 x dx dx x2
= e−3x 2 · x3
D − 8D + 12 d d 12 log x
⇒ x =
e−3x 2 1 2 −1 3 dx dx x
= 1− D− D x
12 3 12 Integrating we get

e−3x 2 1 2 2 1 2 2 dy
= 1+ D− D + D− D x = 6(log x)2 + c1
12 3 12 3 12 dx
3
dy 6 1
+
2 1
D − D2 + · · · x 3 ⇒ = (log x)2 + c1
3 12 dx x x
−3x
Integrating again we get
e 2 13 2 5 3
= 1 + D + D + D + · · · x3
12 3 36 27 y = 2(log x)3 + c1 (log x) + c2 .
6. (a) Solve the differential equation Solution

√ The region R of integration is y = 0 to
a2 − x2 ; x = 0 to a. That is, the first quadrant of
d2 dx the circle x2 + y2 = a2 .
− 4 − 12t = e3t
dt 2 dt Transformation formulae
given that x(0) = 1 and x (0) = −2 using Laplace
transforms. x = r cos θ, y = r sin θ, dxdy = rdrdθ.
Solution The given I.V.P is The limits for r are 0 to a and for θ are 0 to π/2.
x − 4x − 12x = e3t , x(0) = 1, x (0) = −2
The given integral is
√
a a2 −x2
L{x(t)} = x̄(s) ⇒ L{x } = sx̄(s) − x(0), I= y x2 + y2 dxdy
L{x } = s2 x̄(s) − sx(0) − x (0) 0 0
π/2 a
⇒ L{x } = sx̄ − 1, L{x } = s2 x̄ − s · 1 + 2 = r 3 sin θdrdθ
θ=0 r=0
4 a
using the initial conditions. π/2 r a4
The given I.V.P becomes = [− cos θ]0 = .
4 0 4
1
(s2 x̄ − s + 2) − 4(sx̄ − 1) − 12x̄ = 7. (a) If Ā is irrotational vector evaluate div(Ā × r̄)
s−3
where r̄ = xī + yj̄ + z k̄.
1
∵ L e3t =
s−3 Solution
1 div(Ā × r̄) = · (Ā × r̄)
⇒ (s − 4s − 12)x̄ = s − 6 +
2
s−3
= ( × Ā) · r̄ − ( × r̄) · Ā = 0̄
1
⇒ (s + 2)(s − 6)x̄ = (s − 6) +
s−3 because (i) Ā is irrotational ⇒ × Ā = 0̄
1 1 (ii) × r̄ = 0̄
⇒ x̄ = + ;
s + 2 (s − 3)(s − 6)(s + 2)
7. (b) If F̄ = xyī − z j̄ + x 2 k̄ and C is the curve
Taking inverse Laplace transforms, we obtain
x = t 2 , y = 2t, z = t 3 from t = 0 to 1, evaluate
/ , / ,
−1 1 −1 1 F̄d r̄.
x(t) = L +L
s+2 (s−3)(s−6)(s+2) C
/ ,
−2t −1 A B C Solution
=e +L + +
s−3 s−6 s+2 F̄ = xyī − z j̄ + x2 k̄ = 2t 3 ī − t 3 j̄ + t 4 k̄;
41 −2t 1 1
⇒ x(t) = e − e−3t + e6t d r̄ = dxī − dyj̄ + dz k̄ = (2t ī − 2j̄ + 3t 2 k̄)dt
40 15 24
1 1 1 1
∵A= =− ; ¯ =
F̄ · dr (4t 4 − 2t 3 + 3t 6 )dt
(s − 6)(s + 2) s=3 15 t=0 0

B=
1
=
1 4 5 2 4 3 7 1
; = t − t + t
(s − 3)(s + 2) s=6 24 5 4 7 0

1 1 4 1 3 51
C= = = − + = .
(s − 3)(s − 6) s=−2 40 5 2 7 70
6. (b) Evaluate by √
transforming into polar coordi- 8. Verify Stokes’s theorem for F̄ = (y − z + 2)ī+
a a2 −x2
(yz + 4)j̄ − xz k̄ where S is the surface of the cube
nates y x2 + y2 dxdy. x = 0, y = 0, z = 0, x = 2, y = 2, z = 2 above
0 0
the xy-plane. (ii) On S2 : POAQ (y = 0) y = 0; dy = 0;

n̄ds = −jdzdx; z, x vary from 0 to z
Solution Stokes’ theorem is

curl F̄ · nds = F̄d r̄ (1) curl F̄ · n̄ds
S C S2
2 2
ī j̄ k̄ = (−y, z − 1, −1) · (0, −1, 0)dzdx
∂ ∂ ∂
Curl F̄ =
z=0

x=0
2
∂x ∂y ∂z 2 2
z2
y−z+2 yz + 4 −xz = (1−z)dzdx = z− [x]20 = 0
z=0 x=0 2 0
(iii) On S3 : QABR (x = 2) x = 2; dx = 0;
z
n̄ds = −idydz; y, z vary from 0 to z

P S curl F̄ · n̄ds
S3
Q R
2 2
= (−y, z − 1, −1) · (1, 0, 0)dydz
y=0 z=0
C y 2 2 2
A y
B = −ydydz = − [z]20 = −4
y=0 z=0 2
x
Figure 8 (iv) On S4 : RBCS (y = 2) y = 2; dy = 0;
Here the surface S composed of surfaces normal n̄ds = jdzdx; z, x vary from 0 to z
S1 : OCSP(x = 0) − ī
curl F̄ · n̄ds
S2 : POAQ(y = 0) − j̄ S4
2 2
S3 : QABR(x = 2) ī
= (−y, z − 1, −1) · (0, 1, 0)dzdx
S4 : RBCS(y = 2) j̄ z=0 x=0
2 2 2
S5 : SPQR(z = 2) k̄ z2
= (z−1)dzdx = −z [x]20 = 0
z=0 x=0 2 0
ī j̄ k̄
∂ ∂ ∂ (v) On S5 : SPQR (z = 2) z = 2; dz = 0;
Curl F̄ = ∇ × F̄ =
∂x ∂y ∂z n̄ds = −kdxdy; x, y vary from 0 to z
y−z+2 yz + 4 −xz
= (−y, z − 1, −1), i.e., − yī + (z − 1)j̄ − k̄ curl F̄ · n̄ds
S5
(i) On S1 : OCSP (x = 0) x = 0; dx = 0; 2 2
n̄ds = −idydz; y, z vary from 0 to z = (−y, z − 1, −1) · (0, 0, 1)dxdy
x=0 y=0

curl F̄ · n̄ds 2 2
S1 = −1dxdy = [−x]20 [y]20 = −4
2 2 x=0 y=0
= (−y, z − 1, −1) · (−1, 0, 0)dydz

y=0 z=0 Adding (i)–(v) we obtain LHS of (1)

2 2

= ydydz = 4
y=0 z=0 curl F̄ · n̄ds = 4 + 0 − 4 + 0 − 4 = −4
S
The surface S is not included in the xy-plane and (vii) On C2 : x = z = 0, dx = dz = 0; y

the bounding curve C : OABC is the xy-plane so that varies from 0 to 2
x = 0 and dz = 0. So we get 2 2
F̄ · d r̄ = (y + 2) · 0 + 4dy = 4y = 8
C2 0 0
F̄d r̄ = [(y − x + 2)ī + (yz + 4)j̄ + z k̄·
C C
(viii) On C3 : y = 2, z = 0, dy = dz = 0; x
(dxī + dyj̄ + 0)] varies from 2 to 0

0
= (y + z)dx + 4dy 0
C F̄ · d r̄ = (2 + 2)dx + 0 = 4x = −8
C3 2 2
C composed of the following st. line segments:
(ix) On C4 : x = z = 0, dx = dz = 0; y varies
C = C 1 ∪ C2 ∪ C3 ∪ C4 from 2 to 0
0 0
where C1 : OA; C2 : AB C3 : BC and C4 : CO F̄ · d r̄ = 4dy = 4y = −8
2
(vi) On C1 : y = z = 0, dy = dz = 0; x varies C4 2
from 0 to 2
2 Adding (vi)–(ix) we get RHS of Eq. (1) =
2 C
F̄ · d r̄ = 4 + 8 − 8 − 8 = −4. The equality of
F̄ · d r̄ = 2dx = 2x = 4 Eqs. (4) and (6) verifies Stokes’ theorem.
C1 0 0
1. (a) Test the convergence of the series |un | sin √1n

Now lim = lim · n3/2
2 2 · 5 · 8 2 · 5 · 8 · 11 n→∞ vn n→∞ (n − 1)
+ + + ··· . 0
1 1 · 5 · 9 1 · 5 · 9 · 13 sin √1n √1n
= lim =1→0
$ (1 − n )
Solution Let n→∞ 1
$ $
2 · 5 · 8 · 11 · (3n − 1) Both un and $ vn behave alike. Since vn
un = ;
1 · 5 · 9 · 13 · · · (4n − 3) is convergent; so in |un |. ∴ The given series is
2 · 5 · 8 · 11 · (3n − 1)(3n + 2) absolutely convergent.
un+1 =
1 · 5 · 9 · 13 · · · (4n − 3)(4n + 1) 1. (c) State and prove Rolle’s theorem.
un+1 3n + 2 3 + 2n 3
= = → < 1(as n → ∞) Solution Please refer to Theorem 3.1 in Ch. 3
un 4n + 1 4+ n 1
4 on p. 1 (Revised Edition).
$
By D’Alembert’s ratio test, un converges. 2. (a) If x = u cos v, y = u sin v, prove that
% (−1) sin n √1 ∂(x, y) ∂(u, v)
n · = 1.
1. (b) Find whether the series is ∂(u, v) ∂(x, y)
(n − 1)
absolutely convergent or conditionally Solution
convergent. x = u cos v, (1)
y = u sin v (2)
Solution The given series which is of the
sin √1n Differentiating we get
$ ∂x ∂x
form (−1) un where un =
n
for n ≥ 2 is = cos v = −u sin v
(n − 1) ∂u ∂v
an alternating series whose terms are decreasing ∂y ∂y
sin √1n = sin v = u cos v
(u2 > u3 > u4 · · · ) and lim un = lim = 0. ∂u ∂v
n→∞ n→∞ (n − 1) ∂x ∂x
By Leibnitz’s test, the $ given series is $convergent. ∂(x, y) cos v −u sin v
Consider the series |un | and vn (where = ∂u ∂v
∂y ∂y = sin v
∂(u, v) u cos v
1 ∂u ∂v
vn = 3/2 ) which is known to be convergent
n = u(cos2 v + sin2 v) = u
(p-series with p = 3/2 > 1).
By Eq. (2)/(1) we get 2 tan θ/2

= a(θ − sin θ) + 2a ·
y 1 + tan2 θ/2
tan v = y/x or v = tan−1 (3)
x (1 + y12 ) 1 + cot2 θ/2
ȳ = y + = a(1 − cos θ) − 4a
Again (1) + (2) gives
2 2
y2 cosec4 θ/2
u2 = x2 + y2 or u = x2 + y2 (4) = a(1 − cos θ) − 4a sin θ/2
2
Differentiating Eqs. (4) and (3) = a(1 − cos θ) = 2a(1 − cos θ) = −a(1 − cos θ)
∂u x ∂u y Hence the locus of x̄, ȳ is a cycloid.

= =
∂x x2 + y2 ∂y x2 + y2
3. Trace the curve r = a(1 + cos θ). Show that the
∂v − xy2 ∂v − 1x volume of revolution of it about the initial line
= = 8πa3
∂x 1 + yx2
2
∂y 1 + yx2
2
is .
3
−y x
= 2 = 2
x + y2 x + y2 Solution First part: Please refer to Ex. 6.17 on
x y pp. 15–16 of Ch. 6.
∂(u, v) x +y
2 2 x + y2
2 Second part: The given curve is symmetrical about
= −y x the initial line θ = 0 and the volume is generated
∂(x, y)
x2 + y2 by rotating the upper half of the curve when θ varies
x + y2
2
from θ = 0 to π
x2 + y2 1 π
= = ∴ Volume =
2 3
πr sin θdθ
(x2 + y2 )3/2 x2 + y2
0 3
1 π
= , by Eq. (4) 2π
u = π a3 (1 + cos θ)3 sin θdθ
3
∂(x, y) ∂(u, v) 1 0

Hence · =u· =1 2πa3 π
∂(u, v) ∂(x, y) u =− (1 + cos θ)3 d(1 + cos θ)
3 0
2. (b) Show that the evolute of the cycloid x = a(θ− 2πa3 (1 + cos θ)4 π
=−
sin θ), y = a(1 − cos θ) is another cycloid 3 4 0
x = a(θ + sin θ), y = a(1 − cos θ). −πa3 πa3

= (0 − 16) = .
6 3
Solution We have
dy dy/dθ θ 4. (a) Form the differential equation of the family
y1 = = = cot ; of curves log(y + a) = x2 + c, c being the
dx dx/dθ 2
parameter.
d 2y d d dθ
y2 = 2 = (cot θ/2) = (cot θ/2) ·
dx dx dθ dx Solution Differentiating w.r.t x, we get
1 4θ 1 dy dy
= − cosec = 2x or = 2x(y + a)
4a 2 y + a dx dx
dy θ θ which is the required differential equation.
∵ = a sin θ = 2a sin cos ;
dθ 2 2
dx θ 4. (b) Solve the differential equation
= a(1 − cos θ) = 2a sin2
dθ 2 dy
3 − y cos x = y4 (sin 2x − cos x).
If (x̄, ȳ) is the centre of curvature at any point of dx
the curve
Solution Dividing the given equation by 3, we
y1 (1 + y12 ) 4a cot θ/2 dy y y4
x̄ = x − = a(θ − sin θ) + can write it as − cos x = (sin 2x − cos x)
y2 cosec2 θ/2 dx 3 3
dy When t = 10, θ = 65 so that y = 65 − 25 = 40.

which is Bernoulli’s equation of the form + Py =
dx
Qy (n=4 here). Multiplying the equation by y − 4,
n 4
This gives 40 = 50e−10t ⇒ e−10k = (5)
we have 5
dy 1 −3 1 we have to find y for t = 20.
y−4 − y cos x = (sin 2x − cos x)or 2
dx 3 3
∴ y = 50e−20k = 50 e−10k
1 d −3 1 −3 1 2
− (y ) − (y ) cos x = (sin 2x − cos x) 4
3 dx 3 3 = 40 by (5)
dz 5
or + z cos x = cos x − sin 2x
dx 16
or y = 50 = 32 (6)
which is a linear equation with P = cos x, 25
Q = cos x − sin 2x (z = y−3 ) ⇒ θ − 25 = 32 or θ = 32 + 25 = 57 (7)

I.F = e Pdx
=e cos xdx
=e sin x ◦
We have to find t when θ = 55 C. We have
10t
The complete solution is −kt −k t 3 4
30 = 20e ⇒ (e ) = = by Eq. (5)
−3
5 5
y ·e sin x
= c + (cos x − sin 2x)esin x dx
Taking logs on both sides
sin x
=c+ e d(sin x) − 2 sin xesin x d(sin x) t 4 3
log = log or
sin x
10 5 5
= c+esin x−2(sin x−1)e = c+(3−2 sin x)esin x log 3 − log 5
t = 10
1 log 4 − log 5
or 3 = ce− sin x + 3 − 2 sin x.
y 1.0986 − 1.60943
= 10
4. (c) An object whose temperature is 75◦ C cods is 1.38629 − 1.60943

an atmosphere of constant temperature 25◦ C −0.51083
= 10 = 22.89 .
at the rate kθ, θ being the excess temperature −0.22314
of the body over the temperature. If after 10
minutes the temperature of the object falls to
(D3 + 1)y = cos(2x − 1).
65◦ C find its temperature after 20 minutes.
Find the time required to cool down to 55◦ C. Solution The given equation
Solution The differential equation governing (D3 + 1)y = cos(2x − 1) (1)
Newton’s law of cooling is a non-homogeneous linear differential equation.
dθ Its complete solution consists of
= −k(θ − θ0 ) (1)
dt (a) complementary function, yC.F and
where θ0 is the temperature of atmosphere and θ0 = (b) particular integral, yP.I
25◦ C. Writing θ − θ0 = y Eq. (1) becomes To find C.F we have to solve
dy (D3 + 1)y = 0 (2)

= −ky (2)
dt Putting y = e we get A.E√m + 1 = 0
mx 3
whose solution is y = ce−kt (3) 1±i 2

whose roots are m = 1,
Initially when t = 0, θ = 75 and θ0 = 25 2
√ √
⇒ y = c = 50 3 3
−x
∴ yC.F = c1 e +e x/2
c2 cos x+c3 sin x (3)
Substituting in Eq. (3), we get y = 50e−kt (4) 2 2

To find P.I, we write 1 1 sin 2x
= (1 + cos 2x)dx = x+
1 2 2 2
yP.I = cos(2x − 1)
D3 + 1 The complete solution is
1
1 y = yC.F + yP.I = c1 cos x + c2 sin x − − sin2 x cos x
= cos(2x − 1) replacing D2 by − 4 2
−4D + 1 1 1
1 + 4D + x sin x + sin x sin 2x.
= cos(2x − 1) 2 4
1 − (4D)2
6. (a) Find the Laplace transform of the following
cos(2x − 1) − 8 sin(2x − 1)
= function : t e−t sin 2t.
1 − 16(−4)
1 Solution
= [cos(2x − 1) − 8 sin(2x − 1)] 2
65 L{sin 2t} = = f¯ (s)
The complete solution is +4 s2
d
y = yC.F + yP.I ⇒ L{tf (t)} = L{t sin 2t} = f¯ (s)
√ √ ds
3 3 4s
−x
= c1 e + e (c2 cos
x/2
x + c3 sin x) = 2
2 2 (s + 4)2
1 By First shifting theorem
+ [cos(2x − 1) − 8 sin(2x − 1)].
65 4s
5. (b) Solve the differential equation (D2 + 1)y = L{e−t (t sin 2t)} =
(s2 +4)2 s→s+1
cos x by the method of variation of parameters. 4(s + 1)
=
Solution The given equation ((s + 1)2 + 4)
4(s + 1)
(D2 + 1)y = cos x (1) = 2 .
(s + 2s + 5)2
is a non-homogeneous linear differential equation 6. (b) Using Laplace transform solve
with P = 0, Q = 1, R = cos x. Its complete solu- y + 2y + 5y = e−t sin t given that
tion consists of (a) C.F and (b) P.I y(0) = 0, y (0) = 1.
C.F is found by solving the homogeneous equa-
tion. (D2 + 1)y = 0. Solution Please refer to Ex. 12.44(ii) on p. 41
Putting y = emx we get of Ch. 12.
A.E m2 + 1 = 0 ⇒ m = ±i
5 x2
∴ C.F is yC.F = c1 cos x + c2 sin x (2) 6. (c) Evaluate x(x2 + y2 )dx dy.
P.I is found by letting yP.I = A cos x + B sin x (3) 0 0
Solution
and varying the parameters A, B. 5 x2
Let u = cos x, v = sin x [x(x2 + y2 )dy]dx
x=0 y=0
w(u, v)(x) =
cos x sin x
= 1 = 0 2
− sin x cos x 5 x
= (x + xy )dy dx
3 2
vR sin x cos x x=0 y=0
A=− dx = − dx
w 1 5 x 2 5
xy3 x7
1 = x3 y + dx = x5 + dx
= − sin2 x x=0 3 0 x=0 3
2 5

uR cos2 x x6 x8 56
B=− dx = − dx = + = · 29.
w 1 6 24 0 24
7. (a) Evaluate ∇ 2 (log r) where r = x2 + y2 + z 2 . From Eqs. (6) and (4)
Solution Please refer to Ex. 10.19 on p. 442 of 1 2 3

Engg. Maths-I (I Edition) where it was proved that φ= x + 2xy + 4xz − y2 − yz + h(z) (7)
2 2
∇ 2 f (r) = f (r) + 2r f (r).
1 −1 Differentiating Eq. (7) w.r.t z, we get
Take f (r) = log r ⇒ f (r) = , f (r) = 2
r r
we get ∇ 2 (log r) = 0. ∂φ
= 4x − y + h (z)
∂z
7. (b) Find constants a, b, c so that the vector = 4x − y + 2z
Ā = (x + 2y + az)ī + (bx − 3y − z)j̄
⇒ h (z) = 2z ⇒ h(z) = z 2 + c (8)
+(4x + 8y + 2z)k̄ is irrotational. Also find φ
such that Ā = ∇φ. 1 3
∴ φ(x, y, z) = x2 + 2xy + 4xz − y2
2 2
Solution − yz + z 2 + c. (9)
Ā is irrotational ⇔ curl Ā = 0
i j k 8. Verify Stokes’ theorem for the vector field
∂ ∂ ∂ F̄ = (2x − y)ī − yz 2 j̄ − y2 z k̄ over the upper half
= of the surface of x2 + y2 + z 2 = 1 bounded by
∂x ∂y ∂z
x+2y+az (bx−3y−z) (4x+cy+2z) its projection on the xy-plane.
= ī(c + 1) + j̄(a − 4) + k̄(b − 2)

Z
⇒ a = 4, b = 2, c = −1
Ā = ∇φ (1)
⇒ (x + 2y + 4z)i + (2x − 3y − z)j̄ S
+ (4x − y + 2z)k̄ (2)

o Y
∂φ ∂φ ∂φ ∂φ ∂φ ∂φ C
=i +j +k ⇒ , , (3)
∂x ∂y ∂z ∂x ∂y ∂z
X
= (x+2y+4z)ī, (2x−3y−z)j̄, (4x−y+2z)k̄
Figure 8
dφ = (x + 2y + 4z)dx + (2x − 3y − z)dy
+ (4x − y + 2z)dz Solution Let C denote the boundary of the pro-
jection in the xy-plane of the upper half of the sphere
Integrating Eq. (1) partially w.r.t x, we get x2 + y2 + z 2 = 1 so that C is the circle.
1 2 C : x2 + y2 = 1, z = 0
φ(x, y, z) = x + 2xy + 4xz + g(y, z) (4) or x = cos t, y = sin t (0 ≤ t ≤ 2π)
2
Differentiating Eq. (4) partially w.r.t y, we get
∂φ ∂g F̄d r̄ = (2x − y)dx − yz 2 dy − y2 zdz
= 2x + = 2x − 3y − z by Eq. (2)
C
2π
∂y ∂y
∂g (2x − y)dx = − (2 cos θ − sin θ) sin θdθ
⇒ = −3y − z (5) C 0
∂y
1 2π
Integrating Eq. (5) w.r.t y, we get = [− sin2 θ]2π
0 + (1−cos 2θ)dθ
2 0
3 1
g = − y2 − yz + h(z) (6) = 0 + · 2π − 0
2 2
i j k where R is the projection of S on the xy-plane and

∂ ∂ ∂ dxdy
Now curl F̄ = ds =
∂x ∂y ∂z |n̄ · k̄|
2x − y (−yz 2 ) (−y2 z)
curl F̄ · n̄ds = dxdy
= 0 + 0 + k̄ S R
= area of the circle x2 + y2 = 1 = π(1)2 = π

dxdy
curl F̄ · n̄ds = k̄ · n̄ds = k̄ · n̄ The equality of Eqs. (2) and (4) verifies Stokes’
S S R |n̄k̄| theorem.
1. (a), (b), (c) same as in Set 2 pp. 34–39 of this sin θ cos x r cos θ cos φ −r sin θ sin φ
section. = sin θ sin φ r cos θ sin φ r sin θ cos φ
2. (a) If x = r sin θ cos φ, y = r sin θ sin φ and
∂(x, y, z) Expanding by R3
z = r cos θ, prove that = r 2 sin θ.
∂(r, θ, φ) = cos θ r 2 cos θ sin θ(cos2 φ + sin2 φ)
∂(r, θ, φ)
Find . + r 2 sin θ sin2 θ(cos2 φ + sin2 φ)
∂(x, y, z)
= r 2 sin θ.
Solution x = r sin θ cos φ, y = r sin θ sin φ, ∂(r, θ, φ) 1 1
z = r cos θ. = ∂(x,y,z)
= .
∂(x, y, z) r2 sin θ
Differentiating partially w.r.t r, θ, φ, we get ∂(r,θ,φ)
a a
∂x ∂x 2. (b) Find the centre of curvature at a point ,
= sin θ cos φ, = r cos θ cos φ, √ √ √ 4 4
∂r ∂θ of the curve x + y = a. Find also the
∂x equation of the circle of curvature at that
= −r sin θ sin φ
∂φ point.
∂y ∂y
= sin θ sin φ, = r cos θ sin φ, Solution Equation of the curve is
∂r ∂θ
∂y √ √ √
= r sin θ cos φ x + y = a. (1)
∂φ
Differentiating Eq. (1) w.r.t x, we get
∂z ∂z ∂z √
= cos θ, = −r sin θ, =0 1 1 dy y
∂r ∂θ ∂φ √ + √ = 0 or y1 = √
2 x 2 y dx x
∂x ∂x ∂x √
∂r ∂θ ∂φ x · 2√yy −√y· √1
1
dy1 1 2 x
∂(x, y, z) ∂y ∂y ∂y ⇒ y2 = =−
= dx ⎧ x
∂(r, θ, φ) ∂r ∂θ ∂φ ⎨y1 = −1
∂z ∂z ∂z a a
At , , 4
∂r ∂θ ∂φ 4 4 ⎩y2 =
a
If (x̄, ȳ) is the centre of curvature at (x, y) = Solution

a a
, y2 = 4ax (1)
4 4
y1 (1 + y12 ) a a 3a Differentiating Eq. (1) w.r.t x, we get
x̄ = x − = +2 = ;
y2 4 4 4
(1 + y12 ) a a 3a 2yy1 = 4a (2)
ȳ = y + = +2 =
y2 4 4 4 Eliminating 4a between Eqs. (1) and (2), we get
a a
At , , the radius of curvature is dy
4 4 y = 2x (3)
√ √ dx
(1 + y12 )3/2 23/2 2 2a 2a
ρ= = = = which is the required differential equation.
y2 4/a 4 2
a a 4. (c) In a chemical reaction, a given substance is
Equation of the circle of curvature at , is
4 4 being converted into another at a rate propor-
2 tional to the amount of transformed substance,
3a 3a 2 a2
x− + y− = . how much time will be required to transform
4 4 2 one half.
3. Trace the lemniscale of Bernoulli: r 2 = a2 cos 2θ.
Prove that the volume of revolution about the Solution The differential equation governing the
πa3 √ √ Law of decay is
initial line is √ [3 log( 2 + 1) − 2].
6 2 dx
= −kx (k > 0) (1)
Solution First part: Please refer to Ex. 6.22 on dt
pp. 19-20 of Ch. 6. whose solution is x = ce−kt (2)
Second part: Please refer to Solution of 3(b)in JNTU
Exam-Regular May/June 2006 (Set 2) on p. 9 of this (x: amount of substance remaining in grams; t: time
section. in minutes, k: constant of proportionality).
Initially, when t = 0 if x = m then c = m;
m 4m
4. (a) Solve the differential equation: t = 4, x = m − = .
dy tan y 5 5
− = (1 + x)ex sec y. From Eq. (1), we get x = me−kt , which gives
dx 1 + x 4m 1
= me−4k so that k = (log 5 − log 4). We now
Solution The given equation is reducible to linear 5 4
m
type. Multiplying it by cos y, we can write it as find the value of t for x = , we have
2
d(sin y) 1
− (sin y) = (1 + x)ex (linear in sin y) m 1
dx 1+x = me−kt ⇒ kt = log 2 ⇒ t = log 2
1 1 2 k
I.F = e− 1+x dx = e− log(1+x) = . 4 log 2 2.772
1+x ⇒ = = 12.42 min.
log 5 − log 2 0.2231
The complete solution is

