Mathematics For Mechanical Engineers
Uploaded by
Minh Nguyễn AnhMathematics For Mechanical Engineers
Uploaded by
Minh Nguyễn AnhMATHEMATICS FOR
MECHANICAL ENGINEERS
NGUYEN VIET KHOA
I
PREFACE
Science and mathematics are integral parts of engineering. Science
provides the laws of the natural world and describes them as
relationships between different quantities using mathematical
formulas. Mechanical engineering is all about applying science and
mathematics to solve problems we face in our daily lives such as
mechanics of solids, mechanics of fluids, mechanics of structures and
so on. This text book is written to provide mathematical knowledge to
meet the needs of engineering students in the early years of mechnical
engineering degrees provided they have learned calculus I and II.
However, the book will also be valuable to a much wider readership
including those on other courses, e.g. engineering diplomas, and those
who already have some basic knowledge of mechanics, but wish to
explore more advanced topics, such as works done by variable forces,
circulations of fluid in vector fields, vibrations of structures.
The book contains ten chapters. Scalar fields and vector fields are the
subjects of Chapter 1, as they constitute the knowledge of the vector
and scalar fields that is very important tool for describing many
physical concepts such as gravitation, velocity fields of wind and fluid
which affect the behavior of objects over a large region of space. This
chapter presents the basic knowledge of vectors, vector fields and
their properties. The vector fields will be applied to form engineering
problems, e.g. to define work done by a force field in moving an
object along a curve as line integrals, to define the rate of fluid flow
across as surface integrals.
Chapter 2 presents the important operators that can be performed on
scalar and vector fields. This refers to, for instance, engineering,
physics, and computer sciences, in general, but particularly solid
mechanics, aerodynamics, aeronautics, fluid flow, heat flow, robotics
as well as other areas have applications that require an understanding
of vector calculus.
Multiple integrals constitute the object of Chapter 3. The double and
triple integrals and their applications for computing volumes, masses,
and centroids of more general regions are presented in this chapter.
We will learn how to apply different types of coordinates such as
polar coordinates, cylindrical coordinates and spherical coordinates
II
for calculating multiple integrals over specific regions depending on
types of integration regions.
Chapter 4 presents differential forms as an alternative to vector
calculus which is ultimately simpler and more flexible. In this chapter,
a brief supplement is provided to explain how to work with them.
Chapter 5, devoted to define the line integral in vector fields. In
Chapter 4 the line integral using the differential form presents a
coordinate free way to deal with general principles. However, in many
physical circumstances, equations need to be expressed on
coordinates. The line integral in vector fields, therefore, is presented
in this chapter.
Chapter 6 presents the surface intefral which describes the flow of a
field through a surface. This chapter first gives the representations of
parametric surface and show us how to parametrize surfaces to give
the convenient ways of computing surface integrals. Then, we will
learn how to integrate over different types of surfaces. Finally, we will
apply the general surface area formula to special surfaces.
Chapter 7 presents important theorems which give the relationships
between line, surface and multiple integrals. Green-riemann theorem
relates a line integral around a simple closed curve to a double integral
over the plane region. Stoke's theorem establish the relations between
the integral of a vector field along a curve and the flux of that field
over a surface. Frean-Ostrogradski relates a surface integral of a field
to a triple integral over a region bounded by a surface.
Chapter 8 presents Fourier series of an arbitrary function. Fourier
pointed out that an arbitrary function can be represented by a
trigonometric series. Fourier series is advantageous than a power
series in many phenomenons that are periodic such as astronomical
phenomena, heartbeats, tides, vibration.
The Fourier transform is the topic of Chapter 9. Fourier transform is
useful in many fields such as partial differential equation, signal and
image processing, LTI system & circuit analysis. In this chapter the
relationship between the Fourier series and Fourier transform will be
presented. The properties of the Fourier transform are provided and
some of its applications are then explored.
III
Chapter 10 presents the Laplace transform as well as the inverse
Laplace transform. The Laplace transform has many applications in
science and engineering because it is a tool for solving differential
equations. The application of the Laplace transform for solving
ordinary differential equations is demonstrated. Finally, some
engineering problems are presented and solved by using Laplace
transforms.
Each chapter provides some examples as solved problems to clarify
the introduced concepts. Abstraction discussion are avoided and the
mechanical application and illustrations are given. Some appropriate
proofs are provided to help student understand the theorems. Some of
the example are taken from the recent literature to illustrate the
applications in the field of mechanical engineering. At the end of each
chapter, there are exercises to help students better understand the
contents of the chapter and improve the calculating techniques. To
yield the good knowledge from the book, students need to read the
text carefully and do the exercises as much as they can. Some answers
or hints are given to help student do the exercises effectively.
Students can seek more materials beyond the book from a list of
references provided at the end of the book.
Acknowledgments
I would like to thank Thao Thi Bich Dao for her careful reading the
text, checking the examples, and drawing and modifying some of the
figures to make it readable.