1 5. (a) Solve the differential equation
(sin y) = c + ex dx = c + ex (D2 + 5D + 6)y = ex .
1+x
⇒ sin y = (1 + x)(c + ex ).
Solution The given equation
4. (b) Form the differential equation by eliminating
the arbitrary constant y2 = 4ax. (D2 + 5D + 6)y = ex (1)
is a non-homogeneous linear differential equation. Y Parabola

Its complete solution consists of (4a, 4a)
(a) complementary function, yC.F and
(b) particular integral, yP.I y=x
To find C.F we have to solve the homogeneous o X
equation
(D2 + 5D + 6)y = 0 (2) Figure 6b
Putting y = emx , we get A.E. from 0 to 4a. So the region of integration is R bounded
by st. line y = x and parabola y2 = 4ax. Let x =
m2 + 5m + 6 = 0 ⇒ (m + 2)(m + 3) = 0 r cos θ, y = r sin θ ⇒ dxdy = rdrdθ and the limits
⇒ m = −2, −3 for r are r = 0 (at the origin) and r 2 sin2 θ = 4ar cos θ
∴ yC.F = c1 e−2x + c2 e−3x (3) 4a cos θ
or r = .
sin2 θ
To find P.I, we have The limits for θ are θ = π/4 (on y = x) to θ =
π/2 (on y-axis)
1 4a y
yP.I = ex (x2 − y2 )
(D2 + 5D + 6) dxdy
y2 /4a x + y
2 2
ex 0
= 2 ( replace D by a = 1) π/4 4a cos θ/ sin θ
1 +5·1+6
e x = (cos θ − sin2 θ)rdt dθ
yP.I = (4) θ=π/4 r=0
12 π/4 4a cos θ/ sin2 θ

r2
The complete solution is = (cos θ − sin θ)
2 2
dθ
θ=π/4 2 0
ex π/4
y = yC.F + yP.I = c1 e−2x + c2 e−3x + (5) = 8a2 (cot4 θ − cot2 θ)dθ
12
θ=π/4

5. (b) Using the method of variation of parameters, 3π − 8 π π 5
solve the differential equation = 8a2 + − 1 = 8a2 − .
12 4 2 3
d 2y
+ 4y = tan 2x.
dx2 7. (a) Prove that curl (Ā× B̄) = Ā div B̄− B̄ div Ā
+(B̄ · ∇) Ā − (Ā · ∇) B̄.
Solution Please refer to Ex. 2.64 (3) on p.38 of
Ch. 2 (Take a = 2.) Solution Please refer to Ex. 10.4 Identity 4(b)
/ , on p. 15 of Ch. 10.
−1 s2
6. (a) Find L using Convolu-
(s2 + 4)(s2 + 9) 7. (b) Find the directional derivative of φ(x, y, z) =
tion Theorem. x2 yz + 4xz 2 at the point (1, −2, −1) in the
direction of the surface f (x, y, z) = x log z−y2
Solution Please refer to Ex. 12.29(iii) on p. 25
at (−1, 2, −1).
of Ch. 12.
Solution
4a y
(x2 − y2 )
6. (b) Show that dx dy =
0 y2 /4a x2 + y2 φ(x, y, z) = x2 yz + 4xz 2
π 5 ∂φ ∂φ ∂φ
8a2
− . ⇒ ∇φ = ī + j̄ + k̄
2 3 ∂x ∂y ∂z
Solution x varies form y2 /4a to y and y varies = (2xyz + 4z 2 )ī + x2 z j̄ + (x2 y + 8xz)k̄
P = (1, −2, −1) 8. (b) Apply Stokes’ theorem to evaluate

ydx + zdy + xdz where C is the curve of the
(∇φ)P = [(−2)(−1) + 4(−1)2 ]ī + [(−1)2 (−1)]j̄ C
+ [(−1)2 (−2) + 8(1)(−1)]k̄ intersection of the sphere x2 + y2 + z 2 = a2
and the plane x + z = a.
= 8ī − j̄ − 10k̄
Solution The intersection of sphere
The unit normal to the surface is
x2 + y2 + z 2 = a2 (1)
∇f
f (x, y, z) = x log z − y is
2
and the plane x + z = a (2)
|∇f |
∂f ∂f ∂f is the circle in the plane x + z = a whose diameter
∇f = ī + j̄ + k̄
∂x ∂y ∂z is AB where√ A = (a, 0, 0), and B = (0, 0, a).√
x AB = a 2 and radius of the circle r = a/ 2
= log z ī + (−2y)j̄ + k̄
z
Let F̄d r̄ = ydx + zdy + xdz
Q = (−1, 2, −1)
⇒ F̄ = yī + z j̄ + xk̄
(∇φ)Q = log(−1)ī + (−2)2j̄ + k̄ ⇒ n̄ i j k
∂ ∂ ∂
∇f −4j̄ + k̄ curl F̄ =
= = √ . ∂x ∂y ∂z
|∇f | Q 17 y z x
Directional derivative of φ in the direction of n̄ is = −(ī + j̄ + k̄)
∇f −4j̄ + k̄ The unit normal to the surface is

∇φ · = (8ī − j̄ − 10k̄) · √
|∇f | 17 ∇S
4 − 10 −6 n̄ = where S = x + z − a = 0
= √ =√ . |∇S|
17 17 ∇S ī + k̄
⇒ n̄ = = √
8. (a) Apply Green’s theorem to evaluate |∇S| 2
C
C
(2xy − x2 )dx + (x2 + y2 )dy where C is Hence
bounded by y = x2 and y2 = x.
F̄ · d r̄ = curl F̄ · n̄ds
Solution C consists of two parabolas y = x2 C S
and y2 = x intersecting O(0, 0) and P(1, 1). (by Stokes’s theorem)
Here M = 2xy − x2 and N = x2 + y2 ;
∂N ∂M ī + k̄
− = 2x − 2x = 0. =− (ī + j̄ + k̄) · √ ds
∂x ∂y S 2
By Green’s Theorem
1 1
=− √ +√ S
∂N ∂M 2 2
Mdx + Ndy = − dxdy √ πa2
C S ∂x ∂y = − 2S = − √
2
RHS is clearly
=0
Since the area of the circle is
LHS = (2xy − x2 )dx + (x2 + y2 )dy

1 √x
C
a 2 πa2
S = πr = π √
2
= .
= 0dxdy = 0. 2 2
x=0 y=x2
1. (a) Test the

√ convergence
√ of the series By Leibnitz’s test, the given series is convergent.
√ $ $ 1
1 2 3 n Consider the series |un | = Taking
+ + + ··· + 2 . n +1
2
2 3 8 n −1
$ $ 1
vn = we note that
√
Solution The nth term of the series n2
n 1 1 un n2
un = 2 , n ≥ z; u1 = . Let vn = 3/2 be the lim = → 1 = 0 as n → ∞.
n −1 z$ n n→∞ vn 1 + n12
nth term of comparison series vn which converges $ $
√ By comparison
$ test, both un and vn behave
un n n3/2 1 alike. Since vn is p-series with p = 2 > 1
= 2 · = → 1 = 0 as n → ∞.
vn n −1 1 1− n12 is convergent; consequently the given series is
$ $ absolutely convergent.
So both un and vn converge or diverge
$ 1 1. (c) State and prove Generalised Mean Value
together. But is convergent. Theorem.
n3/2
∴ The given series is convergent.
% ∞
cos nπ Solution Please refer to See 3.1.6 on p. 11 of
1. (b) Test whether the series is Ch. 3.
n=1
n2 + 1
absolutely convergent.
2. (a) If r 2 = x2 + y2 + z 2 and u = r m , prove that
Solution The given series is ∂2 u ∂2 v ∂2 w
+ 2 + 2 = m(m + 1)r m−2 .
%∞ ∂x2 ∂y ∂z
cos nπ 1 1 1
2 +1
=− + 2 − 2 + ···
n 2 2 + 1 3 +1 Solution
n=1 r 2 = x2 + y2 + z 2 (1)
(−1) n
+ 2 + ··· Differentiating partially w.r.t x, we get
n +1
% ∂r ∂r x
= (−1)n un (say), (un > 0) 2r = 2x ⇒ =
∂x ∂x r
The series is an alternating series, which observe ∂r y ∂r z
1 Similarly = , = (2)
that (u1 > u2 > u3 · · · ) and lim un = 2 = 0. ∂y r ∂z A
n→∞ n +1
Differentiating u = r m w.r.t x twice y becomes imaginary when x > 3a and x < −a.
∂u ∂r Thus the curve exists only between x = −a and x =
= mr m−1 , = mr m−2 x 3a, i.e., for −a < x < 3a.
∂x ∂x
Asymptotes:
∂2 u m−3 ∂r
= m m(m − 2)r · x + r m−2
· 1 Asymptotes parallel to the y-axis are obtained by
∂x2 ∂x equating to zero the coefficient of the highest power

x 2
of y. So x = −a is a vertical asymptote of the curve.
= mr m−2 (m − 2) 2 + 1
r Since the coefficient of x3 is constant there is no
r m−2 asymptote parallel to x-axis.
= m 2 (m − 2)x2 + r 2 Intersection with the axes:
r
Putting y = 0 in Eq. (1) we have x2 (3a − x) = 0 ⇒
∂2 u ∂2 u ∂2 u x = 0, x = 3a. The curve meets the x-axis at O and
Hence + 2 + 2
∂x2 ∂y ∂z A(3a, 0). Put x = 0 in Eq. (1) we get y = 0. ∴ The
mr m−2 % % curve meets at O.
= (m − 2) x 2
+ r2 Derivative:
r2
Differentiating w.r.t x Eq. (3) gives
mr m−2
= 2
(m − 2)r 2 + 3r 2 dy √ x √
r = a+x − √ + 3a − x
= mr m−2 [(m − 2) + 3] = m(m + 1)r m−2 . dx 2 3a − x

√ 1
2. (b) Find the radius of curvature of the curve −x 3a − x · √ /(a + x)
x2 y = a(x2 + y2 ) at (−2a, 2a). 2 a+x
6a2 − 2x2
=
Solution Please refer to the Solution of 2(b) in 2(a + x)3/2 (3a − x)1/2
JNTU Exam-Regular May/June 2006 (Set 3). 3a2 − x2
=
(a + x)3/2 (3a − x)1/2
3. (a) Trace the curve y2 (a + x) = x2 (3a − x).
dy
Solution The equations of the curve are At x = 3a, is not defined; so the curve has a
dx
vertical tangent at A(3a, 0). The shape of the curve
y2 (a + x) = x2 (3a − x) (1) is as shown in the adjoining figure.
Symmetry:
The curve is symmetric about the x-axis since Eq. (1) y=0
A symtote
contains even power of y.

Tangent
Origin:
Equation (1) is satisfied by x = 0, y = 0 and so the
curve passes through the origin.
Tangents at the origin: A (3a, 0) x
o
Equating to zero the lowest degree√ terms of Eq. (1)
we get ay2 = 3ax2 ⇒ y = ± 3x. Tangents at 0
are inclined at ±π/3 to the x-axis. Origin is a nodal
point of the curve since we have two real and distinct Figure 7
tangents at 0.
Region 3. (b) Find the length of the arc of the curve
Solving Eq. (1) for y, we get x = eθ sin θ, y = eθ cos θ from θ = 0 to π/2.

3a − x 1/2 Solution The equations of the curve are
y = ±x (2)
a+x x = eθ sin θ, y = eθ cos θ
Differentiating we get 1 2
which is linear sin z = . P = −x, Q = −ex /2 sin x
y
dx
= eθ (cos θ + sin θ) = eθ (sin tθ + cos θ) 2 /2
dθ I.F = e Pdx
= e−xdx = e−x .
dy
= eθ (− sin θ + cos θ) = eθ (cos θ − sin tθ) The complete solution is
dθ

Length of the arc from θ = 0 to π/2 1 −x2 /2 2 2
e = c − ex /2 sin x · e−x /2 = c + cos x
y
π/2 2 2
dx dy 1
+ dθ 2
or = (c + cos x)ex /2 .
0 dθ dθ y
π/2
= e2θ(sin θ+cos θ)2 +(−sin θ+cos θ)2 dθ 4. (c) Find the orthogonal trajectories of the family
0 of curves r = 2a(cos θ + sin θ).
√ π/2 θ √
= 2 e dθ = 2(eπ/2 − 1). Solution
0
r = 2a(cos θ + sin θ) (1)
4. (a) Form the differential equation of all circles
passing through the origin with their centres Differentiating the given equation w.r.t θ, we get
lying on the x-axis.
dr
= 2a(− sin θ + cos θ) (2)
Solution Equation of the family of circles pass- dθ
ing through the origin with their centres on the x-axis Dividing Eq. (2) by Eq. (1)
is
1 dr − sin θ + cos θ
x2 + y2 + 2gx = 0 (where g is a parameter ) (1) = ,
r dθ cos θ + sin θ
Differentiating Eq. (1) w.r.t x, we get which is the differential equation of the family of
dr dθ
2x + 2yy1 + 2g = 0 (2) given curves. On replacing by −r 2 , we get the
dθ dz
Eliminating ‘g’ between Eqs. (1) and (2), we get differential equation of the orthogonal trajectories as
the required differential equation as
1 2 dr − sin θ + cos θ
−r = ,
dy r dθ cos θ + sin θ
y2 − x2 = x . (3)
dx dr −(cos θ+sin θ)dθ
4. (b) Solve the differential equation Separating the variables =
r (−sin θ+cos θ)
dy 2 Integrating we get
+ yx = y2 ex /2 sin x.
dx
log r = − log(cos θ − sin θ) + log c
Solution The given equation is
⇒ r(cos θ − sin θ) = c.
dy 2
+ yx = y2 ex /2 sin x (1) 5. (a) Solve the differential equation
dx
(D2 − 2D − 3)y = x3 e−x .
which is Bernoulli’s equation with n = 2. Multiply-
ing Eq. (1) by y−2 we can write Eq. (1) as Solution Please refer to the Solution of 5(a) in
dy JNTU Exam-Regular May/June 2006S (Set 1).
2
y−2 + y−1 x = ex /2 sin x
dx 5. (b) Solve the differential equation
dz 2
or − zx = ex /2 sin x, z = y−1 d 2 y 1 dy 12 log x
dx 2
+ = .
dx x dx x2
Solution Please refer to the Solution of 5(b) in Equating the coefficients of s2

JNTU Exam-Regular May/June 2006S (Set 1).
31
A+C =1⇒A=1−C =1− (4)
6. (a) Find the Laplace transform of the rectified 34
semi-wave function defined by
the constants
sin wt, 0 < t < π/w
f (t) = . −5B + 9C = −4 ⇒ −5B = −4 − 9C
0, π/w < t < 2π/w
279 415
= −4 − =−
Solution f (t) is a periodic function with period 34 34
2π 83
. Therefore, ⇒B= (5)
w 34
2π/w
1 / ,
L{f (t)} = e−st f (t)dt s2 + 2s − 4
1 − exp(−2πs/w) 0 ∴L −1
π/w (s2 + 9)(s − 5)
1
= e−st f (t)dt 1 83 31
1 − exp(−2πs/w) 0 = 3 cos 3t + sin 3t + e5t .
2π/w 34 3 34
+ 0dt
π/w 6. (c) Change the order of integration and evaluate
−st 1 2−x
1 e
= xydxdy.
1 − exp(−2πs/w) s + w2 2 0 x2
π/w
(−s sin wt − w cos wt)]0 Solution Please refer to Ex. 8.13 on p. 8 Ch. 8.
−πs/w
1 e
= (0 + w) 7. (a) Find curl [r̄ f (r)] where r̄ = xī + yj̄ + z k̄,
1 − exp(−2πs/w) s2 + w2
r = |r̄|.
1
− 2 (0 − w)
s + w2 Solution Please refer to the Solution of 7(a) in
w [1 + exp(−πs/w)] JNTU Exam-Regular May/June 2006 (Set 4).
=
1 − exp(−2πs/w) s2 + w 2
w 1 7. (b) Find the work done in moving a particle in the
= · 2 . force field F̄ = 3x2 ī + j̄ + z k̄ along the st. line
1 − exp(−πs/w) s + w2
/ 2 , from (0,0,0) to (2,1,3).
−1 s + 2s − 4
6. (b) Find L .
(s2 + 9)(s − 5) Solution
Solution
F̄d r̄ = (3x2 ī + j̄ + 3k̄)(dxī + dyj̄ + dz k̄)
3 2
/ , x yz
−1 s2 + 2s − 4 = 3x2 dx + dy + zdz = d
L 2
(s2 + 9)(s − 5) (2,1,3) (2,1,3)
/ , / ,
−1 As + B −1 C Work done = ¯ =1
F̄ dr d(x3 yz 3 )
=L +L (1) 2
s2 + 9 s−5 (0,0,0) (0,0,0)
1 3
s2 + 2s − 4 31 = · 2 · 13 = 36.2
where C = 2 = (2) 2
(s + 9)(s − 5) s=5 34
and A and B are to be determined from 8. Verify Stokes’s theorem for F̄ = yī+z j̄ +xk̄ when
surface S is the part of the sphere x2 +y2 +z 2 = 1
(As + B)(s − 5) + c(s2 + 9) = s2 + 2s − 4 (3) above the xy-plane.
Solution We have to verify that The unit normal to S is

¯ = ∇φ 2xī + 2yj̄ + 2z k̄
F̄ dr curl F̄ · n̄ds (Stokes’ theorem) (1) n̄ = =
C S |∇φ| 4x2 + 4y2 + 4z 2
where S : φ(x, y, z) = x2 +y2 +z 2 = 1 is the spherical = xī + yj̄ + z k̄
surface above the xy-plane and C is the circle x2 + = sin θ cos φī + sin θ sin φj̄ + cos θ k̄
y2 = 1 in the xy-plane. is spherical polar coordinates
¯ = (yī + z j̄ + xk̄) · (dxī + dyj̄ + dz k̄)
F̄ dr
⇒ curl F̄ · n̄ = sin θ cos φ + sinθ sin φ + cos θ
= ydx + zdy + xdz

LHS = ¯ = ydx + zdy + xdz
F̄ dr RHS = curl F̄ds
S
C C π/2 2π
= ydx = [sin θ(cos φ + sin φ)
c θ=0 φ=0
+ cos θ sin θ]dθdφ

z = dz = 0 in the xy-plane C : x = cos θ, y = sin θ, π/2
dx = − sin θ dθ; 0 ≤ θ ≤ 2π =− [sin θ sin φ − sin θ cos φ
2π θ=0
= sin θ(− sin θ)dθ 0 sin θdθ

+ φ cos θ]2π
0 π/2

1 2π 1 = −2π cos θ sin θdθ
=− (1 − cos 2θ)dθ = − · 2π + 0 = −π 0
2 0 2 π/2
= −π sin 2θdθ
ī j̄ k̄ 0
∂ ∂ ∂ π/2
Also, curl F̄ = = −(ī + j̄ + k̄) cos 2θ
∂x ∂y ∂x =− = −π = LHS = RHS
2 0
y z x
Stokes’ theorem has been verified.
Code No. RO5010102 Set No.1
I year B. Tech. Regular Examinations, Apr/May, 2007
√
1. (a) Please refer to Question 1(a) and its Solution π 3 3 π 1
1. (c) Prove that + < sin−1 < + .
in JNTU-Regular Exam Apr/May 2007 (Set 2). 6 5 5 6 8
1. (b) Find whether the following series converges Solution f (x) = sin−1 x is continuous in [0,1]
1 1 1 and differentiable in (0,1)
absolutely/conditionally − · 1
6 6 3 f (x) = √
1·3·5 1·3·5·7 1 − x2
+ − + ··· .
6 · 8 · 10 6 · 8 · 10 · 12 By Lagrange’s mean value theorem, there exists c
in (0,1) such that 0 < c < 1 and
Solution We can write the given series as
f (b)−f (a) 1
1% 1 1 3·5 3·5·7 f (c) = =√ 0<a<c<b<1
(−1)nun = 1− + − +· · · b−a 1−c2
6 6 3 8·10 8·10·12
⇒ a2 < c2 < b2 ⇒ 1 − a2
ignoring the first two terms and suppressing the
1 3 · 5 · 7 · · · (2n + 1) > 1 − c2 > 1 − b2
factor . Let un = ; 1 1 1
6 8 · 10 · 12 · · · (2n + 6) ⇒√ <√ <√
1−a 2 1−c 2 1 − b2
un+1 2n + 3
= → 1(as n → ∞) Now we have
un 2n + 8
and D’Alembert’s ratio test fails. 1 sin−1 b − sin−1 a 1
√ < <√
1 − a2 b−a 1 − b2
un 2n + 8
Now n −1 =n −1 f (b) = sin−1 b, f (a) = sin−1 a
un+1 − 2n + 3
b−a b−a
=
5n ⇒√ < sin−1 b − sin−1 a < √
2n + 3 1 − a2 1 − b2
5 5 3/5 − 1/2 3 π 1
= → >1 ⇒ + < sin−1 − <
2+ n 3
2 1− 1 5 6 8
4
√
By Raabe’s test, the series converges, so that the π 3 3 π 1
⇒ + < sin−1 < +
given alternating series is absolutely convergent. 6 5 5 6 8
2. (a) Show that the functions u = x + y + z, shown in the figure.

v = x2 + y2 + z 2 − 2xy − 2yz − 2zx and Differentiating y w.r.t x, we get
w = x3 +y3 +z 3 −3xyz are functionally related.
dy 1 −2x
Solution y1 = = √ a2 − x2 + x · √
ux uy uz dx 2 2a a2 − x2
∂(u, v, w)
= vx vy vz a2 − 2x2
∂(x, y, z) wx wy wz = √ √
2 2a a2 − x2
1 1 1 (a2 − 2x2 )2
= 2(x − y − z) 2(y − x − z) 2(z − y − x) ⇒ 1 + y12 = 1 +
8a2 (a2 − x2 )
3(x2 − yz) 3(y2 − zx) 3(z 2 − xy)
4a − 12a x + 9a4
4 2 2
(3a2 − 2x2 )2
= = 2 2
Applying c1 − c2 , c2 − c 3 8a (a − x )
2 2 2 8a (a − x2 )
0 0 1 Length of the curve =2× length of one loop=4×

=6 2(x − y) 2(y − z) z−y−x length of loop in the first quadrant
(x − y)(x + y + z) (y − z)(x + y + z) z 2 − xy a+
1 1 =4 1 + y12 dx
= 6·2(x−y)(y−z) =0
(x + y + z) (x + y + z) 0
a 2
4 3a − 2x2
Hence functional relationship exists between u, v = √ √ dx
2 2a 0 a2 − x 2
and w. √ π/2 2
2 3a − 2a2 sin2 θ
= a cos θdθ
2. (b) Find the centre of curvature at the point a 0 a cos θ
a a √ √ √
, of the curve x + y = a. Find where we have put x = a sin θ, dx = a cos θdθ
4 4
also the equation of the circle of curvature at √ π/2
that point. = 2a (1 + cos 2θ) dθ
0
Solution Please refer to Solution of 2(b) in √ 1 π/2
√
= 2a(2θ + sin 2θ)0 = 2πa.
JNTU-Suppl. Exam Aug/Sep 2006S (Set 3). 2
3. (b) Find the volume of the solid generated by
3. (a) Find the length of the curve
revolving the lemniscale r 2 = a2 cos 2θ about
x2 (a2 − x2 ) = 8a2 y2 .
the line θ = π/2.
Solution The curve is symmetric about both the
Solution The given curve is symmetrical about
coordinates axes since its equation contains even
the pole. It has two loop which lie entirely within the
powers of x and y.
circle r = a. The upper half of the right loop is traced
Also origin lies on the curve. Equating the lowest
as θ varies from 0 to π/4.
degree term a2 (x2 − 8y2 ) to zero we obtain the
1 Here r = a(cos 2θ)1/2 (Please refer to Fig. 9.17 in
tangents at the origin to the curve y = √ x. This I Edition.)
2 2 Required volume=2× Volume of solid generated
shows that the origin is a node.
by right-half loop about θ = π/2
The curve meets the x-axis at x = 0, ±a, i.e., at π/4
2 3
O, A = (a, 0) and A = (−a,
√ 0). Solving the equa- V =2 πr cos θdθ
3
x a2 − x 2 0

tion for y we get, y = ± √ . The curve lies 4π π/4 3
2 2a = a (cos 2θ)3/2 cos θdθ
between −a ≤ x ≤ a. The shape of the curve is as 3 0
y which is a first order linear differential equation in x

dx
of the form + Px = Q
dy
/4

(–a, 0) A´ 0 A (a, 0) x 1 dy
−1 y
I.F = e Pdy
=e 1+y2 = etan

Figure 3 Lemniscate of Bernoulli x(I.F) = c + Q(I.F)dy
−1 y
−1 y tan−1 yetan
⇒ xetan =c+ dy
3 π/4
1 + y2
4πa 1
= (1 − 2 sin2 θ)3/2 cos θdθ Put tan−1 y = t dy = dt
3 0 1 + y2
1 1 −1 −1 y
Put sin θ = √ sin t ⇒ cos θdθ = √ cos tdt etan y tan
2 2 dy = tet dt = (t − 1)et
1 + y2
θ = 0 ⇒ t = 0 θ = π/4 ⇒ t = π/2 −1 y
= (tan−1 y − 1)etan
4πa3 π/2 1
= (1 − sin2 t)3/2 √ cos tdt ⇒ xetan
−1 y
= c + (tan−1 y − 1)etan
−1 y
3 0 2
−1 y
4πa3 π/2 ⇒ x = cetan + tan−1 y − 1.
= √ cos4 tdt
3 2 0
4πa3 3 1 π π2 a3 4. (c) The temperature of the body drops from
= · · · = √ . 100◦ C to 75◦ C in 10 min when the surround-
3 4 2 2 4 2
ing ar is at 20◦ C.
4. (a) Form the differential equation by eliminating Solution By Newton’s law of cooling, we have
the arbitrary constant from log y/x = cx.
dθ dθ
Solution The given relation is = −k(θ − θ0 ) or = −kdt (1)
dt θ − θ0
log y/x = cx or log y − log x = cx (1) where θ0 = 20◦ C, surrounding air temperature.
Differentiating Eq. (1) w.r.t x, we get Separating the variables and integrating
1 dy 1 log(θ − 20) = −kt + log c ⇒ θ − 20 = ce−kt (2)
− =c (2)
y dx x Initially at t = 0, we have θ = 100
dy y ⇒ c = 80 ⇒ θ − 20 = 80e−kt
x = y + y log or (3)
dx x
dy y When t = 10, θ = 75◦ ,
x = y 1 + log (3)
dx x 11
which is the required differential equation. 75 − 20 = 80e−10k ⇒ e−10k = (4)
16
4. (b) Solve the differential equation When t = 30,

(1 + y2 )dx = (tan−1 y − x)dy. 3
3 11
Solution The given equation can be written as θ − 20 = 80e−30k = 80 e−10k = 80
16

dy 1 tan−1 y 1331
+ x = (1) ⇒ θ = 20 + 80
dx 1 + y2 1 + y2 4096

When θ = 25◦ , 1 f (D) 1
by (xv) = x − V
f (D) f (D) f (D)
25 − 20 = 80e−kt The complete solution is
10t
5 −kt 11 ex 3
⇒ =e = y = yC.F + yP.I = c1 ex + c2 e−x + 2x −3x2 +3x
80 16 2
1
5 11 − (x sin x + cos x).
or log = 10t log 2
30 16
1 log 5 − log 80 5. (b) Solve the differential equation
⇒t= . (x2 D2 + xD + 4)y = log x cos(2 log x).
10 log 11 − log 6
5. (a) Solve the differential equation Solution The given equation is Euler–Cauchy
(D2 − 1)y = x sin x + x2 ex . Equidimensional equation. It is transformed to non-
homogeneous linear differential equation with con-
Solution The equation stant coefficients if we put
d d
(D2 − 1)y = x sin x + x2 ex (1) x = e or log x = z ⇒
z
=θ =x
dz dx
2
is a non-homogeneous linear differential equation. 2 d
x 2 = θ(θ − 1). The equation becomes
Its complete solution consists of dx
(a) complementary function yC.F and [ θ(θ − 1) + θ + 4 ]y = z cos 2z or (θ 2 + 4)y
(b) particular integral yP.I
= z cos 2z (1)
To find C.F
We have to solve the homogeneous equation C. F
A.E is m2 + 4 = 0 ⇒ m = ±2i
(D2 − 1)y = 0 (2)
∴ yC.F = c1 cos 2z + c2 sin 2z
Its A.E is m2 − 1 = 0 ⇒ m = ±1
= c1 cos(2 log x) + c2 sin(2 log x) (2)
∴ yc = c1 ex + c2 e−x (3) P. I
To find P.I 1 ze2iz
1 1 yP.I = z cos 2z = Re
We write 2 x 2 ex + 2 x sin x = I1 + I2 θ2 + 4 θ2 + 4
D −1 D −1
1
= Re e2iz · z , (exponential shift)
1 ex D −1 2 (θ+2i)2 +4
I1 = ex 2 x2 (exp. shift) = 1+ x
D + 2D 2D 2 = Re e2iz · 2
1
z
θ + 4iθ
ex D D2 D3 2
= 1− + − x
2D 2 4 8 e2iz θ −1
= Re 1+ z
ex x 3 x 2 x 1 ex 3 4iθ 4i
= − + − = 2x −3x2 +3x 2iz
2 3 2 2 4 2 e θ θ2
= Re 1− − + ··· z
We omit the last term as it merges with C.F 4iθ 4i 16
2iz 2
1 2D 1 e z z 1
I2 = 2 (x sin x) = x− 2 sin x = Re − −
D −1 D −1 D −1 2 4i 2 4i 16

2D sin x 1 D sin x −i cos 2z + sin 2z z 2 z 1
= x− 2 = − x sin x+ 2 = Re +i −
D − 1 −2 2 −1 − 1 4 2 4 16
2
1 z sin 2z z 1
= − (x sin x + cos x) = cos 2z + −
2 16 4 2 16

z cos 2z + 2z 2 sin 2z 1 ∂ ∂A3 ∂A2 ∂ ∂A1 ∂A3
= , − sin 2z = − + −
16 64 ∂x ∂y ∂z ∂y ∂z ∂x

is omitted as it merges in C.F ∂ ∂A1 ∂A3
+ −
1 ∂z ∂z ∂x
= log xcos(2 log x)+2(log x)2 sin(2 log x)
16 ∂2 A3 ∂2 A2 ∂2 A1 ∂2 A3 ∂2 A2 ∂2 A1
(3) = − + − + − = 0.
∂x∂y ∂x∂z ∂y∂z ∂y∂x ∂z∂x ∂z∂y
7. (b) Evaluate S Ā n̄ ds where Ā = z ī+xj̄ −3y2 z k̄
y = yC.F + yP.I from Eqs. (2) and (3). (4) and S is the surface of the cylinder x2+y2 = 16
included in the first octant between z = 0
/ , ∞
1 and z = 5.
6. (a) Prove that L f (t) = f¯ (s)ds where
t s
Solution Let φ = x2 + y2 − 16 = 0 be the curved
L {f (t)} = f¯ (s).
surface S. The unit normal to S is given by
Solution Please refer to Theorem 12.10 (Divi- ∇φ 2xī + 2yj̄
sion by t) in Ch. 12 on p. 17. n̄ = =
|∇φ| 4x2 + 4y2
6. (b) Find the Inverse Laplace transform of xī + yj̄
= x2 + y2 = 16
3(s2 − 2)2 4
.
2s5 Let R be the projection of S on the yz-plane
Solution
/ , z
−1 3(s2 − 2)2 Z=5
L A C
2s5
/ , B
3 −1 s4 − 4s2 + 4
= L
2 s5 o Q
/ , / , / , y
3 −1 1 −1 1 −1 1 P
= L − 4L 3
+ 4L Z=0
2 s s s5 x
3 1
= − 3t 2 + t 4 Figure 7
2 4

dydz
F̄ · n̄ds = F̄ · n̄
7. (a) For any vector Ā find div (curl Ā). S R |n̄ · ī|
xī + yj̄
Solution Let Ā = A1 ī + A2 j̄ + A3 k̄ F̄ · n̄ = (z ī + xj̄ − 3y2 z k̄) ·
4
xz + zy xī + yj̄ x
ī j̄ k̄ = and n̄ · i = ·i =
∂ ∂ ∂ 4 4 4
curl Ā = ∇ × Ā =
∂x ∂y ∂z Putting x = 0 in x2 + y2 = 16 we get y = 4
A1 A2 A3 (y = −4 is not in octant).

∂A3 ∂A2 ∂A1 ∂A3 ∴ y varies from 0 to 4, and z varies from 0 to 5
= − ī + − j̄
∂y ∂z ∂z ∂x dydz
∴ F̄ · n̄ds = F̄ · n̄
∂A2 ∂A1 |n̄ · ī|
+ − k̄ S S

∂x ∂y xz + xy 4
= dydz
div (curl Ā) = ∇ · (∇ × Ā) R 4 x
4 3 2π
= (y + z)dydz = [(−sin3 θ)(sin θ) + cos3 θ cos θ] dθ
y=0 z=0 0

4 4 4 5 2π
= ydy dz+ dy dz
y=0 z=0 y=0 z=0 = (cos4 θ + sin4 θ)dθ
4 5 π/2
0
π/2
y2 z2
= [z]50 + [y]40 =4 cos4 θdθ + 4 sin4 θdθ
2 0 2 0 0 0
= 40 + 50 = 90 3 1 √ 3 1 √
· ( π)2 · ( π)2 3
=4· 2 2 +42 2 = π
2 · 2 2
· 2 2
8. Verify Stokes’ theorem for F̄ = y3 ī + x3 j̄ in the
region x2 + y2 ≤ 1, z = 0. RHS = (∇ × F̄) · n̄ds = 3 (x2 + y2 )k̄ · n̄ds
S S
Solution F̄ = −y ī + x j̄. 3 3
j̄ ī k̄
The boundary C of the surface S is the circle in ∂ ∂ ∂
∵ ∇ × F̄ = = (3x2 + 2y2 )k̄
the xy-plane: x2 + y2 = 1, z = 0. ∂x ∂y ∂z
Putting x = cos θ, y = − sin θdθ, dy = cos θdθ −y3 x3 0
and 0 ≤ θ ≤ 2π. We have 2π 1
3π
=3 r 2 r dr dφ =
φ=0 r=0 2
F̄d r̄ = curl F̄ · nds (Stokes’ theorem) (1)
c S
where we have put x = r cos φ, y = r sin φ and
0 ≤ φ ≤ 2π. Also, dx dy = rdrdφ. Since LHS =
LHS = F̄d r̄ = −y3 dy + x3 dy RHS in Eq. (1) Stokes’ theorem has been verified.
c c
1. (a) Testthe convergence

of the following series Solution Consider f (x) = tan−1 x defined in
$ n2 1 [a, b] and 0 < c < b < 1.
+ 2 .
2n n Then f (x) is continuous in [a, b] and derivable in
(a, b) and we can apply Lagrange’s Mean Value The-
$ $ n2 orem. By the theorem there exists a point c in (a, b)
Solution Let un = . We know that
2n such that
$ 1
is p-series, with p = 2 > 1 which is convergent. f (b) − f (a)
n2 f (c) = (1)
n2 (n + 1)2 b−a
Consider vn = n ; vn+1 =
2 2n+1 1 1
vn+1 (n + 1) 22 n Since f (x) = ⇒ f (c) = there
= · 2 1+x 2 1 + c2
vn 2n+1 n exists c in (a, b) such that
−1
1 1 2 1 1 tan b − tan−1 a
= 1+ → as n → ∞ = (2)
2 n 2 1 + c2 b−a
$
By D’Alembert’s Ratio test, vn is also Since 1 + an < 1 + c2 < 1 + an we have
convergent. Since both the series are positive
−1
term series which are convergent, their sum 1 tan b − tan−1 a 1
$ $ n2 1 > > (3)
un = + 2 is also convergent. (1 + a )
2 b−a (1 + b2 )
2n n by Eqs. (1) and (2)
(b−a)
1. (b) If a < b prove that < (tan−1 b−tan−1 a)
(1+b2 ) Hence the first result of the question
b−a 4
< using Lagrange’s mean value (i) Taking b = , a = 1, we get
1 + a2 3
theorem. Deduce the following:

4
−1 −1 4 −1
4
−1
π 3 −1 4 π 1 3
< tan − tan 1 < 3
(i) + < tan < + 1 + 16/9 3 1 + 12
4 25 3 4 6
5π + 4 π+2 3 π 4 π 1
(ii) < (tan−1 2) < . ⇒ + < tan−1 < +
20 4 25 4 3 4 6
(ii) Taking b = 1 and a = 1, we get [∵ r = a(1 + cos θ = 2a cos2 θ/2)].

2−1 2−1
< tan−1 2 − tan−1 1 = 3. (a) Find the volume of the solid of revolution of
1 + 22 1 + 12
x3
1 π 1 π y2 = about its asymptote.
⇒ + < tan−1 z < + 2a − x
5 4 2 4
4 + 5π −1 2+π Solution Please refer to Ex. 6.5 for the shape of
or < tan 2 < . the curve cissoid of Diocles on p.6 Ch. 6.
20 4
Let P(x, y) be any point on the curve. If PN is
2. (a) If u = x2 − y2 , v = 2xy where
∂(u, v) perpendicular on the asymptote x = 2a, then PN =
x = r cos θ, y = r sin θ show that = 4r 3 . 2a − x.
∂(r, θ)
x3/2
Solution ∴y= √ (1)
u = x2 − y2 = r 2 (cos2 θ − sin2 θ) = r 2 cos 2θ 2a − x
√
v = 2xy2 = 2r 2 cos θ sin θ) = r 2 sin 2θ dy (2a−x) · 32 x−x3/2 · (−1) (2a−x)−1/2
∴ = 2
∂(u, v) dx (2a − x)
=
ur uθ √
∂(r, θ) vr vθ x(3a − x)
= (2)
(2a − x)3/2
2r cos 2θ −2r 2 sin 2θ
=
2r sin 2θ 2r 2 cos 2θ Volume of solid of revolution=2 Volume generated
= 4r (cos 2θ + sin 2θ)
3 2 by the upper half of the curve
2a
= 4r 3 · 1 = 4r 3 .
V =2 π(2a − x)2 dy
2. (b) For the cardioid r = a(1 + cos θ) prove that x=0
2a
ρ2 /r is constant where ρ is the radius of curvature. x1/2 (3a − x)
= 2π (2a − x)2 · dx
x=0 (2a − x)3/2
Solution Calculating the derivative 2a
= 2π (3a − x) 2ax − x2 dx
r1 = −a sin θ, r2 = −a cos θ (1) x=0
⇒r + 2
r12 = a (1 + cos θ) + (−a sin θ)
2 2 2
Putting x = 2a sin2 θ ⇒ dx = 4a sin θ cos θdθ;
= a [1 + 2 cos θ + cos θ + sin θ]
2 2 2
Limits for θ = 0 to π/2
= a2 (2 + 2 cos θ) = 4a2 cos2 θ/2 (2) π/2
= 16πa 3
(3 sin2 θ cos2 θ − 2 sin4 θ cos2 θ)dθ
and r 2 + 2r12 − rr2 = a2 (1 + cos θ)2 + 2a2 sin2 θ 0
3π π
+ a2 (1 + cos θ) cos θ = 16πa 3
− = 2π2 a3 .
16 16
= a2 [1 + 2 cos θ + cos2 θ + 2 sin2 θ
3. (b) Find the area of the curve r = a(1 + cos θ).
+ cos θ + cos2 θ]
= 3a2 (1 + cos θ) = 6a2 cos2 θ/2 (3) Solution Please refer to Ex. 8.23 on p. 15
(r + r12 )3/2
2 of Ch. 8.
ρ=
r 2 + 2r12 − rr1
(4a cos θ/2)3/2
2 2
4a θ the arbitrary constants from y = (a+x)/(x2+1).
= = cos
6a cos θ/2
2 2 3 2
2 Solution The equation can be written as
ρ2 16a
cos2 θ/2 8a
= 9 = (constant)
r 2a cos θ/2
2 9 (x2 + 1)y = a + x (1)
Differentiating Eq. (1) w.r.t x, we get is a non-homogeneous linear differential equation

with constant coefficients. Its complete solution con-
(x2 + 1)y1 + 2xy = 1 (2) sists of two parts
(a) the complementary function, yC.F
which is the required differential equation. (b) The particular integral yP.I
To find C.F
4. (b) Solve the differential equation We write the
dy
(1 − x2 ) + xy = y3 (sin7 x). A.E is m3 − 7m2 + 14m − 8 = 0
dx
⇒ (m − 1) (m − 2) (m − 4)
Solution The equation may be put in the form ⇒ m = 1, 2, 4
dy x y3
− y = sin−1 x (3) yC.F = c1 ex + c2 e2x + c3 e4x (2)
dx 1 − x2 1 − x2
To find P.I, we write
which is Bernoulli’s equation with x = 3. Multiply-
ex cos 2x
ing by−2y−3 the equation can be written as yP.I =
D3 − 4D2 − D + 4
dy−2 2x −2 sin−1 x =
ex cos 2x
− y = (4) (D − 1)(D − 2)(D − 4)
dx 1 − x2 1 − x2
cos 2x
which is a linear equation y−2 with = ex (exponential shift)
D(D − 1)(D − 3)
−2x sin−1 x cos 2x
P= , Q = ; = ex 3
1 − x2 1 − x2 D − 4D2 + 3D
2 cos 2x
I .F = e Pdx
= elog(1−x ) = 1 − x2 = ex replacing D2 by − 4
−4D−4(−4) + 3D
The general solution of the equation is 16 + D
= ex 2 cos 2x
16 − D2
x
−2 e
y (1 − x ) = c + Q (I.F) dx
2
= (16 cos 2x − 2 sin 2x)
260
= c − 2 sin−1 xdx The complete solution is
y = yC.F + yP.I
= c − 2[x sin−1 x + 1 − x2 ].
ex
= c1 ex +c2 e2x +c3 e4x + (16 cos 2x−2 sin 2x).
4. (c) Prove that the family of confocal conics 260
x2 y2 5. (b) Solve the differential equation
+ 2 = 1 are self-orthogonal.
(a + λ) (b + λ)
2 1
(x2 D2 − 3xD + 1)y = [log x · sin(log x)+ 1].
x
Solution Please refer to Ex. 6.75 on pp. 34–35.
Solution The given differential equation is an
Euler–Cauchy Equidimensional equation.
5. (a) Solve the differential equation d d
(D3 − 7D2 + 14D − 8)y = ex cos 2x Put x = ez ⇒ log x = z, x = θ,
dx dz
2
2 d d2
Solution The equation x 2 = 2 = θ(θ−1) the equation transforms to
dx dz
(D3 − 7D2 + 14D − 8)y = ex cos 2x (1) (θ 2 − 4θ + 1) = ze−z sin x + e−z (1)

which is a non-homogeneous linear equation in y e−z 1
= (5sin z+6cos z)z+ (382cos z+54sin z)
with constant coefficients. Its complete solution con- 61 61
/
sists of 1
(a) complementary function, yC.F and = [5 sin(log x) + 6 cos(log x)] (log x)
61x
(b) particular integral, yP.I ,
1
To find C.F + [382 cos(log x) + 54 sin(log x)] .
We have to solve (θ 2 − 4θ + 1)y = 0 61
A.E. is m2 − 4m + √1 = 0 d 2x
⇒m=2± 3 6. (a) Solve the differential equation + 9x
dt 2
√ √ √ √ = sin t, using Laplace transforms given that
yC.F = e2z (c1 e 3z
+ c2 e − 3z
) = x2 (c1 x 3
+ c2 x − 3 ) x(0) = 1, x(π/2) = 1.
To find P.I, we write Solution Let us assume that x (0) = a. Tak-
1 1 ing Laplace transforms of both sides of x + 9x =
yP.I = e−z + 2 e−z (z sin z) sin t, with x(0) = 1, x (0) = a and x(π/2) = 1.
θ 2 − 4θ + 1 θ − 4θ + 1
/ 1
1 1 We obtain [s2 x̄(s) − sx(0) + x (0)] + x̄(s) =
= + [5 sin(log x) + 6 cos(log x)] log x s2 + 1
6x 61x
, 1
1 ⇒ (s2 + 9)x̄(s) = s + a +
+ [382 cos(log x) + 54 sin(log x)] s2 +1
61
s+a 1
⇒ x̄(s) = +
1 1 s2 + 9 (s2 + 1)(s2 + 9)
∵ e−z = e−z
θ 2 − 4θ + 1 (−1)2 − 4(−1) + 1 s a 1 1 1
⇒ x̄(s) = 2 + 2 + −
e−z 1 s + 9 s + 9 8 s2 + 1 s2 + 9
= =
6 6x Taking inverse transforms on both sides
/ , / ,
s 1 1
1
z e−z sin z x(t) = L−1 2 + a L−1 2 +
θ 2 − 4θ + 1 s +9 s +9 8
/ , / ,
1 1 1 1
= e−z z sin z, L−1 2 − L−1 2
(θ − − 4(θ − 1) + 1
1)2 s +1 8 s +9
(exponential shift) a 1 1
= cos 3t + sin 3t + sin t − sin 3t
1 3 8 24
= e−z Im 2 zeiz x(π/2) = 1
θ − 6θ + 6
1 3π 1 1 3π 1 π
= e−z Imeiz z ⇒ 1 = cos + a sin + sin
(θ + i) − 6(θ + i) + 6
2 2 3 8 2 8 2

(exponential shift) 1 1 1 5
iz ⇒0− a− + ⇒a=−
e 1 3 3 8 2
−z
= e Im z 1
5 − 6i 1 − 5−6i θ + 5−6i
6−2i 1
θ2 x(t) = cos 3t + (sin t − 7 sin 3t)
42 + 26i 8
= e−z Imeiz (5 + 6i)(1 + θ + · · · )z 6. (b) Change the order of integration and evaluate
61 1 2−x
e−z xydydx.
= Im [(5 cos z − 6 sin z)
61 0 x2

42 26 Please refer to Ex. 8.13 on p. 8 Ch. 8.
+i(5 sin z + 6 cos z)] z + + i Solution
61 61
7. (a) Prove that ∇ × (∇ × Ā) = −∇ 2 Ā + ∇(∇ · Ā). Along OB x = 1, z = 0, dx = dz = 0;
Solution Pleas refer to proof of Identity 9 (Art y changes from y = 0 to y = 1

10.4) p. 17 Ch. 10. (2xy2 z + x2 y)dxī + dyj̄ + 0̄
AB
1
7. (b) If φ = 2xy2 z + x2 y evaluate C φdt where C 1
y2 1
consists of st. lines from (0, 0, 0) to (1, 0, 0) = ydyj̄ = j= j̄
0 2 0 2
and then to (1, 1, 0) and then to (1, 1, 1).
Along BC x = 1, y = 1 dx = dy = 0;
Solution z changes from z = 0 to z = 1

φ(x, y, z) = 2xy2 z + x2 y (2xy2 z + x2 y)dxī + dyj̄ + dz k̄
BC
1
r̄ = xī + yj̄ + z k̄ ⇒ d r̄ = dxī + dyj̄ + dz k̄
= (2z + 1)k̄dz = k̄(z 2 + z)10 = 2k̄
∴ φd r̄ = (2xy2 + x2 y)dxī + dyj̄ + dz k̄ z=0
C
C φd r̄ = (2xy2 z + x2 y)dxī + dyj̄ + dz k̄
= + + C C
1
OA AB BC
= 0̄ + j̄ + 2k̄.
2
where O = (0, 0, 0), A = (1, 0, 0),
B = (1, 1, 0) and C = (1, 1, 1) 8. Verify Green’s theorem for c (y − sin x)dx+
cos xdy where c is the triangle formed by the
Along OA y = z = 0, dy = dz = 0; points (0, 0), (π/2, 0) and (π/2, 1)

x varies from x = 0 to x = 1 0 · dx = 0̄; Solution Please refer to Ex. 11.35 on p. 23 of
OA
Ch. 11.
1. (a) (b) (c) Please refer to the Question 1(a), Substituting these value in Eq. (3), we have
(b), (c) and its Solution in JNTU-Regular Exam u
May/June 2006 (Set 4). x + y + z = −(m + n + p) = a or λ
λ
u
= −(m + n + p) (8)
2. (a) Given that x + y + z = a find maximum value a
of xm yn z p . Substituting from Eq. (8) into Eqs. (5), (6), (7)
mu am
Solution Let x=− = ,
λ m+n+p
u(x, y, z) = xm yn z p , (1) an
y= ,
φ(x, y, z) = x + y + z − a = 0 (2) m+n+p
ap
z= (9)
Consider the Lagrangian function m+n+p
Now the maximum value is
F(x, y, z) = u(x, y, z) + λφ(x, y, z) m n p
am an ap
= xm yn z p + λ(x + y + z − a) (3) xm yn z p =
m+n+p m+n+p m+n+p
Differentiating Eq. (3) partially w.r.t x, y, z, we am+n+p mm nn pp
= (10)
obtain m + n + pm+n+p
,
Fx = mxm−1 yn z p +λ, Fy = nxm yn−1 z p +λ, 2. (b) Find the envelope of the family of circles
(4)
Fz = pxm yn z p−1 +λ, through the origin and whose centres lie on
x2 y2
For an extreme value (maximum or minimum) the circle 2 + 2 = 1.
each if Eq. (4) is zero a b
mu Solution Let P = (a cos t, b sin t) be any point

⇒x=− , (5) on the given ellipse. The equation of a circle through
λ
nu O and P as centre is
y=− , (6)
λ (x − a cos t)2 + (y − b sin t)2 = OP 2
pu
z=− (7) = a2 cos2 t + b2 sin2 t (1)
λ
Since t is a parameter we get a family of circles Eliminating a from Eqs. (1) and (2), we get the
for different values of t. Differentiating Eq. (1) w.r.t differential equation as
’t,’ we get
2xy1 = y. (7)
2(x − a cos t)(a sin t) + 2(y − b sin t)(−b cos t)
dy
= 2a2 cos t(− sin t) + 2b2 sin t cos t (2) 4. (b) Solve the differential equation x2 = ey − x.
dx
by
⇒ ax sin t − by cos t = 0 ⇒ tan t = Solution The given equation can be written as
ax
by ax dy 1 −y 1 d 1 1
⇒ sin t = , cos t = (3) e−y + e = 2 or (e−y ) − e−y = − 2
h h dx x x dx x x
where h = a x + b y
2 2 2 2
1 1
which is linear in e−y ; with p = − and Q = − 2
x x
Substituting into Eq. (1) from Eq. (3)
1 1
2 I.F = e x dx = e− log x =
a2 x b2 y ax by x
x− + y− = a2 + b2
h h h h and the general solution is

x2 h2 − 2a2 + a4 + y2 h2 − 2b2 h + b4
1 −y 1 1 1
= a4 x2 + b4 y2 e =c− · dx = c + 2 .
x x2 x 2x
⇒ h2 x2 + y2 = 2h a2 x2 + b2 y2 (4)
4. (c) Find the orthogonal trajectories of the family
which is the required envelope. of curves x2/3 + y2/3 = a2/3 .
Solution The equation of the given family of

3. (a) Trace the curve r = a(1 + cos θ).
curves
Solution Please refer to Ex. 6.17 on p. 15 x2/3 + y2/3 = a2/3 (1)
of Ch. 6.
Differentiating Eq. (1) w.r.t x, we obtain
3. (b) Find the length of the arc of the curve 2 −1/3 2 −1/3 y1/3
x = eθ sin θ, y = eθ cos θ from θ = 0 to x + y y1 = 0 or y1 = − 1/3 (2)
3 3 x
θ = π/2.
which is the differential equation of the family of
Solution Please refer to the Solution of 3(b) in given curves.
JNTU-Suppl. Exam Aug/Sep 2006 (Set 4). Thedifferential equation of the orthogonal
trajec-
1
tories replacing y1 in Eq. (2) by −
4. (a) Find the differential equation of all parabolas y1
having the x-axis as its axis and (a.θ) as its 1 y1/3
focus. − = − 1/3 or x1/3 dx = y1/3 dy (3)
y1 x
Solution Equation of the family of parabolas Integrating we obtain the required orthogonal
y = 4ax
2
(5) trajectories as
y4/3 − x4/3 = c4/3 . (4)
where a is the parameter. Differentiating w.r.t x we
obtain d 2y dy
5. (a) Solve the differential equation +4 +
2yy1 = 4a (6) dx2 dx
5y = −2 cos hx given that y(0) = 0, y (0) = 1.
d is a Legendre’s equation. We change the independent