CONTENTS
INTRODUCTION ............................................................................... I
CHAPTER 1. SCALAR FIELDS, VECTOR FIELDS ................ 1
1.1 Scalar Fields............................................................................... 2
1.1.1 Definition of Scalar Field ................................................ 2
1.1.2 Representations of Scalar Fields...................................... 3
1.2 Vectors ....................................................................................... 6
1.2.1 Adding Vectors ................................................................ 6
1.2.2 Triangle Law.................................................................... 7
1.2.3 Parallelogram Law ........................................................... 7
1.2.4 Scalar ............................................................................... 8
1.2.5 Subtracting Vectors ......................................................... 8
1.2.6 Components of Vectors in 2- or 3-Dimensional.............. 8
1.2.7 Length of Vector ............................................................ 10
1.2.8 Adding Algebraic Vectors ............................................. 10
1.2.9 Multiplying Algebraic Vectors ...................................... 11
1.2.10 3-D Algebraic Vectors ................................................. 11
1.2.11 Components of n-Dimensional Vectors ...................... 12
1.2.12 Properties of Vectors ................................................... 12
1.2.13 Unit Vector .................................................................. 12
1.2.14 Dot Product .................................................................. 13
1.2.15 Direction Angles and Cosines ..................................... 15
1.2.16 Projections ................................................................... 16
1.2.17 Cross Product............................................................... 18
1.2.18 Cross Product in Physics ............................................. 20
1.2.19 Equations of Lines and Planes ..................................... 21
1.2.20 Scalar Equations of a Line ........................................... 22
1.2.21 Equation of Line Segments.......................................... 25
1.2.22 Equation of Planes ....................................................... 26
1.2.23 Parallel planes .............................................................. 28
1.2.24 Nonparallel planes ....................................................... 28
1.2.25 Distance from a Point to a Plane ................................. 30
1.3 Vector Fields ............................................................................ 32
1.3.1 Vector Functions............................................................ 32
1.3.2 Limit of a Vector Function ............................................ 32
1.3.3 Continuous of Vector Function and Space Curve ......... 33
1.3.4 Derivatives of Vector Functions .................................... 34
1.3.5 Integrals of Vector Functions ........................................ 35
1.3.6 Vector Fields.................................................................. 36
1.3.7 Vector Fields on 2 ...................................................... 36
1.3.8 Vector Fields on 3 ....................................................... 38
1.4 Velocity Field .......................................................................... 38
1.5 Gravitational Field ................................................................... 40
1.6 Conservative Vector Fields...................................................... 42
CHAPTER 2. GRADIENT, DIVERGENCE, ROTATIONAL
AND LAPLACIAN OPERATORS ..................... 53
2.1 Gradient Operator .................................................................... 53
2.1.1 Partial Derivative ........................................................... 53
2.1.2 Gradient Operator .......................................................... 56
2.1.3 Directional Derivative ................................................... 57
2.1.4 Tangent Planes to a Level Surfaces ............................... 60
2.1.5 Normal Line ................................................................... 61
2.2 Curl Operator ........................................................................... 62
2.2.1 Conservative Vector Fields ........................................... 64
2.2.1 Physical Meaning of Curl .............................................. 66
2.3 Divergence Operator ................................................................ 67
2.4 Laplace Operator ..................................................................... 68
CHAPTER 3. MULTIPLE INTEGRALS ................................ 73
3.1 Double Integrals over Rectangles ............................................ 73
3.2 Iterated Integrals ...................................................................... 78
3.3 Double Integrals over General Regions ................................... 79
3.3.1 Type I Region ................................................................ 80
3.3.2 Type II region ................................................................ 82
3.4 Double Integrals in Polar Coordinates ..................................... 89
3.5 Triple Integrals......................................................................... 97
3.5.1 Definition of Triple Integrals ......................................... 97
3.5.2 Integral Over a General Bounded Region ................... 100
3.5.3 Applications of Triple Integrals ................................... 107
3.6 Triple Integrals in Cylindrical Coordinates ........................... 110
3.7 Triple Integrals in Spherical Coordinates .............................. 114
3.8 Change of Variables in Multiple Integrals ............................. 118
3.8.1 Change of Variables in Double Integrals .................... 118
3.8.2 Change of Variables in Triple Integrals ...................... 130
CHAPTER 4. DIFFERENTIAL FORMS ...............................139
4.1 1-Forms .................................................................................. 139
4.2 Exactness in R2 ...................................................................... 140
4.3 2-Forms .................................................................................. 148
4.4 Exactness in R3 ...................................................................... 150
4.4.1 Convert a Vector Field to a 2-Form............................. 151
4.4.2 Derivative of 1-Forms ................................................. 152
4.4.3 Curl and Derivative of 1-Forms................................... 152
4.4.4 Derivative of 2-Forms ................................................. 153