Solution If we write = D the given
dx variable x to z by putting
differential equation is
(D2 + 4D + 5)y = −(ex + e−x ) (1) 2x − 1 = ez or z = log(2x − 1)
d d2
which is a non-homogeneous linear equation with ⇒ (2x − 1) = 2θ, (2x − 1)2 2 = 22 θ (θ − 1),
dx dx
constant coefficients. 3
d
Its complete solution consists of (2x − 1)3 3 = 23 θ(θ − 1)(θ − 2),
(a) C.F and (b) P.I. dx
To find C.F, we put y = emx in the homogeneous so that Eq. (1) becomes
equation (D2 + 4D + 5)y = 0 and get the
A.E m2 + 4m + 5 = 0 ⇒ m = −2 ± i 1 1 z
[8(θ 3 − 3θ 2 + 2θ) + 2θ − 2]y =+ e
2 2
yC.F = e−2x (c1 cos x + c2 sin x) (2)
1 ez
or (4θ 3 − 12θ 2 + 9θ − 1)y = + (2)
To find P.I, we write 4 4
1 1 which is a linear equation in y. Its complete solution
yP.I = ex − 2 e−x
D2+ 4D + 5 D + 4D + 5 consists of (a) C.F, and (b) P.I.
1 1 To find C.F, we consider the homogeneous equa-
or yP.I = − ex − e−x tion
10 2
replacing D by 1 and –1. (4θ 3 − 12θ 2 + 9θ − 1)y = 0 (3)
whose A.E is 4m − 12m + 9m − 1 = 0
3 2
y = yC.F + yP.I √
ex e−x ⇒ m = 1, 1 ± 3/2 (4)
= e−2x (c1 cos x + c2 sin x) − − (3)
10 2
y = e−2x (c2 − 2c1 ) cos x + (c − 1 − 2c2 ) 1 4−12 9 −1
4 −8 1
ex e−x 4 −8 1 0
sin x − − (4) √ √
10 2 8 ± 82 − 2 · 4 3
3 = =1±
y(0) = c1 + c2 · 0 = ; 2·4 2
5
1 1 √ √
y (0) = (c2 − 2c1 ) + 0 − + =1 ∴ yC.F = [c1 + (c2 e 3z/2
+ c3 e 3z/2
)]ez (5)
10 2
3 9
⇒ c1 = , c2 = where z = log(2x − 1).
5 5
1 1 1 1
3e−2x ex e−x yP.I = + ez
y= (cos x + 3 sin x) − − . (5) 4θ 3 −12θ 2 +9θ−1 4 4 4θ 3 −12θ 2 +9θ−1
5 10 2 1 z
= − − ez (6)
5. (b) Solve the differential equation 4 12
d 3y dy
(2x − 1)3 3 + (2x − 1) − 2y = x. The complete solution is
dx dx
Solution The given equation y = yC.F + yP.I substituted from Eqs. (5) and (6)
d 3y dy / ,
(2x − 1)3 + (2x − 1) − 2y = x (1) 1
dx3 dx 6. (a) Evaluate L et (cos 2t + sinh 2t) .
2
1 1
Solution We have x2 x3
/ , = − dy
1 0 2 2 y
L cos 2t + sinh 2t 1 2
2 1 1 y y3
= − − − dy
1 2 3 2 3
= L {cos 2t} + L−1 {sinh 2t} by linearity (1) 0
1
2 1 y3 y2 1
s2 1 2 = + − dy = .
= 2 + · = f¯ (s) (2) 0 6 2 2 12
s + 4 2 s2 − 4
7. (a) Evaluate ∇ 2 (log r) where r = x2 + y2 + z 2 .
By first shifting theorem,
/ , Solution Please refer to Ex. 10.19 (ii) on p. 442
1
L e (cos 2t + sin h2t) = f¯ (s − 1)
t of Engg. Maths-I (I Edition).
2
(s − 1)2 1 2
= 2 . ∇ 2 f (r) = f (r) + f (r) f (r) = log r
s − 2s + 5 s − 2s − 3
2 r
/ , 1 1
1 ⇒ f (r) = , f (r) = − 2
6. (b) Find L−1 2 . r r
s + 2s + 5 ∴ ∇ 2 (log r) = 0.
Solution 7. (b) Find constants a, b, c so that the vector
/ , / , Ā = (x + 2y + az)ī + (bx − 3y − z)j̄
1 1
L−1 = L−1 +(4x + 8y + 2z)k̄ is irrotational. Also find φ
s2 + 2s + 5 (s + 1)2 + 22
/ , such that Ā = ∇φ.
1 e−t
= e−t L−1 2 = sin 2t
s +2 2 2 Solution Please refer to Solution of 7(b) in
by first shifting theorem. JNTU-Suppl. Exam Aug/Sep 2006 (Set 2).
6. (c) Evaluate the triple integral 8. Verify Stokes’ theorem for the vector field
1 1 1−x F̄ = (2x − y)ī − yz 2 j̄ 2 − y2 z k̄ 2 over the upper
x dzdxdy. half surface of x2 + y2 + z 2 = 1 bounded by the
0 y 0
projection of the xy-plane.
Solution
Solution Please refer to the Solution for Qn. 8
1 1 1 of JNTU-Suppl. Exam Aug/Sep 2006 on Page 38 in
0 dxdy =
x[z]1−x [x(1 − x)dx]dy this section.
0 y 0
1.(a) Test
√ the convergence
√ of √
the series absolutely.
2−1 3−1 4−1
+ 2 + 2 + ··· . 1. (c) Verify Rolle’s theorem for
32 − 1 4 −1 5 −1
√ x2 + ab
f (x) = log in [a, b], (x = 0)
n+1−1 x(a + b)
Solution Here un = > 0 for all
(n + 2)2 − 1 √
n Solution
n. The series is of positive terms. Take vn = 2 = f (x) = log(x2 + ab) − log x − log(a + b)
n
1 $ 2x 1
3/2
vn is a p-series with p = 3/2 > 1 and is Also f (x) = 2 − ;
n x + ab x
convergent.
+ f (a) = log a(a + b) − log a − log(a + b) = 0
√ 1 + 1n − √1n
un n + 1 − 1 3/2 f (b) = log b(a + b) − log b − log(a + b) = 0
= · n =
vn (n + 2)2 − 1 (1 + 2n )2 − n12
So, f (x) is continuous in [a, b], derivable in (a, b)
→ 1 = 0 as n → ∞
and f (a) = f (b). Thus f (x) satisfies all the con-
$ ditions of Rolle’s theorem. As a consequence there
By comparison test, un is convergent.
must exists a point ‘c’ (a, b) such that
1. (b) Examine whether the following series is 2c 1 √
absolutely convergent or conditionally f (c) = 0 ⇒ − = 0 ⇒ c = ± ab
1 1 1 c2 + ab c
convergent 1 − + − + · · · . √
3! 5! 7! Thus, c = ab (a, b) exists and Rolle’s Theorem
1 1 1 is verified.
Solution Consider the series 1+ + + +· · ·
3! 5! 7!
2. (a) Show that the functions u = x + y + z, v =
un+1 1 x2 + y2 + z 2 − 2xy − 2yz − 2zx and w = x3
= → 0 < 1 as n → ∞
un 2n(2n + 1) +y3 +z 3 − 3xyz are functionally related.
By D’Alembert’s Ratio test, the series converges. Solution Please refer to the Solution of 2(a) in
Therefore, the given alternating series converges JNTU-Regular Exam Apr/May 2007 (Set 1).
2. (b) Find the centre of curvature at that point which is a non-homogeneous linear equation with
a a √ √ √
, of the curve x + y = a. Find constant coefficients. Its solution consists of
4 4 (a) C.F and (b) P.I.
also the equation of the circle of curvature at
To find C.F, we have to solve the homogeneous
that point.
equation
Solution Please refer to the Solution of 2(b) in (D3 + 4D)y = 0 (2)
JNTU-Suppl. Exam Aug/Sep 2006 (Set 3) pp.40–41
of this section. Putting y = emx we obtain A.E m3 + 4m = 0 ⇒
m = 0, ±2i
3. (a) In the evolute of the parabola y2 = 4ax show
that the length of the curve from its cusp yC.F = c1 + c2 cos 2x + c3 sin 2x (3)
x = 2a to the point where it√meets the To find P.I, We write
parabola y2 = 4ax is 2a (2 3 − 1).
1 − 12 cos 2x
Solution Please refer to the Solution of 3(a), yP.I = sin 2x =
D(D2 + 4) D2 + 4
3(b) May/June 2000(Regular) p. 307 this section.
1 x sin 2x x sin 2x
=− · =− (4)
3. (b) Find thelength of 2 4 8
the arc of the curve
ex − 1 The complete solution is
y = log x from x = 1 to 2.
e +1
y = yC.F + yP.I
Solution Please refer to Ex. 7.5 on p. 5 of Ch. 7. x
= c1 + c2 cos 2x + c3 sin 2x − sin 2x. (5)
8
the arbitrary constant log y/x = cx. 5. (b) Solve the differential equation:
d 2y dy
x2 2 − 2x − 4y = x4 .
Solution Please refer to the Solution of 4(a) in dx dx
JNTU-Regular Exam Apr/May 2007 (Set 1) p. 51 of
this section.
d 2y dy
4. (b) Solve the differential equation x2 2
− 2x − 4y = x4 (1)
dx dx
(1 + y2 )dx = (tan−1 y − x)dy.
is an Euler–Cauchy Equidimensional equation. We
Solution Please refer to the Solution of 4(b) in change the independent variable x to z by putting
JNTU-Regular Exam Apr/May 2007 (Set 1) p. 51 of d d2
x = ex or z = log x, x = θ, x2 2 = θ(θ − 1) and
this section. dx dx
Eq. (1) becomes
4. (c) Please refer to the Question 4(c) and its [θ(θ − 1) − 2θ − 4] = e4z
Solution in JNTU-Regular Exam Apr/May 2007
(Set 1) p. 51 of this section. ⇒ (θ 2 − 3θ − 4)y = e4z (2)
which is a non-homogeneous linear differential equa-

tion with the constants coefficients. Its complete
d 3y dy
+ 4 = sin 2x. solution consists of (a) C.F, (b) P.I.
dx3 dx To find C.F, we consider the homogeneous equa-
Solution The given equation is tion
(D3 + 4D)y = sin 2x (1) (θ 2 − 3θ − 4)y = 0 (3)

Putting y = emz we obtain A.E is Solution

m − 3m − 4 = 0 ⇒ m = −1, 4
2 π/4 a sin θ
rdrdθ
√
1 0 0 a2 − r 2
yC.F = c1 e−z + c2 e4z = c1 + c2 x 4 (4) π/4 a sin θ
x −2rdr
=2 √ dθ
To find P.I, we write 0 0 a2 − r 2

1 π/4 2 a sin θ
1 e4z · z x4 log x =− a − r2 dθ
yP.I = e4z = = (5) 2 0 0
θ2 − 3θ − 4 5 5 π/4
= (−1) (a cos θ − a)dθ
The complete solution is 0

π/4 π 1
1 x4 log x = a[θ − sin θ]0 = a −√ .
y = yC.F + yP.I = c1 + c2 x 4 + . (6) 4 2
x 5
7. (a) Provethat
6. (a) Find L{t 2 sin 2t}.
Ā × Ā (2 − x)Ā n(r̄ · Ār̄)
∇× n
= + .
2 r rn r n+2
Solution We have L{sin 2t} = . By mul-
s2 + 4
tiplication by t 2 Solution Differentiating r 2 = x2 + y2 + z 2
partially w.r.t x we obtain
d2 1
L t sin 2t = (−1) 2
2 2
∂r ∂r x
ds s+ 4 = 2x or
2r = (1)
/ , ∂x ∂x r
d −1
=2 ∂r y
ds (s2 + 4)2 · 2s Similarly = ,
/ , ∂y r
d s ∂r z ∂r y ∂r̄
= −4 = and = = ī etc.
ds (s2 + 4)2 ∂z r ∂y r ∂x
(s2 + 4)2 ·1 − 2s(s2 + 4)·2s
= (−4)· We have
(s2 + 4)2
Ā × r̄ % ∂ Ā × r̄
4(3s − 4)
2
∇× = ī ×
= 2 . rn ∂x rn
(s + 4)3
%
/ , ∂ r̄
s+3 = ī × Ā × (2)
6. (b) Find L−1 2 . ∂x r n
s − 10s + 29
∂ Ā × r̄ ∂ 1
Solution We have Now = Ā × r̄ (3)
/ , / , ∂x rn ∂x r n
s+3 (s − 5) + 8
L−1 = L−1 ∂ r̄
s2 − 10s + 29 (s − 5)2 + 22 ⇒ ī × A× n
/ , ∂x r
s+8
= e5t L−1 2 by shifting theorem ī × (Ā × ī) nx
s + 22 = − n+2 (ī × (Ā × r̄))
/ , / , r n r
s 2
= e5t L−1 2 + 4L −1 (ī · ī)Ā − (ī · A)ī nx
s + 22 s 2 + 22 = − n+2 [(ī · r̄)Ā − (ī · Ā)r̄]
rn r
= e5t (cos 2t + 4 sin 2t) . Ā − A1 ī nx
= − n+2 (xĀ − A1 r̄), (4)
π/4 a sin θ rn r
r dr dθ
6. (c) Evaluate √ . Ā = A1 ī + A2 j̄ + A3 k̄
0 0 a2 − r 2
%
∂ r̄ (2) Along the st. line AB where A = (1, 0, 0)
LHS = ī × Ā × n
∂x r B = (1, 1, 0)
2Ā n nr̄ Here x = 1, z = 0, dx = 0, dz = 0 and y varies
= n
− n+2 (r 2 Ā) + n+2 (A1 x + A2 y + A3 z) form 0 to 1
r
r r
2−n nr̄ 1
= Ā + n+2 (Ā · r̄) = RHS. (5) F̄d r̄ = (−6yz)dz = 0
rn r
AB y=0
F̄ = (x − 27)ī − 6yz j̄ + 8xz k̄ evaluate

2 2
7. (b) If (3) Along the st. line BC where B = (0, 1, 0)
C
F̄d r̄ from the point (0, 0, 0) to the point C = (1, 1, 1)
(1, 1, 1) along the straight line from (0, 0, 0) Here x = 1 = y, dx = dy = 0, z varies form 0
to (1, 0, 0) to (1, 1, 0) and (1, 1, 0) to (1, 1, 1). to 1
1 1
Solution F̄d r̄ = 8xz dz =
2
8z 2 dz
BC z=0 z=0
F̄ = (x2 − 27)ī − 6yz j̄ + 8xz 2 k̄ 3
1
8z 8
= =
r̄ = xī + yj̄ + z k̄ ⇒ d r̄ = dxī + dyj̄ + dz k̄ 3 0 3
F̄d r̄ = (x2 − 27)dx − 6yzdy + 8xz 2 dz Adding Eqs. (1) and (3)
(1) Along the st. line OA where O = (0, 0, 0)
88
A = (1, 0, 0) F̄d r̄ =
C 3
Here y = z = 0 and dy = dz = 0 and x varies
form 0 to 1 8. Verify Stokes’ theorem for F̄ = x2 ī − yj̄ + k̄
1 3 1 integrated around the square x = 0, y = 0,
x
F̄d r̄ = (x − 27)dx =
2
− 27x z = 0, x = 1, y = 1 and z = 1.
OA 0 3 0
1 80 Solution The question is ambiguous. No
= − 27 = −
3 3 solution is possible.
1. (a) Test the convergence

of the series Solution Please refer to Ex. 4.8 on p.132 of Engg.
% n2 1 Maths-I (I Edition).
+ 2 . [5]
2n n x
3. Trace the curve y = a cosh and find the volume
Solution Please refer to the Solution of 1(a) on p. a
got by rotating this curve about the x-axis
54 in JNTU-Regular Exam Apr/May 2007 (Set 2)(in between the ordinates x = ±a. [16]
this section).
Solution
1. (b) Find the interval of convergence of the series Symmetry:
% (−1)n (x + 2)
whose nth term is . [5] The curve is symmetrical about the y-axis since
(2n + 5) x
cosh is an even function of x.
Solution The given series is alternating. If un = a
x+2 Origin:
then un > un+1 for all n. ∴ 2n+1 > 2n Putting x = 0 we get y = a so that the curve does
2n + 5 not pass through the origin.
x+2 x+2 x+2
⇒ n > n+1 and lim un = lim n = 0. Special points:
2 + 5 2 +5 n→∞ n→∞ 2 + 5
The curve cuts the y-axis at (0, a). For no value of x,
By Leibnitz’s Test the series converges for all x, x
y = 0, i.e., cosh = 0 for any x. The curve does not
i.e., −∞ < x < ∞ absolutely. a
cut the x-axis.
1. (c) Please refer to Question 1(c) and its Solution Asymptotes:
in JNTU-Regular Exam Apr/May 2007 (Set 2). [6] As x → ±∞, y → ±∞; so, there are no asymptotes
to the curve parallel to the axes.
2. (a) If x = r sin θ cos φ, y = r sin θ sin φ and
∂(x, y, z) Region:
z = r cos θ, prove that = r 2 sin θ.[6] x
∂(r, θ, φ) cosh is an even function which increases with x so
a
Solution Please refer to the result in Ch. 3 y is non-negative and increases with |x|.
3.11(iii) on p. 104 of Engg. Maths-I (I Edition). Tangents:

dy x dy
2. (b) Find the radius of curvature at any point on = sinh and = 0. The tangent at
x dx a dx (0,a)
the curve y = C cosh . [10] (0, a) is horizontal.
c
y 4. (c) Find the orthogonal trajectories of the family:

r n sin nθ = bn . [6]
Solution Please refer to Ex. 7.78 on p. 270 of

Engg. Maths-I (I Edition).
5. (a) Solve the differential equation y" − 4y + 3y

= 4e3x ; y(0) = −1, y (0) = 3 [8]
A (0, a)
o x Solution The equation y" − 4y + 3y = 4e3x
is a non-homogeneous linear equation with constant
Figure 6 coefficients.
This curve is called a Catenary or a chainette and Its complete solution consists of (a) complemen-
its shape is as shown in the figure. tary function, yC.F and (b) Particular integral, yP.I .
Volume of revolution The complementary function is obtained by solv-
a a ing the homogeneous equation y" − 4y + 3y = 0.
x
V = πy2 dx = 2πa2 cosh2 dx The auxiliary equation is m2 − 4m + 3 = 0 ⇒
x=−a 0 a (m − 3)(m − 1) = 0 ⇒ m = 3, 1.
a
2x
= πa2 1 + cosh dx ∴ yC.F = c1 ex + c2 e3x
0 a
a
a 2x To find P. I.
= πa2 x + sinh
2 a 0 1
a yP.I = 4e3x
= πa a + sinh 2
2 D2 − 4D + 3
2 1
= πa3 (1 + sinh 1 cosh 1) . = e3x 4,
(D + 3 − 3)(D + 3 − 1)
4. (a) Form the differential equation by eliminating (exponential shift)

the arbitrary constant from e 3x
D −1
x2 = 1+ 4
sec y +sec x = c + [3] 2D 2
2
e3x D D2
= 1− + − ··· 4
Solution Differentiating w.r.t. x gives sec y tan y 2D 2 4
dy
+ sec x tan x = x which is the required differen- 1 1 D
dx e 3x
− + − · · · 4 = e3x (2x − 1)
tial equation. 2D 4 8
4. (b) Solve the differential equation (2y sin x+cos y)
dx = (x sin y + 2 cos x + tan y)dy [7] y = yC.F + yP.I = c1 ex + c2 e3x + 2xe3x ,
Solution Grouping the terms, we can write (−e3x ) merges with C.F.
d 2y
(cos ydx − x sin ydy) ± (2 cos xdy − 2y sin xdx) 5. (b) Solve the differential equation (1 + x)2
dx2
+ tan ydy dy
+(1 + x) + y = 4 cos log(1 + x). [8]
⇒ d(x cos y) = 2d(y cos x) − d(log cos x) dx
Integrating we get Solution This is an Euler–Cauchy equidimen-

sional equation. We solve it by changing the indepen-
x cos y − 2y cos x + log cos x = c. dent variable by putting 1+x = ez or z = log(1+x).
Therefore Solution
dy d 2y 2
(1 + x) = θy, (1 + x)2 2 = θ(θ − 1)y L{sin 2t} = ,
dx dx s2 + 4
2 2
and the given equation becomes Le−t sin 2t = = 2
(s + 1)2 + 4 s + 2s + 5
/ ,
d 2
[θ(θ − 1) + θ + 1]y = 4 cos z −t
L{te sin 2t} = −
ds (s2 + 2s + 5)
or (θ 2 + 1)y = 4 cos z
−1
= −2 2 (2s + 2)
This is a non-homogeneous linear equation with (s + 2s + 5)2
constant coefficients whose complete solution con- 4(s + 1)
= 2 .
sists of (a) complementary function, yC.F and (b) (s + 2s + 5)2
particular integral, yP.I . 6. (b) Using Laplace transform solve y" + 2y + 5y =
To find C.F, we have to solve (θ 2 + 1)y = 0. Its e−t sin t given that y(0) = 0, y (0) = 1.
auxiliary equation is m2 + 1 = 0 ⇒ m = ±i
Solution Taking Laplace transforms of both
∴ yC.F = c1 cos z + c2 sin z
sides
= c1 cos[log(1 + x)] + c2 sin[log(1 + x)]
[s2 ȳ(s) − sy(0) − y (0)] + 2[sȳ(s) − y(0)] + 5ȳ(s)
1
=
1 1 (s + 1)2 + 1
yP.I = 2 4 cos z = 4Re 2 eiz ⇒ s2 ȳ(s) − 1 + 2sȳ(s) + 5ȳ(s)
θ +1 θ +1
1 1 1
= 4Re · eiz = 4Reeiz 2 ⇒ (s2 + 2s + 5)ȳ(s) = 1 +
(θ + i)2 + 1 θ + 2iθ s2 + 2s + 2

iz 1 1 −1 s2 + 2s + 3
= 4Re e 1+ θ = 2
2iθ 2i s + 2s + 2

eiz 1 s2 + 2s + 3
= 4Re 1 − θ + ··· ⇒ ȳ(s) =
2iθ 2i (s2 + 2s + 2)(s2 + 2s + 5)

eiz 1 1 1/3 2/3
= 4Re − = 2 +
2i θ 2i s + 2s + 2 s2 + 2s + 5

cos z + i sin z 1 / , / ,
= 4Re z− 1 −1 1 2 −1 1
zi 2i ∴ y= L + L
3 (s+1)2 +1 3 (s+1)2 +22
Re (cos z + i sin z)(−2iz + 1) = cos z + 2z sin z / , / ,
1 −t −1 1 2 −t −1 1
= e L + e L
The complete solution of the equation is 3 s2 + 1 3 s 2 + 22
1
= cos[log(1 + x) + 2 log(1 + x) sin log(1 + x)] = e−t (sin t + sin 2t).
3
y = yC.F +yP.I = c1 cos[log(1+x)]+c2 sin[log(1+x)] 5 x2
+ 2 log(1 + x) sin log(1 + x), 6. (c) Evaluate x(x2 + y2 )dxdy.
0 0
the term cos log(1 + x) merges in the C.F. Solution

5 x2 5
x x 2
6. (a) Find the Laplace transformation of the func- (x3 + xy2 +)dydx = x3 y + y3 dx
tion te−t sin 2t. [5] x=0 y=0 x=0 3 0
5 6 5
1 7 x 1 x8 = 4 · 22 + 4 · 1 · 2 · 3 = 40
x + x dx =
5
+
3 6 3 8 0 ∂φ
0
= (8xy + 2x2 z)
5 6
1 ∂y A (1,2,3)
= + (5)8 . = 8 · 1 · 2 + 2 · 12 · 3 = 22
6 24
∂φ
7. (a) Prove that div (Ā × B̄) = B̄ · curl Ā − Ā · curl B̄. = (0 + 2x2 y)
∂z A (1,2,3)
=2·1 ·2=4
2
Solution Please refer the Bookwork and Art
10.4 in Ch. 10 on p. 14 of Ch. 10. ∂φ ∂φ ∂φ
∇φ = , ,
A
∂x ∂y ∂z A
7. (b) Find the directional derivative of the scalar = (40, 22, 4)
point function φ(x, y, z) = 4xy2 + 2x2 yz at
the point A(1, 2, 3) in the direction of the line Directional derivative at A = (1, 2, 3)
AB where B(5, 0, 4).
4 2 1
∇φ · ē = (40, 22, 4) · √ , − √ , √
Solution 21 21 21
AB¯ = (5, 0, 4) − (1, 2, 3) = (4, −2, 1); 40 × 4 − 22 × +4 × 1 120
= √ =√ .
√ 21 21
¯ = 42 + (−2)2 + 12 = 21
|AB|
8. Verify Stoke’s theorem for the vector field
¯
Unit vector along AB F̄ = (2x − y)ī − yz 2 j̄ − y2 z k̄ over the upper
half surface of x2 + y2 + z 2 = 1 bounded by the
¯
AB 4 2 1
ē = √ , −√ , √ projection of the xy-plane.
¯
|AB| 21 21 21
∂φ Solution Please refer to Ex. 11.42 on p. 29 of
= (4y + 4xyz) (1,2,3)
2
Ch. 11.
∂x A
1. (a), (b), (c) Please refer to Question 1 (a), (b), (c) We can find points in the nbd of (0, 0), u assumes
and its Solution in JNTU-Regular Exam May/June values greater than and less than f (0, 0), u assumes
2007 (Set 2) [5 + 5 + 6] . values greater than and less than f (0, 0)which show
2. (a) Locate the stationary points and examine that u(0, 0) is not an extreme value.
their nature of the following function:
u = x4 +y4 −2x2 +4xy −2y2 (x > 0, y > 0). x2 y2
2. (b) From any point of the ellipse =1
a2 b2
Solution perpendiculars are drawn to the coordinate
∂u axes. Prove that the envelope of the st. line
= 4x3 − 4x + 4y and (1) joining the feet of these perpendiculars is the
∂x
x 2/3 y 2/3
∂u curve + = 1.
= 4y3 + 4x − 4y (2) a b
∂y
∂u Solution If P(a cos θ, b sin θ) be any point on the
= 0 ⇒ x3 − x + y = 0, (3)
∂x ellipse
∂u x2 y2
=0⇒y +x−y =0
3
(4) + =1 (1)
∂y a2 b2
√ √ then M = (a cos θ, 0) and N = (0, b sin θ) are feet
√ √ we get (x, y) = (0, 0), ( 2,− 2),
Solving
of perpendiculars.
(− 2, 2)
y
∂2 u ∂2 u
Now l = = 12x2 − 4, m = = 4,
∂x 2 ∂x∂y
P (a cos Φ, bsin θ)
∂2 u N
n = 2 = 12y2 − 4
∂y
o M x
√ √
At ( 2, − 2), ln − m2 = 20 × 20 − 42 > 0 and
u has a minimum
√ √
At (− 2, 2), ln − m2 > 0 and u has a minimum.
At (0, 0), ln − m2 = 0 and we cannot say anything
Figure 2b
without further investigation.

Equation of MN is dy 1 x−a
Differentiating we get =± √ √
dx 2 a x
x 1 y 1
+ =1 (2) dy 1
a cos θ b sin θ = ± √ . The curve has a pair of tangents of
dx x=3a 3
Differentiating Eq. (2) w.r.t. ‘θ’ we get π
x = 3a which are inclined to the x-axis at ± .
6
x sin θ y cos θ Region:
· − · =0
a cos2 θ b sin2 θ The curve exists for x > 0. There is a loop between
x = 0 and x = 3a and the curve extends to ∞ in I
cos3 θ sin3 θ (cos3 θ)2/3 + (sin3 θ)2/3 and IV quadrants.
⇒ = = x 2/3 y 2/3
x/a y/b + Asymptotes:
a b
The co-efficient of highest power of x is constant. So,
1
= 2/3 2/3 there is no asymptote parallel to the x-axis. Similarly
x
a
+ by there is no asymptote parallel to the y-axis. There is
x/a no oblique asymptote as well.
⇒ cos3 θ = 2/3 y 2/3 , The shape of the curve is as shown in the adjoining
x
a
+ b
figure.
y/b
sin3 θ = 2/3 y 2/3 (3)
y=0
x
+ b
A symtote
Tangent
a
Substituting for cos θ and sin θ from Eq. (3) to Eq.

(1), we get
1/3
x (x/a)2/3 + (y/b)2/3 A (3a, 0) x
o
a (x/a)1/3
1/3
y (x/a)2/3 + (y/b)2/3
+ = 1 (4)
b (y/b)2/3
x 2/3 y 2/3 Figure 5
+ =1 3. (b) Find the surface area got by rotating one
a b
loop of the curve r 2 = a2 cos 2θ about the
which is the envelope of the st. line joining the feet
initial line.
of perpendiculars.
Solution Please refer to Ex. 6.33 on p. 213 of
3. (a) Trace the curve 9ay2 = x(x − 3a)2 . Engg. Maths-I (I Edition).
Solution
Symmetry:
the constant from x2 + y2 − 2ay = a2 .
Since the equation remains unchanged when y is
replaced by −y, the curve is symmetric about the Solution
x-axis. x2 + y2 − 2ay = a2 (1)
Origin:
y = 0 ⇒ x = 0, x = 3a. The curve passes through Differentiating w.r.t. x, we get
the origin. Also, it cuts the x-axis at x = 3a. x + yy1
2x + 2yy1 − 2ay1 = 0 ⇒ a = (2)
Tangents: y1
Equating to zero the lowest degree term we get x = 0.
Eliminating ‘a’ from Eqs. (1) and (2), we get the
Thus the y-axis is tangent to the curve at (0, 0).
required differential equation
1 √
9ay2 = x(x − 3a)2 ⇒ y = ± √ (x3/2 − 3a x) (x + yy1 )(x + 3yy1 ) = y12 (x2 + y2 ) (3)
3 a
4. (b) Solve the differential equation 5. (a) Solve the differential equation
dy 2 (D3 − 1)y = ex + sin3 x + 2.
(x + y3 + xy) = 1.
dx
Solution The equation is a non-homogenous lin-
Solution The question in wrong. ear differential equation with constant coefficients.
Its complete solution consists of
4. (c) If the air is maintained at 15◦ C and the (a) complementary function yC.F and
temperature of the body cools from 70◦ C (b) particular integral, yP.I .
to 40◦ C in 10 min. Find the temperature after To find C.F, we have to solve (D3 − 1)y = 0.
30 min. Its A.E. is m3 − 1 = 0 √
−1 ± i 3
Solution By Newton’s Law of Cooling whose roots are m = 1,
2
dθ √ √
= −k(θ − θ◦ ) (k > 0) (1) 3x 3x
dt x −x/2
∴ yC.F = c1 e +e c2 ·cos +c3 ·sin
2 2
where θ◦ = 15.
Separating the variables and integrating
dθ
= −k dt 1
θ − 15 yP.I = (ex + sin3 x + 2)
⇒ log(θ − 15) = −kt + C (2) D3 − 1
1 1 1
= 3 ex + 3 sin3 x + 3 2
when t = 0, θ = 70. D −1 D −1 D −1
= I1 + I2 + I3 (say)
From Eq. (2) we get C = log(70 − 15) = log 55
(3) 1 1 x x
I1 = 3 ex = e = ex
D −1 3D2 3
Putting in Eq. (2) 1 1
I2 = 3 (3 sin x − sin 3x)
D −14
log(θ − 15) = −kt + log 55
3 1 1 1
or kt = log 55 − log(θ − 15) (4) = sin x − sin 3x
4 −D − 1 4 −3 D − 1
2
3 D−1 1 9D − 1
When t = 10, θ = 40 so that we get from Eq. (4) − sin x + sin 3x
4 D2 − 1 36 81D2 − 1
10k = log 55 − log(40 − 15) 3 1
= (cos x−sin x)− (27cos 3x−sin 3x)
= log 55 − log 25 (5) 8 36×730
1
Dividing Eq. (4) by Eq. (5) I3 = 3 · 2 = −2
D −1
t log 55 − log(θ − 15)
= The complete solution is y = yC.F +yP.I .
10 log 55 − log 25
11 5. (b) Solve the differential equation:
⇒ log(θ − 15) = log 55 − 3 log
5 1
55 × 53 (x3 D3 + 2x2 D2 + 2)y = 10 x + .
= log for t = 30 x
113
6875 Solution This is Euler–Cauchy equidimensional
or θ = 15 + = 20.165 equation.
1331
We change the independent variable by putting x =
Temp. after 30 min = 20.165◦ C. ez or z = log x ⇒ xD = θ, x2 D2 = θ(θ − 1) and
d 1 1
x3 D3 = θ(θ − 1)(θ − 2) where θ = . sin t −
= cos 3t + sin 3t.
dz 8 24
The equation becomes 6. (b) Change the order of integration and evaluate
1 2−x
[θ(θ−1)(θ−2)+20(θ−1)+2]y = 10(ez +e−z )
x dydx.
⇒ (θ 3 − θ 2 + 2)y = 5 cosh z 0 x2
y
The A.E. of (θ 3 − θ 2 + 2)y = 0 is
m3 − m2 + 2 = 0 y = x2
(0, 2)
B
⇒ (m + 1)(m3 − 2m + 2) = 0 m = −1, 1 ± 2i
D2
yC.F = c1 e −z
+ e (c2 cos 2z + c3 sin 2z), z = log x
z M P (1, 1)
D1
P.I. is
0 A x
1 (2, 0)
yP.I = 10 (ez + e−z ) y=2–x
θ3 − θ2 + 2
ze−z ze−z Figure 6b
= 5ez + 10 · 2 = 5ez + 10
3θ − 2θ 5 Solution The region of integration is D above
−z 2 the parabola y = x2 and below the st. line y = 2 −
= 5e + 2ze = 5x + log x
z
x x between x = 0 and x = 14. D consists of two
The complete solution is domains
D1 and D2 : D = D1 ∪ D2
1 √
y = yC.F + yP.I = c1 · + x[c2 cos(2 log x) where D1 : x = 0 to x = y; y = 0 to 1
x D2 : x = 0 to x = 2 − y; y = 1 to 2.
2
+ c3 sin(2 log x)] + 5x + log x. 1 2−x 1 √y
x
2
∴ x dydx = x dxdy
d x x=0 y=x2 y=0 x=0
6. (a) Solve the differential equation 2 + 9x 2 2−y
dt
= sin t using Laplace transforms given that + x dxdy
x(0) = 1, x (0) = 0. y=1 x=0
1 √y 2 2−y
x2 x2
Solution Applying Laplace transforms on both dy + dy
sides, we have 0 2 0 1 2 0
1
1 1 1 2
[s2 x̄(s) − sx(0) − x (0)] + 9x̄(s) = 2 = ydy + (4 − 2y + y2 )dy
s +1 2 0 2 1
1 2
1 1 y2 1 1
⇒ (s2 + 9)x̄(s) = s + 2 = + 4y − y2 + y3
s +1 2 2 0 2 3
s 1 1 1

⇒ x̄(s) = 2 + · 1 1 1
s + 9 s2 + 1 s2 + 9 = + (4 · 1) − (22 − 1) + (23 − 1)
4 2 3
s 1 1 1 1
= 2 + · 2 − · 2 1 1 7 23
s +9 8 s +1 8 s +9 = + 4−3+ = .
4 2 3 12
Taking inverse Laplace transforms
/ , / , 7. (a) If φ1 = x2 y and φ2 = xz −y2 find×(φ1×φ2 ).
s 1 1
x(t) = L−1 2 + L−1 2
s +9 8 s +1 Solution
/ ,
1 −1 3 ∂ ∂ ∂
− L φ1 = ī + j̄ + k̄ (x2 y) = ī2xy + j̄x2 + 0
8×3 s2 + 9 ∂x ∂y ∂z

∂ ∂ ∂ z
φ2 = ī + j̄ + k̄ (xz − y2 )
∂x ∂y ∂z
= īz + j̄(−2y) + k̄x
o´ (0, 0, 8)B´
(φ1 × φ2 ) = īx3 + j̄(−2x2 y) + (¯ − 4xy2 )īj̄ k̄ A´
S2
× (φ1 × φ2 ) = ī(−8xy) + j̄ · 4y2 + k̄(−4xy)īj̄ k̄ S5

S3 S4
∂ ∂ ∂ 3
x −2x2 y−4xy2 = ī(−8xy)+j̄·y2 +k̄(−4xy). o
∂x ∂y ∂z S1 B (0, 1, 0) y
A
)
7. (b) If F̄ = (3x + 6y) ,0
ī − 14yz j̄ + 20xz k̄ evaluate
2 2
0 x2+y2 = 9
the line integral C F̄ · d r̄ from (0, 0, 0) to x (1,
(1, 1, 1) along x = t, y = t, z = t 3 . Figure 8
Solution The solid region is covered when x varies from 0

to 9 − y2 , y varies from 0 to 3 and z varies from 0
F̄d r̄ = (3x2 + 6y)dx − 14yzdy + 20xz 2 · dz to 8 (height of the cylinder)
= (3t 2 +6t)dt −14t ·t 3 ·dt +20·t ·t 6 ·3t 2 dt √ 8 3 9−y2
∵ x = y = tz = t 3 and RHS = 1dxdydz
z=0 y=0 x=0
dx = dy = dt dz = 3t dt 2
8 3
= (3t + 6t − 14t + 60t )dt
2 4 9
= 9 − y2 dydz
z=0 y=0
The limits for t are t=0 to 1 8 3
y 9 y
1 = 9 − y2 + sin−1 dz
z=0 2 2 3 0
F̄ · d r̄ = (60t 9 − 14t 4 + 3t 2 + 6t)dt
π 9 8
C

0
1 dz = 18π (2)
14 z=0 2 2
= 6t 10 − t 5 + t 3 + 3t 2 . 5
5 0
%
LHS = F̄ · n̄ = F̄ · n¯r dSr (3)
8. Verify divergence theorem for F̄ = 6z ī + (2x + y) S r=1 Sr
j̄ − xk̄ taken over the region bounded by the where Sr (r = 1, 2, . . . 5)

surface of the cylinder x2 + y2 = 9 included in
z = 0, z = 8; x = 0 and y = 0. are the five surfaces bounding the volume and n̄r are
the outward drawn unit normals to these surfaces.
Solution By Gauss’s divergence theorem
S1 : OAB, z = 0, n̄ = −k̄,

F̄ · n̄ds = · F̄dV (1) F̄ · n̄ = [0 + (2x + y)ī − xk̄] · (−k̄) = x
S C 3 √9−y2
where V is the volume bounded by a closed sur- F̄ · n̄dS = xdxdy
S1 y=0 x=0
face S and F̄ is a vector function with continuous √
derivatives. 3 9−y2
x2
= dy
F̄ = (F1 , F2 , F3 ) = (6z, 2x + y, −x) 0 2 0

∂F1 ∂F2 ∂F3 1 1 3 3
· F̄ = + + = 9y − y =9 (4)
∂x ∂y ∂z 2 3 0
=0+1−0=1 S2 : OA B , z = 8, n̄ = k̄,

A normal to cylinder x2 + y2 = 9 is
F̄ · n̄ = [48ī+(2x+y)j̄−xk̄]· k̄ = −x (x2 + y2 ) = 2xī + 2yj̄ unit normal to s is
3 √9−y2
2xī + 2yj̄ xī + yj̄
F̄ · n̄ds = − xdxdy n̄ = =
S2 y=0 x=0 (2x)2 + (2y)2 3
3
1 ∵ x2 + y2 = 9 on s
=− (9 − y2 )dy
2 0 xī + yj̄
3 F̄ · n̄ = (6z ī + (2x + y)j̄ − xk̄) ·
1 1 3
= − 9y − y3 = −9 (5)
2 3 1

0
= (6zx + 2xy + y2 )
S3 : AOO A y = 0, n̄ = −j̄, 3
xī + yj̄ y
F̄ · n̄ = [6z ī + (2x + 0)j̄ − xk̄] · (−j̄) = −2x n̄ · j̄ = · j̄ =
3 3

6zx + 2xy + y2
8 3 F̄ · n̄ds = dxdz
y
F̄ · n̄ds = (−2x)dxdz S5 R
S3 z=0

x=0 3 3
8 x
= −x 2 3
dz = −72 (6) = 3×64√ dx+8 (2x+ 9−x2 )dx
0 x=0 9−x2 x=0
z=0
3
S4 : OBB O , x = 0, n̄ = −ī, = 192 − 9 − x2 + 8 × 9
0

F̄ · n̄ = [6z ī + yj̄ + −0 · k̄] · (−ī) = −6z x 9 x 3
+8 9 − x2 + sin−1
2 2 3 0
3 8 9 π
F̄ · n̄ds = −6z dz dy = 192 × 3 + 8 × 9 + 8 × ×
S4 y=0 z=0
2 2
= 576 + 72 + 18π (8)
3 2 8
= −3z 0 dy = −576 (7)
y=0
%
S5 : ABB A (curved surface) F̄ · n¯r dsr = 9−9−72−576+576+72+18π
Sr
Project s on the xz-plane. Call it R. We have = 18π

dxdz ∴ The divergence theorem has been verified.
F̄ · n̄ds = F̄ n̄
S5 R |n̄ · j̄|
1. (a) Test for convergence of series 1

un+1 =
%∞
(n + 1)(1 − x)n+1
( n4 + 1 − n4 − 1). un+1 n(1 − x)n
1 lim = lim
n→∞ un n→∞ (n + 1)(1 − x)n+1
Solution Let un = n4 + 1 − n4 − 1, 1 1
$ = lim =
un > 0; un is a positive term series n→∞ (1 + 1 )(1 − x)
n
1−x
√ √
n4 + 1 + n4 − 1 By D’Alembert’s Ratio test, series converges
= ( n + 1 − n − 1) √
4 4 √
n4 + 1 + n4 − 1
2 1
=√ √ ⇔ < 1 ⇔ |1 − x| > 1
n + 1 + n4 − 1
4 1−x
⇔ The interval of convergence is (−∞, 0)∪(2, ∞).
$ $ 1
Take vn = as comparison series, which
n2
is convergent. π 1 3 π 1
1. (c) Prove that − √ > cos−1 > · · ·
3 5 3 5 3 8
un 2x2 using Lagrange’s Mean Value Theorem.
lim = lim √ √
n→∞ vn n→∞ n4 + 1 + n4 − 1
2 Solution Let f (x) = cos−1 x. Then f (x) is
= lim + + = 1 = 0 defined and continuous on [0, π] and derivable on
n→∞
1 + n4 + 1 − n14
1
1
(0, π). f (x) = − √ .
$ 1 − x2
By comparison test un is convergent. By Lagrange’s MVT, there exists a ‘c’ such that
for 0 < a < c < b < π
1. (b) Find the interval of convergence of the series
1 1 1 f (a) − f (b) −1
+ + + ··· . f (c) = =√
1 − x 2(1 − x) 2 3(1 − x)3 b−a 1 − c2
Solution Let
Now a < c 1 − c 2 > 1 − b2
n(1 − x)n

1 1 where λ is to be determined so that
⇒ −√ > −√
1 − a2 1 − c2
∂F ∂F ∂F
1 = 0, = 0, =0
> −√ ∂x ∂y ∂z
1 − b2
−1
1 cos b − cos−1 a 2x 2y
⇒ −√ > ⇒ 8yz + λ = 0, 8zx + λ = 0,
1 − a2 b−a a2 b2
1 2z
> −√ 8xy + 2 λ = 0 (3)(a)(b)(c)
1 − b2 c

b−a From Eq. (3)(a), (b), (c), we obtain
⇒ −√ > (cos−1 b − cos−1 a)
1 − a2

b−a a2 yz b2 zx c2 xy λ
> −√ ;b > a > 0 = = =− (4)
1 − b2 x y z 4

1 3 From the first equation, we get
Taking a = , b =
2 5
⎛ ⎞ a2 y b2 x x2 y2
= ⇒ 2 = 2 (5)
3
− 12 ⎟ x y a b
⎜ −1 3 −1 1
⇒ ⎝− + 5
⎠ > cos − cos
1 − ( 12 )2 5 2 y2 z2
Similarly, we obtain = 2
⎛ ⎞ b2 c
Substituting from Eqs. (5) and (6) we obtain from
⎜
3
−1 ⎟
> ⎝− + 5 2 ⎠ Eq. (1)
1 − ( 35 )2
a b c
π 1 3 π 1 x= √ ; y= √ ; z= √
⇒ − √ > cos−1 > − . 3 3 3
3 5 3 5 3 8
so, a possible extreme point is
2. (a) Find the volume of the largest rectangular

parallelepiped that can be inscribed in the a b c
ellipsoid of revolution 4x2 + 4y2 + 9z 2 = 36. p= √ ,√ ,√ .
3 3 3
Solution Let 2x, 2y, 2z be the length, breadth and We note that for fixed x and y, the volume function
height of the rectangular parallelepiped that can be V = 8xyz is increasing function of z, and if z = 0
inscribed in the ellipsoid the parallelepiped reduces to a two-dimensional
lamina with V = 0. a
x2 y2 z 2 As z increase, V also increases. But x = √ ,
+ + −1 = 0 (where a = b = 3, c = 2) (1) 3
a2 b2 c2
b
The centroid of the parallelepiped coincides with y = √ the maximum value of z satisfying
3 √
the centre O(0, 0, 0) of the ellipsoid and the corners
of the parallelepiped lie on the surface of Eq. (1). is z = c/ 3. Thus, V is maximum at
Eq. (1)
a b c θabc √
We have to find the maximum value of V = 8xyz, p = √ ,√ ,√ and max= √ = 16 3
the volume of the parallelepiped subject to condition 3 3 3 3 3
c.u. (a = b = 3, c = 2).
Eq. (1). Consider the Lagrangian function
2
x y2 z 2 2. (b) Find the envelope of the family of curves
F(x, y, z) = V + λ 2 + 2 + 2 −1 , (2) ax by
a b c + = a2 − b2 , where α is a
(V = 8xyz) cos α sin α
parameter. (-sign occurs here for +)
Solution Differentiating the given equation Pole:

pi
Putting r = 0, we get sin 2θ = 0 ⇒ θ = 0, , π, 3π.
ax by 2
+ = a2 − b2 (1) So, the curve passes through the pole. The lines θ =
sin 2α sin α
0, π/2 are tangents to the curve.
w.r.t. ‘α’ we get Limits:
The least value of sin 2θ is –1 and the greatest value is
sin α cos α 1. The corresponding values of r are −a and a. Hence
ax − by =0 (2)
cos2 α sin2 α the curve ties entirely within the circle of radius a.
(by)1/3 Asymptotes:
tan α = ··· The curve has no asymptotes as r does not tend to α
(ax)1/3
for any value of 0.
1 (ax)2/3 + (by)2/3
sec α = = ; Region or Extent:
cos α (ax)1/3 The values of r corresponding to some special values
1 (ax)2/3 + (by)2/3 of θ are as follows:
cosec α = = (3)
sin α (by)1/3 θ =0 π/4 π/2 3π/4 π 5π/4 3π/2 7π/4
r =0 a 0 −a 0 a 0 −a
when θ increases from 0 to π/4, r increases from 0 to
a and when θ increases from π/4 to π/2, r decreases
2
3
)
1
by
(by ) 3 form a to a.
+(
2
3
Thus one loop is formed between the rays θ = 0

x)
(a
and θ = π/2 (I quadrant) which is bisected by the

ray θ = π/4. Similarly three more loops exist in the
α
other three quadrants which are bisected by the rays
(ax )
1
3 θ = 3π/4, 5π/4 and 7π/4, respectively.
Figure 2b The curve is known as four-leafed rose. When θ
is replaced by (π/4 + 0) the equation becomes r =
Substituting these values in Eq.
(1) a cos 2θ, which represents the same curve, rotated
1 1 through π/4 (Please refer to Ex. 5.21 on pp. 177–178
(ax) · + (by) ·
(ax)1/3 (by)1/3 of Engg. Maths-I (I Edition)).
1/2
⇒ (ax)2/3 + (by)2/3 = (a2 − b2 ) 3π y
θ= π
3/2 4 θ=
(ax)1/3 + (by)2/3 = (a2 − b2 ) 2
θ=
π
4
⇒ (ax) 2/3
+ (by) 2/3
= (a − b )
2 2 2/3
x´ θ=π x
which is the required equation of the envelope. o
7π
θ=
2
3. (a) Trace the curve r = a sin 2θ. 5π 3π
θ= θ=
4 2
Solution y´
Symmetry: Figure 5
Changing θ to –0 we notice that n becomes negative, Author’s Note: The curve is in red lines: Print
and the equation remains the same. So, the curve is here Fig. 5.21 p. 170 Engg.Maths-I old Ed.
symmetrical about the initial line θ = 0. If we rotate
through π/2 the curve remains the same and so it is 3. (b) Find the whole length of the curve
symmetrical about θ = π/2 also. δa2 y2 = x2 (aA2 − x2 ).
Solution Please refer to Ex. 6.3 p193 of Engg. disappear.

Math I old Ed.
Solution Let m be the amount of radio-active
4. (a) Form the differential equation by eliminating substance present at time t and m0 the initial (original)
the arbitrary constants from y = a sec x amount at t = 0. By the decay rule
+b tan x. dm
= −km (k > 0) ⇒ m(t) = cekt (1)
dt
Solution The given equation can be written as
Initial condition:
a + b sin x − y cos x = 0 (1)
m0 − m(0) = ce0 = c ⇒ c = m0 (2)
Differentiating w.r.t. x, we have
From Eqs. (1) and (2), we get
0 + b cos x − y cos x + y sin x = 0 (2)
m(t) = m0 e−kt (3)
0 − b sin x−y cos x+2y sin x+y cos x = 0 (3)
From Eqs. (2) and (3), we obtain Amount of substance at t = 10 is

70 7 30
b = y − y tan x = −y cot x + 2y + y cot x m0 = m0 ⇒ m(10) = 1 − m0
100 10 100
⇒y − y tan x − y sec x cosec x = 0.
70
= m0 = 0.7m0 = m0 e−10k
4. (b) Solve the differential equation 100
(y4 + 2y)dx + (xy3 + 2y4 − 4x)dy = 0. 1 log 10 − log 7
⇒ k = − log 0.7 = (4)
10 10
Solution The equation can be written as
If t is the time required for 90% to disappear
dx y3 − 4 2y3
+ 4 x=− 3 90 m0
dy y + 2y y +2 1− m0 = = m0 e−kt ⇒ kt = log 10
100 10
y3 − 4 3y2 2
= 3 − where k is given by Eq. (4)
y + 2y
4 y +2 y
log 10 10 log 10
which is linear in x. ⇒t= = .
k log 10 − log 7
3
y −4
I.F = exp pdy = exp dy 5. (a) Solve the differential equation
y4 + 2y
(D2 + 4D + 4)y = 18 cosh x.
3y2 2
= exp − dy
y3 + 2 y Solution A.E is m2 + 4m + 4 = 0
3
y +2 y3 + 2
= exp log = ⇒ (m + 2)2 = 0 ⇒ m = −2, −2.
y2 y2
C.F
The complete solution is yC.F = (c1 + c2 x)e−2x ;
3
y +2 2y3 y3 + 2 To find P.I, we write
x = c − · dy
y2 y3 + 2 y2 1
= c − y2 yP.I = 9 (ex + e−x )
(D + 2)2
or x(y3 + 2) + y4 = cy2 . ex 1
=9 +9 · e−x ±ex +9e−x
(1+2) 2 (4+2)2
4. (c) If 30% of a radio-active substance disappears
in 10 days, how long will it take for 90% to The complete solution is y = yC.F + yP.I
5. (b) Solve the differential equation (D2 + 4)y = 1 1 1

= − · 1 + · t + · e−2t .
cos x. 4 2 t
Solution The given equation is a non-

6. (c) Evaluate r sin θdrdθ over the cardioid
homogeneous linear differential equation with con- r = a(1 − cos θ) above the initial line. [5]
stant coefficients. Its complete solution consists of
two parts: Solution The cardioid r = a(1 − cos θ) is sym-
(a) complementary function yC.F and metrical about the initial line θ = 0. It passes through
(b) particular integral, yP.I . the pole since θ = 0 ⇒ r = 0 The limits for r are
To find C. F, we have to solve r = 0 to r = a(1 − cos θ) and for θ, 0 to π.
(D2 + 4)y = 0 (1) y
θ=a(1−cos θ) θ= π
A.E. is m + 4 = 0 ⇒ m = ±2i
2
(2) 2
yC.F = c1 cos 2x + c2 sin 2x (3) θ =π θ=0
0 x
1 1 1
yP.I = 2 cos x = = cos x (4)
D +4 −1 + 4
2 3
The complete Solution is Figure 6c
1
y = yC.F + yP.I = c1 cos 2x + c2 sin 2x + cos x.
3 π a(1−cos t)
n
d ∴ sin θ dθ
6. (a) Show that L{t n f (t)} = (−1)n n {f¯ (s)}, θ=0
rdr
ds r=0
a(1−cos t)

n = 1, 2, 3. [5] π
r2
= sin θ dθ
Solution Please refer to Ch. 12 Laplace trans- 0 2 0

forms. a2 π
= (1 − cos θ)2 d(1 − cos θ)
/ , 2 0
1
6. (b) Evaluate L−1 . [6] a2 π a2 4a2
s2 (s + 2) = (1 − cos θ)3 0 = (23 − 0) = .
6 6 3
Solution Let
7. (a) Evaluate ∇ log r where r = x2 + y2 + z 2
1 A B C 1
= + 2+ = 2 [8]
s2 (s + 2) s s s+2 s (s + 2)
Solution Please refer to Solution of 7(b) in
⇒ As(s + 2) + B(s + 2) + Cs2 = 1 JNTU-Regular Exam Apr/May 2007 (Set 3). Also
1 refer Ex. 10.19 (ii) on p. 442 of Engg.Maths-I (I
s = 0 ⇒ 2B = 1 ⇒ B =
2 Edition).
1
s = −2 ⇒ C(−2)2 = 1 ⇒ C =
4 7. (b) Please refer to Question 7(b) and its Solution
1 in JNTU-Regular Exam Apr/May 2007 (Set 3) and
Coefficient of s2 ⇒ A + C = 0 ⇒ A = − JNTU-Suppl. Exam Aug/Sep 2006 (Set 2)
/ , / , 4
/ ,
1 1 1 1 1 C
L−1 2 = − L−1 + L−1 2 8. Verify Green’s theorem for C (3x2 − 8y2 )dx+
s (s + 2) 4 s 2 s
/ , (4y − 6xy)dy where C is the bounding curve of
1 −1 1 region defined by x = 0, y = 0 and x + y = 1.
+ L
4 s+2 [16]
Solution By Green’s theorem, we have y

B (0, 1)
∂N ∂M
Mdx + Ndy = − dxdy (1)
C S ∂x ∂y
x
+
M = 3x2 − 8y2 , N = 4y − 6xy,
y
=
∂N ∂M x=0
1
S
= −6y, = −16y
∂x ∂y

0 x
Now, LHS = Mdx + Ndy y = 0 A (1, 0)
C Figure 8
= + + (2)
OA AB BO
0
Along OA y = 0, dy = 0 Mdx + Ndy = 4ydy = [2y2 ]01 = −2 (5)
1 BO 1
Mdx + Ndy = 3x2 dx = [x3 ]10 = 1 (3)
8 5
OA 0 ∴ LHS = Mdx + Ndy = 1 + − 2 = (6)
3 3
Along AB x + y = 1 ⇒ dy = −dx, x = 1 − y C
∂N ∂M
and y varies from 0 to 1 RHS = − dxdy
S ∂x ∂y
Mdx + Ndy = [3(y − 1)2 − 8y2 ](−dy) 1 1−x
AB = (−6y + 16y)dxdy
+ [4y + 6y(y − 1)]dy x=0

y=0

1
1 1−x 1
= 10 ydy dx = 5 (y2 )1−x
0 dx
= (11y + 4y − 3)dy
2
x=0 y=0 0

0
1 1
y3
1
(1 − x)3
= 11 · + 2y2 − 3y =5 (1 − x)2 dx = −5
3 0 3 0
0
11 8 −5 5
= +2−3= (4) = (0 − 1) = (7)
3 3 3 3
Along BO y = 0, dy = 0 the limits for y are LHS = RHS.

y = 1 to 0 Green’s theorem is thus verified.
1. (a) Test the convergence of the series n $ $1

Solution Let un = . Take vn =
3 3·6 2 3·6·9 3 3n2 − 2 n
1+ x+ x + x +· · · (x > 0). which is divergent.
1 7 · 10 7 · 10 · 13
un n2 1
Solution Let Now = 2 =
3 · 6 · 9 · · · 3n vn 3n − 2 3 − 2/n2
un = xn 1
7 · 10 · 13 · · · (3n + 4) → = 0 as n → ∞
3 · 6 · 9 · · · 3n(3n + 3) 3
⇒ un+1 = xn+1 $
7 · 10 · 13 · · · (3n + 4)(3n + 7) By comparison test, un is divergent. Also
un → 0 as n → ∞ and u1 > u2 > · · · .
and u0 = 1 % (−1)n
Therefore, by Leibnitz’s rule is
un−1 3n + 3 3 + 3n % 3n2 − 2
lim = lim x = lim x=x n
n→∞ un n→∞ 3n + 7 n→∞ 3 + 7 convergent but not . The given series is
n 3n2 − 2
By D’Alembert’s ratio test, the series is convergent conditionally convergent.
for x < 1 and divergent for x > 1, and the test fails
for x = 1. For x = 1, we have 1. (c) Expand ex sec x as a power series in x up to the
term containing x3 . [6]
un 3n + 7
n −1 =n −1
un+1 3n + 3 Solution We know that
4n 4 x x2 x3 x4
= = ex = 1 + x + + + + ··· (1)
3n + 3 3 + 3/n 1! 2! 3! 4!

un 4 4 x2 x4 x6
∴ lim n − 1 = lim = >1 cos x = 1 − + − + ··· (2)
n→∞ un+1 n→∞ 3 + 3 3
n 2! 4! 6!
$
By Raabe’s test, un is convergent. Dividing one by the other we obtain
Hence the series converges if x ≥ 1 and diverges 2 3 4
ex 1 + x + x2 + x6 + x24 + · · ·
if x > 1. e sec x =
x
= 2 4 x6
cos x 1 − x2 + x24 − 144 + ···
1. (b) Test the following series for absolute/
% (−1)n n 2
conditional convergence . = 1 + x + x2 + x3 + · · · (3)
3n2 − 2 3
2. (a) Find the points on the surface z 2 = xy + 1 1 + y12 a2 sin2 θ + b2 cos2 θ a2 sin3 θ
= ·
that are nearest to the origin. y2 a2 sin2 θ −b
sin θ 2 2
Solution Let P(x, y, z) be any point on the surface =− (a sin θ + b2 cos2 θ)
b
z 2 = xy + 1.
Then OP = x2 + y2 + z 2 or u = x2 + y2 + z 2 = Coordinates of the centre of curvature
x + y2 + xy + 1
2
1 + y12
x̄ = x − y1
∂u y2
P= = 0 ⇒ 2x + y = 0
∂x b cos θ −sin θ 2 2
= acos θ− · (a sin θ+b2 cos2 θ)
∂u −a sin θ b
q= = 0 ⇒ 2x + y = 0
∂y cos θ 2 2 2
= (a −a sin θ−b2 cos2 θ)
⇒ x = 0, y = 0 corresponding values for z are a
cos3 θ 2
z 2 = 1 or z = ±1 from Eq. (1) = (a − b2 )
a

∴ The stationary points are P1 = (0, 0, 1) 1 + y12
ȳ = y +
and P2 = (0, 0, −1). At these points y2
rt − s2 = 2 · 2 − 12 = 3 > 0. Also r = 2 > 0. sin θ 2 2
∴ u has minimum value at P1 and P2 . = b sin θ − (a sin θ + b2 cos2 θ)
b
Hence the points on the surface z 2 = xy+1 nearest sin θ 2
to the origin are P1 = (0, 0, 1) and P2 = (0, 0, −1). = (b − a2 sin2 θ − b2 cos2 θ)
b
− sin3 θ 2
2. (b) Prove that if the centre of curvature of the = (a − b2 )
x2 y2 b
ellipse 2 + 2 = 1 at one end of the minor
a b The ends of minor axis are B = (0, b), correspond-
axis lies at the other end, then the eccentricity ing to θ = π/2 and B = (0, −b)
1
of the ellipse is √ .
2 b2 −a2
∴ = −b or 2b2 −a2 = 0. But b = a(1−e2 )
b
Solution Let P = (a cos θ, b sin θ) be any point 1 1
on the ellipse. ∴ 1 − e2 = or e = √ .
2 2
x2 y2 3. (a) Trace the lemniscale of Bernoulli
+ =1 (1)
a2 b2 r 2 = a2 cos 2θ.
dy dy/dθ b cos θ
We have y1 = = = (2)
dx dx/dθ −a sin θ Solution Please refer to Ex. 6.22 on p. 19 of Ch.
6.
cos θ
dy b d cos θ b d
3. (b) The segment of the parabola y2 = 4ax which is
y2 = 2 = − =− dθ
dxsinθ
dx a dx sin θ a dθ
cut off by the latus rectum revolves about the
directrix. Find the volume of rotation of the
b sin θ(− sin θ) − cos2 θ 1
=− · annular region.
a sin θ
2
−a sin θ
=−
b
(3) Solution Volume of area L AL when rotates about
a3 sin3 θ the directrix x = −a is
a2 sin2 θ + b2 cos2 θ a
1 + y12 = √ √
a2 sin2 θ V =2 2π(a + x)ydy when y = 2 a x
x=0
N y L 4. (c) In a chemical reaction, a given substance is

being converted into another at a rate propor-
M tional to the amount of substance unconverted.
P (x, y) 1
If th of the original amount has been trans-
5
Z A S (a, 0) x formed in 4 min, how much time will be
required to transform one half.
Solution Let m be the amount of the substance

L´ at time t and m0 be the original substance at t = 0.
Figure 3b By decay rule
dm
= −km(k > 0) (4)
√ √ a dt
= 2 · 4π a (a x + x3/2 )dx ⇒ m = ce−kt where c = ce−k(0) = m0
0
√ 2a 3/2 2 5/2 a ∴ m(t) = m0 e−kt (5)
= 4π a x + x
3 5 0
1
√ 1 1 128πa3 th of m0 decays in t = 4, i.e.,
= 2 · 4π a2a5/2 + = c.u. 5
3 5 15
1
4. (a) Form the differential equation by eliminating 1− m0 = m(4) = m0 e−k(4)
the arbitrary constant y2 = 4ax. 5
1
y2 ⇒ −4k = log 4/5 ⇒ k = (log 5 − log 4)
Solution Differentiating = 4a w.r.t x, we get 4
x
If T is the time for half-decay, then
a2yy − y2 · 1
= 0 ⇒ 2xy1 = y
x2 1
m0 = m0 e−kT ⇒ −kT = − log 2
which is the required differential equation. 2
log 2 11 log 2
∴T = = .
4. (b) Solve the differential equation k log 5 − log 4
dy tan y
− = (1 + x)ex sec y. 5. (a) Solve the differential equation
dx 1 + x
y − y − 2y = 3e2x ; y(0) = 0, y (0) = 2
Solution Multiplying by cos y, the equation can
be written as Solution The equation is a non-homogeneous
linear differential equation with constant coeffi-
d(sin y) 1 cients. Its complete solution consists of two parts:
− (sin y) = (1 + x)ey
dx 1+x (a) complementary function, yC.F and
which is linear in sin y (b) particular integrals, yP.I .
To find C.F, we have to solve
1 1
I.F = e− 1+x dx = e− log(1+x) =
1+x d
(D2 − D − 2)y = 0, D= (1)
The complete solution is dx

sin y
1
= c + (1 + x)ex ·
1
dx A.E. is m2 −m−2 = (m−2)(m+1) = 0 ⇒ m = 2, −1.
1+x 1+x
= c + ex or sin y = (1 + x)(c + ex ). yC.F = c1 e2x + c2 e−x , (2)
/ ,
where c1 , c2 are arbitrary constants −1 s+1
6. (b) Find L log .
To find P.I, we write s−1
1 Solution Let
yP.I = 3e2x s+1
−D−2
D2 f¯ (s) = L{f (t)} = log
x s−1
= 3e2x = xe2x (3) = log(s + 1) − log(s − 1)
(2D − 1)D=2
d 1 1
The complete solution is ⇒ L{tf (t)} = (−1) f¯ s = −
ds s−1 s+1
y(x) = yC.F + yP.I = c1 e2x + c2 e−x + xe2x (4) = L{et − e−t } = 2 sinh t
2
Differentiating Eq. (4) w.r.t. x, we get ⇒ f (t) = sinh t
t
√
y (x) = 2c1 e2x − c2 e−x + 1 · e2x + 2xe2x (5) 1 1+x2
dxdy
⇒ y(0) = 0 = c1 · 1+c2 ·+1 ⇒ c1 +c2 = 0 (6) 6. (c) Evaluate
0 0 1 + x2 + y2

⇒ y (0) = 2 = 2c1 −c2 +1+0
⎛
Solution √ ⎞
⇒ 2c1 −c2 = 1 (7) 1 1+x2
⎝ dy ⎠ dx
Adding Eqs. (6) and (7), we get 0 0 1 + x2 + y2
1 1 √
3c1 = 1 ⇒ c1 = and c2 = − . 1 1+x2
3 3 −1 y
The required solution is = tan √ dx
0 1 + x2 0
1 2x 1
y = yC.F + yP.I = (e − e−x ) + xe2x . (8) = (tan−1 1 − tan−1 0)dx
3
0
5. (b) Solve the differential equation 1
π π π
(D2 + 1)y = cosec x by variation of = · dx = [x]10 = .
0 4 4 4
parameters method.
7. (a) Find a and b such that the surfaces
Solution Please refer to Ex.2.64(1) on p. 38 Ch. ax2 − byz = (a + 2)x and 4ax2 + z 3 = 4 cut
2. orthogonally at (1, −1, 2).
(a is a misprint in the second equation)
6. (a) Find the Laplace transformation of the
following function: e−3t (2 cos 5t − 3 sin 5t). Solution Let
(x, y, z) = ax2 − byz − (a + 2)x = 0 (1)
Solution We have (x, y, z) = 4x y + z − 4 = 0 2 3
(2)
L{2 cos 5t − 3 sin 5t} Clearly P = (1, −1, 2) lies on Eq. (2) P lies on
= 2L{cos 5t} − 3L{sin 5t} by linearity Eq. (1)
s 5 2s − 15 ⇒a + 2b − a − 2 = 0 ⇒ b = 1 (3)
=2· 2 −3· 2 = 2
s + 25 s + 25 s + 25 ∂φ ∂φ ∂φ
Now ∇φ|p = , ,
By first shifting theorem ∂x ∂y ∂z p
= (2ax − a − 2, −bz, −by)p
2s − 15
L{e−3t (2 cos 5t − 3 sin 5t)} = = (1, −1, 2) = (a − 2, −2b, b) (4)
s2 + 25 s→s+3
2s − 9 ∂ψ ∂ψ ∂ψ
= 2 . ∇|P = , ,
s + 6s + 34 ∂x ∂y ∂z p
= (8xy, 4x2 , 3z 2 )p=(1,−1,2) bounded by x = 0, x = 1, y = 0, y = 1, z = 0

= (−8, 4, 12) (5) and z = 1.
The surfaces φ and ψ cut orthogonally Solution By Gauss’s divergence theorem, we

⇔ ∇φ · ∇ψ = 0 have

⇔ −8(a − 2) + 4(−2b) + 12b = 0
F̄ · n̄ds = ∇ · F̄dv (1)
⇒ −2a + b + 4 = 0 (6) S V
⇒ a = 5/2 (∵ b = 1). (7)

here V is the volume bounded by a closed sur-
7. (b) Show that F̄ = (2xy + 2 )ī + x j̄ + 3xz k̄ is
3 2 2
z
a conservative force field. Find the scalar pot-
ential and the work done by F̄ in moving an
c Q
object in this field from (1, −2, 1) to (3, 1, 4).
P S
Solution Force field F̄ is a conservative force o
field iff it is the gradient of some scalar field φ, which B y
A
is
called the potential of F̄. In this case, the work done R
B
F̄ · d r̄ is path independent. x
A
Figure 8 w
Further the integrand F̄ ·d r̄ = Mdx+Ndy+Pdz =
(2xy + z 3 )dx + x2 dy + 3xz 2 dz is an exact differential. face S and F̄ is a vector function with continuous
A necessary condition for this is that derivatives
∂M ∂N ∂M ∂p ∂N ∂p
= ; = ; = ⇔ curl F̄ = Ō. RHS = ∇ F̄dv
∂y ∂x ∂z ∂x ∂z ∂y
V
ī j̄ k̄ ∂ ∂ ∂
= (4xz) + (−y2 ) + (yz) dv
∂ ∂ ∂ ∂x ∂y ∂z
Here ∇ × F̄ = V
∂x ∂y ∂z
= (4z − y)dv
2xy + z 3 x2 3x23 V
1 1 1
= ī(0)+ j̄(3z 2 −3z 2 )+ k̄(2x−2x) = ō
= (4z − y)dzdydx
= (2xydx + x2 dy) + z 3 dx + 3xz 2 dz x=0 y=0 z=0

= d(x2 y) + d(xz 3 ) = d(x2 y + xz 3 )
1 1 1
= 2z 2 − yz z=0
dydx
Now dφ = (2xy + z 3 )dx + x2 dy + 3xz 2 dz
x=0

y=0
1 1
3
⇒ φ(x, y, z) = x2 + xz 3 = C = (2 − y)dydx = (2)
x=0 y=0 2

is the scalar potential work done by F̄ is independent
of the path joining A = (1, −2, 1) and B = (3, 1, 4) LHS = F̄ · n̄ds = (4xz ī − y2 j̄ + yz k̄) · n̄ds
S S
B B 6
%
∴ F̄ · d r̄ = dφ = φB − φA = F̄ n¯r dsr
A A Sr
3,1,4 n=1
= x2 y + xz 3 1,−2,1 = 202.
where Sr are the six faces of the cube and x¯r are the
8. Verify divergence theorem for F̄ = 4xz ī respective unit normals to them.
−y2 j̄ + yz k̄ where S is the surface of the cube S1 : ARSP, x = 1, n̄ = ī, ds = dydz,
F̄ = 4z ī − y2 j̄ + yz k̄ F̄ · n̄ = F̄ · −̄i = 0
1 1
F̄ · n̄ = F̄ · ī = 4z
1 1 = 0dzdx = 0 (6)
S4 0 0
= 4zdzdy = 2 (3)
S1 0 0
S5 : AOCP, y = 0, n̄ = −̄j, ds = dzdx,
F̄ = 4xz ī − 0j̄ + 0k̄
S2 : BQSR, y = 1, n̄ = j̄, ds = dzdx,
F̄ = 4xz ī − j̄ + z k̄ F̄ · n̄ = F̄ · −̄j = 0
1 1
F̄ · n̄ = F̄ · j̄ = −1
1 1 = 0dzdx = 0 (7)
S5 0 0
= (−1)dzdx = −1 (4)
S2 0 0 ¯ ds = dxdy,
S6 : BOAR, z = 0, n̄ = −k,
F̄ = 0ī − y j̄ + 0
2
S3 : CPSQ, z = 1, n̄ = k̄, ds = dxdy,
F̄ = 4xī − y2 j̄ + yk̄ ¯ =0
F̄ · n̄ = F̄ · −k
1 1
F̄ · n̄ = F̄ · k̄ = y
1 1 = 0dydx = 0 (8)
1 S6 0 0
= ydydx = (5)
S3 0 0 2 1 3
Now, F̄ · n̄ds = 2 − 1 +
+0+0+0=
2 2
S4 : OBQC, x = 0, n̄ = −̄i, ds = dydz, S
∴ Divergence theorem has been verified.
F̄ = 0ī − y2 j̄ + yz k̄
Code No. 07A1BS02 Set No.1
I year B. Tech. Regular Examinations, May/Jun, 2008
1. (a) Solve y (2x2 y + ex )dx = (ex + y3 )dy. Solution Let T (t) be the temperature of the
object at time t.
T (t) = Ta + ce−kt (1)
Mdx + Ndy = 0 (1) Ta = ambient temperature = 80 ◦
where M = 2x2 y2 +yex and N = −ex −y3 (2) ∴ T (t) = 80 + ce− kt; (2)
computing My = 4x y + e and Nx = −e
2 x x
(3) when t = 0, T = 300◦ (3)
(Nx − My ) −2e − 4x y
x 2
−2 300 = 80 + c · e 0
we have q(y) = =
(M ) 2x y + ye
2 2 x y ⇒ c = 300◦ − 80◦ = 220◦ (4)
a pure function of y −kt
∴ T (t) = 80 + 220e (5)
2 1
I.F = e q(y)dy = e− y dy = e−2 log y = 2 (4) We use the condition T (10) = 250 to determine k.
y We have 250 = 80 + 220 e−k(10)
Multiplying Eq. (1) by Eq. (4) we get 250 − 80 17
⇒ e−10k = e−10k = =
x 220 22
1 x e 1 22
2x + e − 2 + y dy = 0
2
⇒k= log = 0.02578 (6)
y y 10 17
1 The required solution is
⇒ 2x2 dx − ydy + d ex = 0 (5)
y
T (t) = 80 + 220 e−0.2578t (7)
Integrating Eq. (5) we get
which gives the temperature of the object at time t.
2 3 1 2 1 x
x − y + e = constant = c. (6) 2. (a) Solve (4D2 − 4D + 1)y = 100.
3 2 y
1. (b) Suppose that an object is heated to 300◦ and Solution This is a non-homogeneous linear equa-
allowed to cool in a room whose air tempera- tion with constant coefficients. Its complete solution
ture is 80◦ . If after 10 min the temperature consists of
of the object is 250◦ . What will be its (a) complementary function yC.F and
temperature after 20 min. (b) particular integral yP.I .
To find yC.F . Solution Let x, y, 100 − x − y be the three

By putting y = em x in (4D2 − 4D + 1)y = 0, we numbers. Let their product be
obtain the auxillary equation
A.E 4m2 − 4m + 1 = (2m − 1)2 = 0 f (x, y) = xy(100 − x − y)
1 1 = 100xy − x2 y − xy2 (1)
⇒m= ,
2 2 f is maximum ⇒ fx = 0, fy = 0 (2)
∴ yC.F = (c1 + c2 x)ex/2
Differentiating Eq. (1) partially w.r.t x and y and
To find yP.I setting fx and fy equal to zero
1
yP.I = (100) = 100
(2D − 1)2 fx = 100y − 2xy − y2 = 0; (3)
The complete solution is fy = 100x − x − 2xy = 0
2
(4)
y = yC.F + yP.I = (c1 + cx )ex/2 + 100. ⇒ 100 − 2x − y = 0; and (5)
⇒ 100 − x − 2y = 0 (6)
2. (b) Solve (D3 − 6D2 + 11D − 6)y = e−2x + e−3x .
If we delete the trivial case x = y = 0 adding Eqs.
Solution This is a non-homogeneous linear equa-
(5) and (6)
tion with constant coefficients.
Its complete solution consists of 200
(a) complementary function (yC.F ) and 200 − 3(x + y) = 0 ⇒ x + y = (7)
3
(b) particular Integral(yP.I ).
Subtracting Eq. (5) from Eq. (6) x = y (8)
To find yC.F .
By putting y = emx in (D3 −6D2 +11D −6)y = 0, 100
Solving x = y = and computing
we obtain the auxiliary equation 3
A.E m3 − 6m2 + 11m − 6 = 0 r = fxx = −2, s = fxy = −1; t = fyy = −2 (9)
⇒ (m − 1)(m − 2)(m − 3) = 0 ⇒ m = 1, 2, 3 · · · We have r = −2 < 0 and
∴ yC.F = c1 ex + c2 e2x + c3 e3x rt − s2 = (−2)(−2) − (−1)2 = 3 > 0
100
To find P.I Hence fmax occurs at x = and
3
1
yP.I = e−2x 100 100 200 100 3
(D − 1)(D − 2)(D − 3) fmax = × 100 − = .
1 3 3 3 3
+ e−3x
(D − 1)(D − 2)(D − 3) √ 1
1 1 3. (b) If f (x) = x and g(x) = √ . Prove that
= e−2x + e−3x x
(−3 − 4 − 5) (−4 − 5 − 6) Cauchy‘s generalised mean value theorem
1 −2x 1 −3x is the geometric mean of a and b for any
= e − e a > 0, b > 0.
60 120
The complete solution is Solution Please refer to Ex. 2.22 on p. 75 of
Engg maths-I (I Edition).
y = yC.F + yP.I
1 −2x 1 −3x √ √ θ
= c1 ex + c2 e2x + c3 e3x − e − e . 4. (a) Find the radius of curvature of a = r cos
60 120 2
at (r, θ) (parabola).
3. (a) Find three positive numbers whose sum is 100
and whose product is maximum. Solution Please refer to Ex. 4.16 on p. 137 of
−1 1 1
Engg Maths-1 (I Edition) and take n = = 1 − x(1 − x − y)dy dx
2 0 0
1−x
ra n
θ 1
1
ρ= = 2r sec = y − xy − y2 dx
(n + 1)r n n=− 12 2 0 2 0
1 1 1
r − 2 = a− 2 cos(−θ/2) 1
* = 1 − x − x + x − (1 − 2x + x ) dx
2 2
a 0 2
⇒ = cos θ/2. 1
r 1 1 2 1 3 1 1 1 1
= − x + x = − ·1+ ·1= .
2 2 6 0 2 2 6 6
4. (b) Find the envelope of the straight lines x/a+
y/b = 1 where a2 + b2 = 4. 5. (b) Find the surface area of the solid generated by
revolving the area of the parabola x2 = 12y
Solution Let
bounded by the latus rectum about the y-axis.
x
f (x, y, a) = + y 4 − a2 − 1 = 0 (1)
a Solution Volume of solid of revolution
(∴ b = 4 − a2 ) 3 3
∂f x a V = πx2 dy = 12π ydy = π6[y2 ]30
= − 2 + y. =0 0 0
∂a a (4 − a2 )3/2 = 6π9 = 54π c.u.
a3 (4 − a2 )3/2
⇒ = (2)
x y 6. (a) Examine the convergence of
1 2 1·2 3 1·2·3 4
a (4 − a ) 2 1/2
a2 + (4 − a2 ) x + x + x + ··· (x > 0).
⇒ = = 3 3·5 3·5·7
x1/3 y1/3 x2/3 + y2/3
2 Solution Let
=
x2/2 + y2/3 1 · 2 · 3···n
un = xn+1
2x1/3 3 · 5 · 7 · · · (2n + 1)
⇒a= ,
x2/3 + y2/3 1·2·3· · ·n(n+1)
Then un +1 = xn+2
2y1/3 3·5·7· · ·(2n+1)(2n+3)
b= 4 − a2 = (3) un + 1 n+1
x2/3 + y2/3 Now = x
un 2n + 3
x2/3 + y2/3 x2/3 + y2/3 1 + 1/n 1
1=x 1/3
=y (4) = x → x as n → ∞
2x 2y1/3 2 + 3/n 2
By D’Alembert’s ratio test,
⇒ (x2/3 +y2/3 )3/2 = 2 ⇒ x2/3 +y2/3 = 22/3 . (5) %
x<2⇒ un is convergent
1 1−x 1−x−y %
5. (a) Evaluate dx dy dz. x>2⇒ un is divergent
0 0 0
x = 2, the test fails
Solution
1 1−x 1−x−y x = 2 we have
I= (dz) dx
0 0 0 1 · 2 · 3···n
1 un = · 2n+1 ,
1−x
3 · 5 · 7 · · · (2n + 1)
= [z]1−x−y dydx
0 1 · 2 · 3 · · · n(n + 1)
0 0
un+1 = · 2n+2
3 · 5 · 7 · · · (2n + 1)(2n + 3)

un 2n + 3 Along C1 :
Now n −1 =n 12 − 1
un +1 n+1 y = 0, z = 0, dy = 0, dz = 0; x varies from 0 to a
n a
= a3
2x + 2 F̄ · d r̄ = (x2 − 0)dx + 0 = (3)
1 1 C1 x=0 3
= → as n → ∞
2+2/n 2 Along C2 :
$
By Raabe’s test, un is divergent . x = a, z = 0, dx = 0, dz = 0; y varies from 0 to b
b
6. (b) Examine
$ the convergence of F̄ · d r̄ = (a2 − y2 ) · 0 + 2aydy = ab2 (4)
[(n + 1)!]2 n−1 C2 y=0
x .
n Along C3 :
Solution y = b, z = 0, dy = 0, dz = 0; x varies from a to 0
[(n + 1)!]2 n−1
Let un = x ; 0
n a3
F̄ · d r̄ = (x2 − b2 )dx + 2bx = − + ab2
[(n + 2)!2 ] n C3 x=a 3
Then un + 1 = x
n+1 (5)
un + 1 [(n + 2)!]2 n
Now = · x Along C4 :
un n+1 [(n + 1)!]2 x = 0, z = 0, dx = 0, dz = 0; y varies from b to 0

1 0
= (n + 2) 2
x
1 + 1n F̄ · d r̄ = (0 − y2 ) · 0 + 2 · 0 · ydy = 0 (6)
C4 b
→ ∞ as n → ∞
$ Adding these we get RHS
By theorem which states that if un is a series
un + 1
of positive terms such that → ∞ as n → ∞ = F̄ · d r̄ = 2ab2 (7)
$ un c
then un diverges to +∞, the above series diverges
to +∞. Now we compute
ī j̄ k̄
7. Verify Stokes’ theorem for F̄ = (x2 − y2 )ī + 2xyj̄ ∂ ∂ ∂
over the box bounded by the planes x = 0, ∇ ×F = = 4yk̄
∂x ∂y ∂z
x = a, y = 0, y = b, z = c. x2 − y2 2xy 0
Solution Stokes’ theorem :
D LHS of Eq. (1) = ∇×F̄ n̄ds
(∇ × F)n̄ds = F̄ · d r̄ (1) s
S C a b
dxdy
F = (x −y , 2xy, 0);
2 2
= 4yk̄ · n̄ = 4ydy dx
R |k · n̄| x=0 y=0
r̄ = (x, y, z), d r̄ = (dx, dy, dz) a
D D
= [2y2 ]b0 dx = [2b2 x]a0 = 2ab2 (8)
F̄ · d r̄ = (x2 − y2 )dx + 2xydy (2) 0
C
C
LHS of Eq. (1) = RHS of Eq. (1)
= + + + F̄ ·d r̄ Hence Stokes’ theorem is verified.
C1 C2 C3 C4
where C1 , C2 , C3 and C4 are line segments OA, AB, 8. (a) Using Laplace transform solve
BC and CO, respectively. (D2 + 2D − 3)y = sin x, y(0) = y (0) = 0.
Solution Let L{y(x)} = ȳ(s). 1 1

C= =
Taking Laplace transforms on both sides of (s2 + 1)(s + 3) s=1 8
(D2 + 2D − 3)y = sin x; (1) 1 1

D= =−
y(0) = y (0) = 0 we get (2) (s2
+ 1)(s − 1) s=−3 40
1 1 s 1 1
s2 ȳ−sy(0)−y (0)+2[sȳ−y(0)]−3ȳ = ∴ ȳ(s) = − · 2 − ·
s2 +1 10 s + 1 5 s2 + 1
1 1 1 1 1
⇒ (s2 + 2s − 3)ȳ = + −
+1s2 8 s − 1 40 s + 3
on using conditions Eq. (2) Inverting we get
1
⇒ ȳ = 2 (3) 1 1 1 1
(s + 1)(s − 1)(s + 3) y(x) = − cos x − sin x + ex − e−3x .
10 5 8 40
Putting the RHS of Eq. (3) into partial fractions , let
8. (b) Using
∞ Laplace transforms
evaluate
1 e−t − e−2t
dt.
(s2 + 1)(s − 1)(s + 3) 0 t
A B C D
= + + + Solution / ,
s−i s+i s−1 s+3 f (t) ∞
(A + B)s + (A − B)i C D If L{f (t)} = f¯ (s) then L = f¯ (s)ds

= + + t s
s2 + 1 s−1 s+3 1 1
1 Now L{e−t − e−2t } = − = f¯ (s), say
A= s+1 s+2
(s + i)(s − 1)(s + 3) s=i / −t , ∞
e − e−2t
1 ∴L = f¯ (s)ds
= t s
2i(−1 + 2i − 3) ∞ ∞
1 1
1 −1 + 2i = ds − ds
=− = s s + 1 s s + 2
4 + 8i 20 ∞
s+1 s+1
−1 − 2i = log = log 1 − log
B= s+2 s s+2
20
1 1 s+2
A + B = − , (A − B)i = − = log .
10 5 s+1
dy Aliter: We may write the equation as

1. (a) Solve x + y = x3 y6 .
dx d(xy) = (xy)6 x−3 dx
1 variables and integrating
Separating the
Solution Multiplying the equation by , we get
x (xy)−6 d(xy) = x−3 dx + c1
dy 1 1 1
+ y = x2 y6 (1) ⇒ − (xy)−5 + x−2 = c1 .
dx x 5 2
which is a Bernoulli’s equation. 1. (b) A bacterial culture growing exponentially
1 increases from 200 to 500 gms in the period
Multiplying again by 6 , we get
y from 6 A.M to 9 A.M. How many grams will
be the present at noon.
1 dy 1 1
+ = x2 (2)
y6 dx x y5 Solution The governing differential equation is
1 1 dN dn
Putting 5 = u ⇒ −5 6 dy = du or (3) = kN ⇒ = kdt
y y dt N
1 1 du 1 ⇒ log N = kt + log c
dy = − du we have − 5 u = −5x2 (4)
y6 5 dx x ⇒ N = cekt (1)
which is linear in u. When t = 0, N = 200 ⇒ 200 = ce 0
1 1 ⇒ N = 200ekt (2)
I.F = e−5 x dx
= e−5 log x =
x5 t = 3 hrs Nt=3 = 500 = 200e k×3
The complete solution of the equation is 500 5

⇒ e3k = = (3)
200 2
1 1 1 2
u · 5 = 5 5 = c1 − 5 x2 · 5 dx 5
x x y x t = 6 hrs Nt=6 = 200ek×6 ⇒ 200 ×
2
= c1 − 5 x−3 dx After 3hrs the number of bacteria present will be
1/3
5 5
x−5 y−5 = c1 + x−2 (c1 ; arbitrary const). (5) N = 200 (3) = 814.325 gm (4)
2 2
2. (a) Solve (D2 − 4D + 13)y = e2 x. Solution By Lagrange’s Mean Value Theorem,

f is monotonically increasing in [a, b] if
Solution This is a second order non-
homogeneous linear differential equation with con- f (p) − f (a)
f (c) = > 0 for a < c 0
To find C.F, ⇒ x < −2 < 0 ⇒ x > −2
We put y = emx in (D2 − 4D + 13)y = 0 and obtain
auxillary equation A.E as m2 − 4m + 3 = 0 ⇒ Thus, f is monotonically increasing in (−∞, −2)
(m − 2)2 = −9 ⇒ m = 2 ± 3i and monotonically decreasing in (−2, ∞).
yC.F = (c1 cos 3x + c2 sin 3x)e2 x 3. (b) Find the minimum value of x2 + y2 + z 2 given
x + y + z = 3a.
Solution Let
1 1
yP.I = 2 e2x = 2 e2x f (x, y, z) = x2 + y2 + z 2 = x2 + y2 + z 2 , (1)
D − 4D + 13 2 − 4 · 2 + 13
φ(x, y, z) = x + y + z − 3a (2)
(replacing D by 2)
The complete solution is Differentiating f partially w.r.t x and y, we get
fx = 2x fy = 2y fz = 2z
1
y = yC.F + yP.I = (c1 cos 3x + c2 sin 3x)e + e2x .
2x
φx = 1 φy = 1 φz = 1
9
By Lagrange’s method of multipliers
2. (b) Solve (D2 + 16)y = e−4x .
−λ
Solution This is a second order non- fx + λφx = 0 ⇒ x =
2
homogeneous differential equation with constant −λ
coefficients. Its complete solution consists of fy + λφy = 0 ⇒ y =
2
(a) complementary function yC.F and −λ
(b) particular Integral yP.I . fz + λφz = 0 ⇒ z =
2
To find C.F.
We put y = emx in (D2 + 16)y = 0 and obtain Substituting these values in Eq. (1) we get
auxillary equation A.E m2 + 16 = 0 ⇒ m = ±4i
λ = −2a ∴x=y=z=a
∴ yC.F = c1 cos 4x + c2 sin 4x The possible extreme point is (a, a, a)
To find P.I, we write ∴ fmin = 3a2 .
1 1 −4x
yP.I = 2 e−4x = e , replacing D by –4 4. (a) Show that the evolute of the parabola
D + 16 32
y2 = 4ax is 27ay2 = 4(x − 2a)3 .
1 −4x Solution Please refer to Ex. 5.23 on p. 13 of Ch.

y = yC.F + yP.I = c1 cos 4x + c2 sin 4x + e . 5.
32
3. (a) Find the region in which f (x) = 1 − 4x − x2 4. (b) Find the equation of the circle of curvature of
is increasing and the region in which it is the curve x = a(cos θ + sin θ)
decreasing using Mean Value Theorem. y = a(sin θ − cos θ).
Solution Differentiating w.r.t θ, we have log 2

dx = [e3x (x − 1) + ex ]dx
= a(− sin θ + sin θ + θ cos θ) = aθ cos θ 0
dθ log 2
dy e3x e3x
= a(cos θ − cos θ + θ sin θ) = aθ sin θ = (x − 1) − +e x
dθ 3 9 0
∴ y1 = tan θ, 1 + y12 = sec2 θ 8 8 1 1
= (log 2 − 1) − + 2 − + − 1
dy1 dθ sec2 θ 1 3 9 3 9
y2 = · = = sec3 θ 8 19
dθ dx aθ cos θ aθ = log 2 − .
3 9
(1 + y12 )3/2
⇒ρ= = aθ
y2 5. (b) Find the volume of the solid that results when
Coordinates of the centre of curvature x2 y2
the region enclosed by the ellipse 2 + 2 = 1
y1 (1 + y12 ) a b
x̄ = x − (0 < b < a) rotates about its major axis.
y2
tan θ sec2 θ Solution Volume of the generated solid
= a(cos θ + θ sin θ) − = a cos θ
1
sec3 θ a a
aθ 2πb2 a 2
1+ y12 =π y2 dx = 2πb2 dx −
x dx
ȳ = y + = a(sin θ − θ cos θ) + aθ cos θ −a 0 a2 0
y2 x2 y2 b2 2
= a sin θ + = 1 ⇒ y 2
= b 2
− x
a2 b2 a2
The equation of the circle of curvature is 2πb2 1
= 2πb2 [x]a0 − 2 [x3 ]a0
a 3
(x − x̄)2 + (y − ȳ)2 = ρ2 2
2πb 4π 2
⇒ (x − a cos θ)2 + (y − a sin θ)2 = a2 θ 2 = 2πab2 − a3 = ab c.u
3a2 3
log 2 x x+log y √
$√
5. (a) Evaluate ex+y+z dzdydx 6. (a) Test the convergence of ( n3 − 1 − n3 ).
0 0 0
√ √
√Let un = ( n − 1 − n ) =
Solution 3 3
Solution √
log 2 x x+log y
( n3 − 1 + n3 )−1 be the nth term of the given
1
I= ex · ey · ez dzdydx series and vk = 3/2 be the nth term of compari-
0 0 0 n
log 2 x $ 1
= e · ey [ez ]0
x x+log y
dydx son series which is convergent being p-series
n3/2

0

0 with p = 3/2 > 1.
log 2 x
= ex · ey (ex+log y − 1)dydx un n3/2

0 0 Now, lim = lim √ √
log 2 x n→∞ vn n→∞ ( n3 − 1 + n3 )
= (e2x yey − ex · ey )dydx 1 1

0 0 = lim + =
log 2 n→∞
1+ 1 +1 2
= [e2x (y − 1)ey − ex · ey ]x0 dx n3

0 $ $
log 2 By comparison test, both
$ un and vn converge
= [e (x − 1)e − (0 − 1) − e (e − 1)]dx
2x x x x
or diverge together. But vn is convergent.
0
log 2 $Therefore,
$ √the given series
√
= [e3x (x − 1) + e2x − e2x + ex ]dx un = ( n3 − 1 − n3 ) also converges.
0
% xn / t ,
6. (b) Test the convergence of (x > 0). = L{1} − L{e−t } + L y(t − u) sin udu
xn−1 0
n n
x nx 1 1 1
Solution Let un = = n be the nth term = − + ȳ(s) · 2
xn−1 x s s−1 s +1
of the given series.
1
1 (xn )1/n n1/n where = L{sin t} = L{g(t)}ḡ(s)
lim unn = lim =0 s2+1
n→∞ n→∞ n

$ 1 1 1
∴ un is convergent by Cauchy’s root test. ⇒ ȳ(s) 1 − 2 = −
s +1 s s+1

7. Verify Stokes’ theorem for F̄ = (2x − y)ī s2 + 1 1 1 s2 + 1
⇒ ȳ(s) = − = 3
−yz 2 j̄ − y2 xk̄ where S is the upper half surface s s s+1 s (s + 1)
x2 + y2 + z 2 = 1 of the sphere and C is the
boundary. s2 1
ȳ(s) = +
s3 (s
+ 1) s3 (s + 1)

Solution Please refer to Ex. 11.42 on p. 29 of 1 1 1 − s + s2 − (s3 /s + 1)
Ch. 11. = − +
s s+1 s3
2 2 1 1
8. (a) Using Laplace transform evaluate = − − 2+ 3
∞ s s+1 s s
cos at − cos bt
.
0 t Putting into partial fractions;
Solution Please refer to Ex. 12.20 (v) on p. 17
1 s3
of Ch. 12. = (1 + s)−1 = 1 − s + s2 −
s+1 s+1
/ ,
8. (b) Using Laplace transform
t solve −1 2 2 1 1
y(t) = 1 − e−t + 0 y(t − u) sin udu. ∴ y(t) = L − − + 3
s s + 1 s2 s
1 2
Solution Let L{y(t)} = ȳ(s) = 2 − t + t − 2e−t .
2
dy dN (H )
1. (a) Solve x3 sec2 y + 3x2 tan y = cos x. = kN (H ) ⇒ N (H ) = cekt
dx dt
Solution The equation after division by x3
When t = 0, N (0) = 100 = ce0 ⇒ c = 100
becomes
Eq. (1) now becomes N (t) = 100ekt
dy 3 1
sec2 y + tan y = 3 cos x
dx x x t = 10 N (10) = 100e10k = 400
Putting tan y = u ⇒ sec2 ydy = du, we obtain 400
⇒ e10k = =4
100
du 3 1
+ u = 3 cos x, t = 3 N (3) = 100e3k = 100(ek )3
dx x x
= 100(41/10 )3 = 100(43/10 ).
which is linear in u = tan y
3
I.F = e Pdx
= e x dx = e3 log x = x3 2. (a) (b) Same as in Set 1 May/June 2008 Exam.
3. (a) Find the maxima and minima of the function
1 f (x) = 2(x2 − y2 ) − x4 + y4 .
ux = x tan y = c +
3 3
cos x · x3 dx
x3
Solution
= c + sin x
f (x, y) = 2(x2 − y2 ) − x4 + y4
Aliter: Nothing that the LHS is an exact differential
we have
Differentiating partially w.r.t x and y, we get
d(x3 tan y) = cos xdx ⇒ x3 tan y = c + sin x.
fx = 4x − 4x3 and fy = −4y + 4y3
1. (b) A bacterial culture, growing exponentially
increases from 100 to 400 gms, in 10 hrs. How
much was present after 3 hrs? Stationary points are obtained by solving
fx = 0, fy = 0 as O = (0, 0) and
Solution The differential equation is A1 = (1, 1), A3 = (1, −1), A4 = (−1, 1)
Differentiating again we get Differentiating again

r = fxx = 4 − 12x 2
s = fxy = 0 (1 − 2y)(6x + 2) − (3x2 + 2x − 1)(−2y1 )
y2 = ;
t = fyy = −4 + 12y 2
(1 − 2y)2
At O(0, 0) r > 0, y2 |(0,0) = 4
δ = rt − s2 = 4(−4) − 0 = −16 < 0 Radius of curvature at P = (0, 0)
A1 (1, 1) r < 0,
(1 + y1 )3/2 [1 + (−1)2 ]3/2
δ = rt − s2 = (−8)(8) − 0 = −64 < 0 ρp = =
y2 P 4
These are saddle points. The function has √
2 2 1
minimax at these points. Same is the situation at = =√
4 2
A2 , A3 , A4 .
Coordinates of the centre of curvature
Note
y1 (1 + y1 )2 (−1)2 1
In the question, it should be f (x, y) and not f (x). x̄ = x − =0− =−
y2 4 2
3. (b) Prove, using mean value theorem (1 + y1 )2 2 1
ȳ = y + =0+ =
| sin u − sin v| ≤ |u − v|. y2 4 2
Solution Let f (x) = sin x, g(x) = x. Then f , g The required circle of curvature at (0, 0) for the
are continuous on any interval [a, b] and derivable in given curve is
(a, b) and g (x) = 1 = 0 for all x. By Cauchy mean
(x − x̄) + (y − ȳ) = ρ2
value theorem
f (c) f (b) − f (a) 1 2 1 1
= ⇒ x+ + y− = √
g (c) g(b) − g(a) 2 2 d
we have f (u) = sin u, f (v) = sin v, ⇒ x2 + y2 + x − y = 0.
g(u) = u, g(v) = v x2 y2
f (v) − f (u) sin v − sin u 4. (b) Find the evolute of the ellipse 2
+ 2 = 1.
∴ = a b
g(v) − g(u) v−u
Solution Please refer to Ex. 4.24 on p. 145
f (c) cos w
= = (u < w < v) Engg. Maths-I (I Edition).
g (c) 1
f (v) − f (u) x2
Now, = | cos w| ≤ 1 5. (a) Find the volume of the solid when ellipse +
g(v) − g(u) a2
y2
⇒ |f (u) − f (v)| = |u − v|. = 1 (0 < b < a) rotate about its minor axis.
b2
4. (a) Find the circle of curvature at (0, 0) for
x + y = x2 + y2 + x3 . Solution Volume of solid of revolution
b b
a2
Solution Differentiating x + y = x2 + y2 + x3 = 2π x2 dy = 2π a2 y − 2 y3
with respect to x, we get 0 3b 0
2

a 3 4πba2
1 + y1 = 2x + 2yy1 + 3x2 = 2π ba − 2 b =
2
c. u.
3b 3
⇒ y1 (1 − 2y) = 3x2 + 2x − 1
5. (b) By transforming
into polar coordinates
3x2 + 2x − 1 x2 y2
y1 = ⇒ y1 |(0,0) = −1
1 − 2y evaluate
x2 + y2
dxdy over the annular
$ $ $
region between the circles x2 + y2 = a2 (−1)n−1 un where |(−1)n−1 un | = un is
and x2 + y2 = b2 with b > a. convergent.
Therefore, the given series is absolutely conver-
Solution Put x = r cos θ, y = r sin θ gent.

xy 2 2
I= dxdy
x2 + y2 7. Evaluate F̄ · d s̄ if F̄ = yz ī + 2y2 j̄ + xz 2 k̄ and
b 2π 4 S
r cos2 θ sin2 θ S is the surface of the cylinder x2 + y2 = 9 con-
= rdθdr
r=α θ=0 r2 tained in the first octant between the planes z = 0
4 b 2π and z = 2.
r sin 2θ 2
= dθ
4 a 0 2
2π
Solution
By Gauss’s divergence theorem,
1 4 1 − cos 4θ F̄d s̄ = δ · F̄dv
= (b − a4 ) dθ
16 0 2 S V
2π
b4 − a4 1 1 π 4 ∂ ∂ ∂
= θ − sin 4θ = (b − a4 ). δ · F̄ =(yz) + (2y2 ) + (xz 2 )
16 2 8 0 16 ∂x ∂x ∂x
% 1 = 0 + 4y + 2xz
6. (a) Examine the convergence of . √
n(2n + 1) 3 9−x2 2
RHS of Eq.(1) = (4y + 2xz)dzdydx
1 1 x=0 y=0
√
z=0
Solution Let un = and vn = 2 be
n(2n + 1) n 3 9−x2 2
$ 1 = 4yz + xz 2 0 dydx
the nth term of the comparison series which x=0 y=0
n2 √
is known to be convergent, being p-series with p = 3 9−x2
2 > 1. = (8y + 4x)dydx
x=0 y=0
un n2 1 √
= = as n → ∞ 3 9−x2
vn n(2n + 1) (2 + 1/n) = 4y2 + 4xy dx
$ $ x=0
y=0
Therefore, by comparison test$ un and vn 3

converge or$diverge together. But vn is convergent = 4(9 − x2 ) + 4x(9 − x2 )1/2 dx
and hence un is also convergent. x=0
3 3
4
6. (b) Examine the convergence of = 36x − x3 −2 (9 − x2 )1/2
3 0 x=0
1 1 1 1
− + − . d(9 − x2 )
1 · 3 · 5 3 · 5 · 7 5 · 7 · 9 7 · 9 · 11
3
4 2
Solution Let = 36x − 33 − 2(9 − x2 )3/2 ·
3
1 1 3 3 0
un = and vn = 3
(2n − 1)(2n + 1)(2n + 3) n 4
= 108 − 36 + × 9 × 3 = 108.
un n3 3
=
vn (2n − 1)(2n + 1)(2n + 3) 8. (a) Using Laplace transform, solve
1 1 (D2 + 5D − 6) = x2 e−x , y(0) = a, y (0) = b.
= →
(2 − 1/n)(2 + 1/n)(2 + 3/n) 8
Solution Let L{y(x)} = ȳ(s). Applying Laplace
$ $ 1 transforms on both sides of the differential equation
Since vn = is convergent.
n3 $
Hence, by comparison test un is also con- 2
vergent. The given series is an alternating series [s2 ȳ−sy(0)−y (0)]+5[sȳ−y(0)]−6ȳ =
(s+1)3
as + 5a + 6 2 1 1 2 1
⇒ ȳ(s) = + + +
(s + 6)(s − 1) (s + 6)(s − 1)(s + 1)3 28 s − 1 875 s + 6
as+5a+6 a−b
A= = a − b −6x 6a + b x 1
(s−1) (s=−6) 7 y(x) = e + e − x2 e−x
7 7 50

as+5a+6 6a + b 19 −x 1 x 2 −6x
B= = − e + e + e .
(s+6) (s=1) 7 500 28 875
as + 5a + 6 A B 8. (b) Using
∞ Laplace transform evaluate
Let = + cos 5t − cos 3t
(s + 6)(s − 1) s+6 s−1 dt.
t
a−b 1 6a + b 1 0
= +
7 s+6 7 s−1 Solution We know that
2
Let L{cos at − cos bt} = L{cos at} − L{cos bt}
(s + 1) (s − 1)s + 6
3
s s
2 1 1 = 2 − 2
= − s + a2 s + b2
7(s + 1)3 s − 1 s + 6

2 1 1 By the theorem on integration
= −
7 (s + 1)3 (s − 1) (s + 1)3 (s + 6) / , ∞
f (t)
2 1 1 L = F(u)du where F(s) = L{f (t)}
= − t s
7 s13 (s1 − 2) s13 (s1 + 5)

2 1 1 −1 1 1 −1 / , ∞
= − 3 1− s1 − 3 1+ s1 cos at −cos bt s s
7 2s1 2 5s1 5 ∴L = − ds
t s2 +a2 s2 +b2
s
2 1 1 1 2 1 s13 1 s 2 + a2 ∞
= − 3 1 + s1 + s1 + = log 2
7 2s1 2 4 8 1 + 12 s1 2 s + b2 s
2
1 s + a2
1 1 1 2 1 s13 = 0 − log 2
− 3 1 − s1 + s1 − 2 s + b2
5s1 5 25 125 1 + 15 s1
1 s +a
2 2
2 1 1 11 11 1 1 = log 2
= − · 3− 2− + + 2 s + b2
7 2 s1 4 s1 8 s1 8 s1 − 2
Here a = 5, b = 3 and s → 0
11 1 1 1 1 1
− 3+ − + ∞/ ,
5 s1 25 s12 125 s1 125 cos 5t − cos 3t 1
∴ dt = log (3/5)2
1 1 3 1 19 1 t 2
=− − − s
5 (s + 1)3 50 (s + 1)2 500 s + 1 = log 3/5.
a2 (xdy − ydx) N (3/2) = 100ek(3/2) = 100(ek )3/2

1. (a) Solve xdx + ydy = .
x2 + y2 = 100(33/2 ) by Eq. (3) ⇒ N (3/2) = 519.6
√ √
Solution We may write the equation as ( 3 = 1.732, 3 3 = 5.196).
a2 xdy−ydx 2. (a) Solve (D2 + 1)y = sin x sin 2x + ex x2 .

1 x2
d(x + y ) =
2 2
2 1 + (y/x)2 Solution The given equation is
1 d(y/x)
⇒ d(x2 + y2 ) = a2 (D2 + 1)y = sin x sin 2x + x2 (1)
2 1 + (y/x)2
Integrating we get the solution as which is a second order non-homogeneous linear
equation with constant coefficients.
1 2 Its complete solution consists of
(x + y2 ) = a2 tan−1 (y/x) + c.
2 (a) complementary function, yC.F and
Please refer to Ex. 7.36 on p.241 of Engg. Maths-I (I (b) particular integral, yP.I .
Edition). To find C.F, we have to put y = emx in
1. (b) The number N of bacteria in a culture grew at (D2 + 1)y = 0 (2)
a rate proportional to N . The value of N was
initially 100 and increased to 332 in 1 hr. and obtain auxiliary equation
What was the value of N after 1 12 hrs?.
A.E m2 + 1 = 0 ⇒ m = ±i (3)
Solution The differential equation of the yC.F = c1 cos x + c2 sin x (4)
problem is
To find P. I, we write
dN 1 1
= k · N ⇒ N (t) = cekt (1) yP.I = sin x sin 2x + 2 x 2 ex
dt D2 + 1 D +1
When t = 0, N = 100 ⇒ N (t) = 100ekt (2) 1 1 1 1
= · 2 cos x − cos 3x
When t = 1, N = 332 ⇒ N (t) = 100e kt
(3) 2 D +1 2 D2 + 1
1 1
When t = 1 , N (t) =? +x x2
2 (D + 1)2 + 1
1 −yz z y
(∴ sin x sin 3x =cos x − cos 3x)
2 x2 x x
z −xz x
x 1 1 ex = = x−2 y−2 z −2
= sin x − · cos 3x + y y2 y
4 2 −32 + 1 2 z x −xy
−1
D2 z z z2
1+D+ x2
2 −yz zx xy
x 1 ex = yz −xz xy
= sin x + cos 3x + yz xz −xy
4 16 2

D2 −1 1 1
1−D− + D + · · · x2
2 (yz)(zx)(xy)
2 = 1 −1 1
x2 y2 z 2 1 1 −1
x 1 1
= sin x + cos 3x + ex (x2 − 2x + 1)
4 16 2 −1 1 1
The complete solution is = 0 0 2 = 4.
0 2 0
x
y = yC.F + yP.I = c1 cos x + c2 sin x + sin x
4
of curvature of x + y = 3axy
3 3
1 ex 4. (a) Find
the radius
+ cos 3x + (x − 1)2 (5) 3a 3a
16 2 at , .
2 2
3. (a) Using Rolle’s theorem show that g(x) = 8x3 −
6x2 − 2x + 1 has a zero between 0 and 1. Solution Please refer to Ex. 5.6 on p. 4 of Ch. 5.
Solution Let x y
4. (b) Find the envelope of + = 1 where
a b
f (x) = g(x) = 8x3 − 6x2 − 2x + 1 (1) am bn = cm+n
⇒ f (x) = 2x − 2x − x + x + c
4 3 2
(2)
Solution Please refer Ex. 5.32(e) on p. 17 in Ch.
f (x) being a polynomial function is continuous and 5 of Engg. Maths-I (Revised Edition).
derivable for all x. In particular, f is continuous in
[c, 1] and derivable in (0, 1). 5. (a) By changing the order of integration evaluate
√
a a2 −x2
Also, f (1) = 2 · 1 − 2 · 1 − 1 + 1 + c
4 3 2
a2 − x2 − y2 dydx.
⇒ f (0) = c = c ⇒ f (0) = f (1) 0 0
This f satisfies all the conditions of Rolle’s Solution Let

theorem and hence there exists at least one point x
in (0, 1) such that f (x) = g(x) = 0. √
a a2 −x2
So, 8x3 −6x2 −2x+1 has a zero between 0 and 1. I= a2 − x2 − y2 dydx (1)
0 0
yz xz xy ∂(u, v, w)
3. (b) If u = ,v = ,w = find .
x y z ∂(x, y, z) The double integral √ is to be evaluated first w.r.t y
from y = 0 to y = a2 − x2 and then w.r.t x from
Solution
∂u ∂u ∂u x = 0 to x = a.
∂x ∂y ∂z Thus the domain of integration is the portion of
∂(u, v, w) ∂v ∂v ∂v the circle x2 + y2 = a2 in the quadrant I.
=
∂(x, y, z) ∂x ∂y ∂z If we change the order of integration, then we have
∂w ∂w ∂w to evaluate it first w.r.t x from x = 0 to x = a2 − y2
∂x ∂y ∂z and then w.r.t y from y = 0 to y = a.
⎛ √ ⎞
a a2 −x2 x2
⎝ = → x2 as n → ∞
∴I = (a2 − y2 ) − x2 dx⎠ dy (1 + )
1 3/2
n+2
y=0 0
∴ By D’Alembert’s ratio test
a
x a2 − y 2 $
= (a2 − y2 ) − x2 + u converges for x2 < 1 (x < 1)
y=0 2 2 $ n
√ vn converges for x2 > 1 (x > 1)
a2 −y2 1 1
x For x = 1, un = ; take vn = 3/2 so that
sin−1 dy (n + 2) 3/2 n
a − y2 0
2
$ $ 1
a 2 vn = being p-series with p = 3/2 > 1 is
a − y2 π n3/2
= 0+ − 0 dy convergent.
y=0 2 2
a
π un n3/2
= (a2 − y2 )dy Now lim = lim
4 0 n→∞ vn n→∞ (n + 2)3/2
a
π 2 1 π 1
= a y − y3 = a3 . = lim =1
4 3 6 n→∞ (1 + 2 )3/2
0 n
$ $
5. (b) Evaluate (x + y + z) dzdydx where R By comparison
$ test, both un and vn$
behave
R alike. Since vn is convergent
$ therefore vn is
is the region bounded by the planes x = 0, also convergent. Finally, un is convergent for
x = 1, y = 0; z = 0, z = 1. x ≤ 1 and is divergent for x > 1.
Solution % 1
6. (b) Examine the convergence of .
1 1 1
(n3/2 +n+1)
I= [(x + y + z) dz] dydx
x=0 y=0 z=0
$ $ 1
Solution Take vn = as a comparison
1 1
1 1 n3/2
= xz + yz + z 2 dy dx series which is convergent being p-series with
x=0 y=0 2 0
1 1 p = 3/2 > 1.
1 1
= x+y+ dydx Let un = 3/2
x=0 y=0 2 (n + n + 1)
1 un n3/2
1 1 1 Now = 3/2
= xy + y2 + y dx vn n +n+1
x=0 2 2 0
1 1 1
1 2 = → 1 as n → ∞
= (x + 1) dx = x +x 1 + n1/2
1
+ n3/2
1
2 $ $
x=0 0 By comparison theorem both $un and vn con-
1 3 verge or diverge together. Since vn is convergent
= +1= . $ $
2 2 1
therefore vn = is convergent.
6. (a) Examine the convergence or divergence of (n3/2 + n + 1)
% x2n
√ x > 0. 7. (a) Prove that grad (F̄ · Ḡ) = F̄ × ( × Ḡ)
(n + 2) n + 2 +Ḡ × ( × F̄) + (F̄ · )Ḡ + (Ḡ · )F̄.
Solution Let Solution Please refer to Ex. 4 (c) on p. 15 of
x2n Ch. 10 (Section 10.4).
un =
(n + 2)3/2
un+1 x2n+2 (n + 2)3/2 7. (b) Evaluate F̄d r̄ where F̄ = 3xyī − y2 j̄ and C
⇒ = · c
un (n + 3)3/2 x2n is the parabola y = 2x2 from (0, 0) to (1, 2).
Solution For parabola, c : y = 2x2 the para- 8. (b) Using Laplace transform evaluate
∞ −at 2
metric equations are x = t, y = 2t 2 from t = 0 to e sin t
t = 1. dt.
0 t
Also, r̄ = x(t)ī + y(t)j̄ = t ī + 2t 2 j̄ Solution The given integral is the Laplace
d r̄ sin2 t
⇒ = ī + 2t j̄ transform of with s = a
dt t
1 )
∞
d r̄ = [3t · 2t 2 ī = (2t 2 )2 j̄] · (ī + 2t j̄)dt sin2 t −st sin2 t
c 0 L = e dt
1 1 t s=a 0 t s=a
6 4 16 6
= (6t − 16t )dt =
3 5
t − t
0 4 6
0
∞
3 8 7 sin2 t
= − =− . = e−at dt
2 3 6 0 t
/ , 1
−1 e−2s L{sin2 t} = L{1 − cos 2t}
8. (a) Find L . 2
s2 + 4s + 5
1 1 1 s
= · − · 2
Solution 2 s 2 s + a2
I shifting theorem:
1 1 s
= − 2
2 s s +4
L−1 {f¯ (s)} = f (t) ⇒ L−1 {f¯ (s − a)} = eat f (t)
By definition of t
II shifting theorem:
)
sin2 t 1 ∞ 1 s
L−1 {f¯ (s)} = f (t) ⇒ L−1 {e−as f¯ (s − a)} L = − 2
t 2 s s s +4
f (t − a)U (t − a), t > a ∞
= 1 1
0, t<a = log s − log(s2 + 4)
2 2
/ , / , s
−1 1 −1 1 1 s ∞
∴L =L = log √
s2 + 4s + 5 (s + 2)2 + 1 2 s2 + 4 s
−2t
= e sin t by Eq. (1) 1 s
= 0 − log √
/ , 2 s2 + 4
1 √
L−1 e−2s 1 s +4
2
s2 + 2s + 5 = log
∞ 2 s
e−2(t−2) sin(t − 2)U (t − 2), t>2 sin 2
t ∞
sin2 t
= e −at
dt = e −st
dt
0, t<2 0 t 0 t s=a
√
e−2t+4 sin(t − 2)U (t − 2), t>2 1 a2 + 4
= = log .
0, t<2 2 a
Bibliography
Kreyszig, E, Advanced Engineering Mathematics, 8th Ed., John Wiley, 2000.
Lang, Serge, Linear Algebra, 3rd Ed., Springer, 1987.
O’ Neil, Peter V, Advanced Engineering Mathematics, 5th Ed., Thomson, Brooks/Cole, 2003.
Pipes, LA and Harvill, LR, Applied Mathematics for Engineers and Physicists, McGraw-Hill, 1970.
Potter, Merle C, Goldberg, JL, and Aboufadel, Edward F, Advanced Engineering Mathematics, 3rd
Ed., Oxford University Press, 2005.
Ross, SL, Introduction to Ordinary Differential Equations, 4th Ed., John Wiley, 1989.
Wylie, CR and Barrett LC, Advanced Engineering Mathematics, 5th Ed., McGraw-Hill, 1982.
B-1
Index
A Cauchy’s mean value theorem 3-9
Cauchy’s nth root test 9-20
absolute convergence 9-35
chain rules for partial differentiation 4-5
algebra of derivatives 10-2 change of scale property 12-34
alternating series 9-35 column 2-55, 2-57
angle of contingence 5-1 comparison tests
antiderivative 7-1 type I 9-15
applications of integration 7-1 type II 9-16
Cartesian coordinates 7-1 comparison theorem-limit form 9-17
parametric coordinates 7-1 complementary function 2-9
polar coordinates 7-1 concavity, convexity 6-5
applications of ordinary differential equation 1-30 conditional convergence 9-35
law of natural growth and decay 1-31 constrained maxima and minima 4-27
Newton’s law of cooling 1-30 continuity 3-1, 3-3, 3-7, 4-1, 4-2, 4-20, 10-1, 10-7, 11-11
arc length 10-3 vector function 10-1
volume elements 10-20 convergent sequences 9-3
area convergent, divergent and oscillatory series 9-11
between two curves 8-21 convolution functions 10-6
surface 4-24, 7-1, 7-16–22 properties of 10-6
under a curve 8-21 critical damping 2-47
astroid 6-23 critical points 4-23
asymptotes curvature
inclined (oblique) 6-25 centre of 5-13
of a curve 6-4 measure of 5-1
parallel to the axes 6-24 of a circle 5-1
radius of 5-2
B curve
Bernoulli’s equation 1-28 closed 11-3, 11-21, 11-22
Bernoulli-Euler law 2-54 concave downward 6-5, 6-21
bounded sequences 9-2 concave upward 6-5, 6-21
in simply connected domain 11-1, 11-3
intersection with coordinate axis 6-4, 6-6, 6-8, 6-21
C intersection with the axes 6-8, 6-9, 6-16, 6-21
cantilever 2-55 passing through the origin 6-3
Cauchy–Maclaurin integral test 9-26 region or extent of 6-5
Cauchy’s condensation test 9-27 rising and falling 6-5
Cauchy’s general principle of convergence 9-7 symmetry about 6-1, 6-3
first theorem on limits 9-8 curve tracing 6-1
second theorem on limits 9-8 Cartesian equation 6-1
I-1
2 Index
curve tracing (Cont’d) differential operators 10-9

cycloid 6-24 gradient of a scalar point function 10-8
cycloid (inverted) 6-22 level surface 10-8
ellipse 6-23 physical interpretation of divergence of a vector
equiangular spiral 6-23 field 10-10
hyperbolic or reciprocal spiral 6-14, 6-15 potential 10-9
hypocycloid 6-14, 6-23 scalar 10-9
lemniscate of Bernoulli 6-13 sets of points 10-6
parabola with its axis along the x-axis and touching vector 10-9
the y-axis 6-22 Dirac delta function 12-22
parabola with its axis along the x-axis and vertex 6-23 Laplace transform of 12-23
parametric coordinates 6-20 divergent sequence 9-4
Pascal’s limacon 6-17 oscillatory sequence 9-4
polar coordinates 6-13 double integral 8-1
rectangular hyperbola 6-22 calculation of 8-2
spiral of Archimedes 6-13 geometrical interpretation 8-1
three-leafed rose 6-13, 6-18, 6-19
tractrix 6-24 E
curvilinear coordinates 10-19
curl in 10-23 ellipse 6-23
divergence 10-23 envelopes 5-15
expressions for the gradient, divergence and curl in equi-dimensional equation 2-31
terms 10-21 method of obtaining an envelope 5-16
gradient 10-22 equi-angular spiral 6-23
orthogonal curvilinear coordinates 10-20 Euler’s theorem on homogeneous functions 4-9
transformation of coordinates 10-20 extension of 4-10
Euler-Cauchy equations 2-31
D evolute involute 5-13
exact equations 1-5, 1-13
D’Alembert’s ratio test 9-22 procedure for solving 1-14
damped oscillations 2-45, 2-46 exactness and independence of path 11-7
in electrical circuits 2-46 criterion for 11-7
in mechanical systems 2-45 exponential shift 2-8
De Morgan–Bertrand test 9-33 extrema 4-23
deflection of beams 2-44, 2-54 necessary and sufficient conditions for 4-23
deflection curve 2-54
elastic curve 2-54
F
neutral surface 2-54
derivability 10-1 filtering property of Dirac delta function 12-23
differentiability 3-1 final value theorem 12-19
differential equations 1-1 first degree differential equations 1-4
degree of 1-2 continuity 3-1
formation of 1-2 differentiability 3-1
general (complete) solution of 1-4 first order linear equation 1-5
linear 1-2 first shifting theorem 12-5
non-linear 1-2, 1-3 applications of 12-6
order of 1-2 flux across a surface 11-11
ordinary 1-1 forced oscillation with damping 2-49, 2-50
partial 1-1 in electrical systems 2-50
particular solution 1-4 in mechanical systems 2-49
singular solution 1-4 forced oscillation without damping 2-47, 2-48
solution of 1-2 in electrical systems 2-48
Index 3
in mechanical systems 2-47 inverse transforms 12-32

free oscillations 2-44, 2-45 derivatives 12-35
in a spring 2-44 integrals 12-36
in electrical circuits 2-45 the method of partial fractions 12-32
in mechanical systems 2-44 irrotational vector point function 10-16
function of several variables 4-1
function of two variables 4-1 J
continuity of 4-1
limit of 4-1 Jacobian 4-13
functional dependence 4-15 of implicit functions 4-15
properties of 4-14
standard 4-15
G
Gauss’s divergence theorem 11-16 L
some deductions from 11-17
geometric series 9-11 Lagrange’s mean value theorem 3-5, 3-6–8
gravity 8-25, 8-32, 8-36 geometrical interpretation 3-6
Green’s theorem 11-21 method of multipliers 4-27
vectorial form 11-24 Laplace transform 12-1
change of scale theorem 12-10
division by t 12-17
H
exponential order 12-9
Heaviside’s function 12-45 integrals 12-13
higher order derivatives 10-2 inverse 12-5
homogeneous equations 1-7, 2-2 linearity 12-2
fundamental theorem for 1-3 multiplication by t 12-16
superposition or linearity principle 2-2 of derivatives 12-11
homogeneous function 1-7, 4-1, 4-8 piece-wise continuous function 12-3
hyper harmonic series 9-14 Legendre’s linear equation 2-35, 2-36
hypocycloid 6-23 Leibnitz test 9-35
line integral 11-1–4, 11-21, 11-25, 11-26
I linear dependence and linear independence of
equations 2-4
identities 10-9, 10-10, 10-14 linear differential equation 1-1, 1-24, 2-1, 2-2, 2-7–9,
improper integral 9-26 2-32, 2-41, 2-44, 12-1, 12-40, 12-43
independence of path 11-3, 11-6, 11-7 applications of 2-44
inexact equation 1-5, 1-17 with constant coefficients 2-9
infinite series 9-10 with variable coefficients 2-31
telescoping series 9-10 linear equation 1-24
initial value problem 2-3 first order linear equation 1-5, 1-24
initial value theorem 12-19 linear operator 2-8
integral in special cases 2-17 linearity principle 2-2
line 11-1–4, 11-21, 11-25, 11-26 logarithmic test 9-31
multiple 8-1
surface 11-11
M
triple 8-14, 8-17
integrating factor 1-17 Maclaurin’s series 3-12
integration theorem 3-12
around closed curves 11-6 mass of a plane lamina 8-25
order of 8-8–10, 8-16, 8-17, 12-17, 12-24 solid 8-25
intrinsic equation 5-2 maxima and minima 4-19
inverse operator 2-11 of functions of two variables 4-22
4 Index
mean value theorem 3-1, 3-5–11, 12-23 polar coordinates 8-12

Cauchy’s mean value theorem 3-9 properties of continuous functions 3-1
constrained 4-27 properties of series 9-13
generalized mean value theorem 3-11
Lagrange’s mean value theorem 3-5 Q
measure of curvature 5-1
method of proportions 1-12 quadrature 8-21
moment of inertia 8-25 exponential curve 8-22
of a solid 8-25 Neil’s parabola 8-22
rectangular hyperpola 8-22
plane lamina 8-25
transcendental curve 8-23
monotonic sequence 9-5
monotone convergence theorem 9-5
monotonic decreasing sequence 9-5 R
monotonic increasing sequence 9-5 Raabe’s test (higher ratio test) 9-29
multiple integrals 8-1 radius of curvature
centre of mass of a solid 8-25 at the origin 5-11
lemniscate of Bernoulli 8-33 Cartesian formula 5-3
mass of a solid 8-25 formula 5-2
moment of inertia 8-25 of a pedal curve 5-10
moment of inertia of a plane lamina 8-25 parametric form 5-5
physical applications of 8-22, 8-31 polar form 5-7
theorem of parallel axis 8-25 real sequence 9-1
theorem of perpendicular axis 8-25 alternating sequences 9-2
bounded sequence 9-2
N Fibonacci sequence 9-2
Riemann sums 7-2
Newton’s law of cooling 1-30 properties of 7-3
non-homogeneous equations 1-9, 2-3 rising and falling curves 6-5
Rolle’s theorem 3-1
O geometrical interpretation of 3-3
oblique asymptotes 6-4
order completeness axiom 9-5 S
order of integration 8-8
orthogonal coordinate systems 10-21 saddle point 4-23, 4-24, 4-26
trajectories 1-33 scalar and vector point functions 10-7
second shifting theorem 12-20
semi cubical parabola 6-1, 6-7, 7-5
P
series of positive terms 9-14
parabola 8-23 short method for find the particular integrals 2-17
parametric coordinates 7-1 simple harmonic motion 2-44
partial differential coeffieient 4-2 simply connected domain 11-1, 11-3
higher order 4-3 solenoidal vector point function 11-12
particular integral 2-12 solid of revolution 7-8-12, 7-15, 7-16, 7-19
method of finding 2-11 special orthogonal coordinate systems 10-21
pedal equation 5-9 spherical coordinates 10-21, 10-22, 10-24, 10-26
periodic function 12-29 Stokes’ theorem 11-26
saw-tooth wave 12-29 Strut 2-55–57
square wave 12-29 surface area of solid of revolution 7-16
triangular wave 12-29 surface integrals 11-11
physical applications of 8-31 definition of 11-11
Index 5
T work done by a variable force 11-3

calculus 10-1
Taylors’ series 3-1, 3-11, 3-12 curves in a simply-connected domain 11-1
for a function of two variables 4-19 differential calculus 10-1
theorem 3-11, 4-1 differential operator 10-6
tests for convergence and divergence of series 9-14 field 10-11
total differential cooefficient 4-5 identities 10-14
tracing of curves 6-12 independence of path: conservative field and scalar
Cartesian curves 6-12 potential 11-3
polar curves 6-12 integral calculus 11-1
transcendental curve 8-23 limit of a vector function 10-1
transformation between line and surface integral 11-26 line integral 11-1
transformation between surface inegrals and volume of constant direction 10-4
integrals 11-16 of constant magnitude 10-4
transformation of coordinates 8-11, 8-18, 8-26, 8-30 point function 10-2
cylindrical coordinates 8-18
scalar differential operator 10-9
spherical polar coordinates 8-12
vector function of a scalar variable 10-1
triple integrals 8-17
volume integral 11-15
volumes of solids of revolution 7-8
U Cartesian coordinates 7-9
unit impulse function 12-22 polar coordinates 7-10
unit step function 12-9, 12-20, 12-22, 12-44
unit vectors in curvilinear system 10-20 W
V working rule for oblique coordinates 6-25

Wronskian determinant 2-4, 2-8
variables separable equations 1-5
variation of parameters 2-37
Z
vector
applications 11-3 zero test 9-13

E. Rukmangadachari - Engineering Mathematics - Volume - 1-Pearson Education (2009)

Uploaded by

Copyright:

Available Formats

E. Rukmangadachari - Engineering Mathematics - Volume - 1-Pearson Education (2009)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

E. Rukmangadachari - Engineering Mathematics - Volume - 1-Pearson Education (2009)

Uploaded by

Copyright:

Available Formats

Engineering

Chennai • Delhi • Chandigarh

Laser typeset by MacroTex Solutions, Chennai.

Printed in India at Chennai Micro Prints Pvt. Ltd., Chennai.

UNIT II FUNCTIONS OF SINGLE VARIABLE

UNIT III APPLICATION OF SINGLE VARIABLES

UNIT V DIFFERENTIAL EQUATIONS OF FIRST ORDER AND THEIR APPLICATIONS

UNIT VI HIGHER ORDER LINEAR DIFFERENTIAL EQUATIONS AND THEIR APPLICATIONS

UNIT VII LAPLACE TRANSFORM AND ITS APPLICATIONS TO ORDINARY DIFFERENTIAL

UNIT VIII VECTOR CALCULUS

2. Linear Differential Equations of 4. Functions of Several Variables

4.6 Chain Rules for Partial Exercise 5.5 5-20

8.5 Multiple Integrals— 9.20 Cauchy’s nth Root Test 9-20

Exercise 11.3 11-20 12.8 Periodic Functions 12-29

If we differentiate Eq.(1.1) n times with respect to Example 1.6

are particular solutions of the second order and ﬁrst

and if we write Example 1.11

EXERCISE 1.1 (b) ex tan y dx + (1 − ex ) sec2 y dy = 0.

(a) y = ax + bx2 (a, b). (d) 1 − x2 sin−1 x dy + y dx = 0.

(i) Form the differential equation of all parabo- dy  

  dy where v = y/x (or v = x/y) and c is an arbitrary

Example 1.14 Example 1.18

Separating the variables 2. xdy − ydx = x2 + y2 dx.

where c is an arbitrary constant. 6. (x2 − y2 )dx = xydy.

dy y Integrating, we obtain the general solution as

Solution Here Separating the variables,

Since a = m = 0, the variables are separable. a = 1, b = −2, c = 4

Put u = 3x + y Solution Here,

Separating the variables and grouping the terms du

(x − 2)dx + ydx + xdy − (y − 4)dy = 0 Integrating, we get

x2 + 2xy − y2 − 4x + 8y + C = 0. (4x + 8y + 1)5 = Ce(4x−8y) .

(2x + 3)dx − 4(ydx + xdy) + (8y − 1)dy = 0 Put x = X + h, y = Y + k

dv 2 + 5v This method applies if l = b and m = a.

Alternative method (Method of proportions) dy 2x − y + 1

dy x+y+1 Proof: (i) The condition is necessary: Suppose

y 1 1 Alternative method: Grouping the terms

The equation is exact since ∂M /∂y = ∂N /∂x ∂v

Example 1.34 The equation is exact.

Example 1.35 Example 1.37

Integrating, we get the general solution as y(xy + ex )dx − ex dy

2. (3x2 − 9x2 y2 + 2xy)dx

1 derive the general solution as follows:

dx dy     3. xdy + 2ydx = 2y2 xdy

ydx − xdy 2 5. 2xydy − y2 dx + x3 ex dx = 0

Multiplying by the integrating factor, Solution This equation is of the form

Solution The given differential equation is Example 1.48

Exactness condition is which is the required solution.

Example 1.49 ⇒ (k + 2)xh+1 yk+1 + 2(k + 3)xh+2 y(k+2)

Now Eq.(1) becomes

Equating My = Nx , we have EXERCISE 1.6

k = h + 2 k + 3 = −h Solve the following equations by ﬁnding integrating

7. (x2 + y2 + 2x)dx + 2ydy = 0 and comparing it with

8. (y + log x)dx − xdy = 0 we have M = Py − Q, N = 1

1.3.6 Linear Equations ye pdx = c + Qe pdx dx (1.22)

(3) Write the general solution of the  differen- which is linear in x.

Po y = c + Qdx. (1.25) = elog x = x

Example 1.51 Multiplying the equation by x, we have

where c is an arbitrary constant. Example 1.54

This is a linear equation in y Example 1.56

(i) Form the differential equation of all parabo- dy

dy where v = y/x (or v = x/y) and c is an arbitrary

dx dy 3. xdy + 2ydx = 2y2 xdy

(3) Write the general solution of the differen- which is linear in x.

Integrating factor = e Pdx = elog(log x) 1 1 −1

involving two arbitrary constants c1 and c2 . An ini- y1 y2

where V is a function of x. Changing D to D + α in [∵ emx = 0]

The auxiliary equation is

2 is a function of x containing no arbitrary constant

1 3x To ﬁnd the particular integral,

P.Is in special cases when Q(x) is of the form