4.4.5 Surface Integrals .......................................................... 154
CHAPTER 5. LINE INTEGRALS AND CIRCULATION OF A
FIELD .............................................................165
5.1 Line Integrals ......................................................................... 165
5.1.1 Line Integrals in 2-D Coordinates ............................... 165
5.1.2 Interpretation of a Line Integral................................... 167
5.1.3 Piecewise-smooth Curve ............................................. 168
5.1.4 Mass ............................................................................. 169
5.1.5 Curve Orientation ........................................................ 173
5.1.6 Line Integrals in 3-D Coordinates ............................... 175
5.1.7 Line Integrals of Vector Fields .................................... 178
5.1.8 Line Integral in Differential Forms and Vector Fields 165
5.2 Fundamental Theorem for Line Integrals .............................. 183
5.3 Line Integrals as Circulation.................................................. 186
5.4 Physical Interpretation of Line Integrals ............................... 192
CHAPTER 6. FLOW OF A FIELD THROUGH A SURFACE 203
6.1 Parametric Surfaces ............................................................... 203
6.1.1 Tangent Planes ............................................................. 211
6.1.2 Smooth Surface............................................................ 211
6.1.3 Surface Area ................................................................ 213
6.2 Surface Integral ...................................................................... 218
6.3 Flow of a Field Through a Surface ........................................ 227
CHAPTER 7. GREEN, STOKE AND OSTROGRADSKI
THEOREMS....................................................239
7.1 Green Theorem ...................................................................... 239
7.1.1 Reverse Direction ........................................................ 244
7.1.2 Regions with Holes ...................................................... 247
7.2 Stokes Theorem ..................................................................... 249
7.3 Ostrogradski Theorem ........................................................... 257
CHAPTER 8. FOURIER SERIES ..........................................271
8.1 Fourier Series Definition ....................................................... 271
8.2 Lemma ................................................................................... 271
8.3 Fourier Convergence Theorem .............................................. 272
8.4 Functions with Period 2L....................................................... 279
8.5 Complex form of Fourier series ............................................. 284
8.6 Complex form of Fourier series ............................................. 286
CHAPTER 9. FOURIER TRANSFORM ................................299
9.1 From Fourier Series to Fourier Transform ............................ 273
9.2 Theorems of Fourier Transform ............................................ 301
9.3 The Sampling Theorem ......................................................... 313
9.4 Discrete Fourier Transform ................................................... 315
9.4.1 Forward Transform ...................................................... 317
9.4.2 Inverse Transform........................................................ 318
9.6 Spectrum Analysis ................................................................. 319
CHAPTER 10. LAPLACE TRANSFROM .............................327
10.1 Definition, Basic Properties ................................................. 327
10.2 Further Laplace Transform Theorems ................................. 334
10.3 The Inverse Laplace Transform ........................................... 338
10.4 Laplace Transform for Solving Ordinary Differential
Equations .............................................................................. 344
10.4.1 First Order Differential Equations ............................. 345
10.4.2 Second Order Differential Equations ........................ 348
10.4.3 Some Applications of Laplace Transform to Mechanical
Engineering........................................................................... 349
REFERENCES .....................................................................371
Scalar and vector fields 1
CHAPTER 1. SCALAR FIELDS, VECTOR FIELDS
Engineering problems are not only the physical sciences but they are
also mathematical sciences. The underlying concepts and principles of
engineering problems have a mathematical basis. In daily life we can
describe things in nature by words. For example, to describe the
motion of objects we can use words and phrases such as going fast,
stopped, slowing down, speeding up, and turning. In engineering we
will be expanding upon this vocabulary list with words such as
distance, displacement, speed, velocity, and acceleration. These words
are associated with mathematical quantities that have strict definitions.
The mathematical quantities that are used to describe the motion of
objects can be divided into two categories. The quantity is either a
vector or a scalar. These two categories can be distinguished from one
another by their distinct definitions:
Scalars are quantities that are fully described by a magnitude
(or numerical value) alone.
Vectors are quantities that are fully described by both a
magnitude and a direction.
This chapter will give the definitions and examples of scalar and
vector quantities. As you proceed through this text book, give careful
attention to the scalar and vector nature of each quantity. This is very
important for engineers to understand the nature of engineering
problems. The concepts of scalar fields are presented first since it is
familiar with the students. The vector fields will be presented later in
details because it will provide a new mathematical presentation of
engineering problems.
The knowledge of the vector fields is very important tool for
describing many physical concepts such as gravitation and
electromagnetism, which affect the behavior of objects over a large
region of space. This chapter presents the basic knowledge of vectors,
vector fields and their properties. This will be the basis of vector
calculus in next chapters. The vector fields will be applied to form
engineering problems, e.g. to define work done by a force field in
moving an object along a curve as line integrals, to define the rate of
fluid flow across as surface integrals. The line and surface integrals
together with the single, double, and triple integrals will provide the
2 Mathematics for mechanical engineers
higher-dimensional versions of the Fundamental Theorem of
Calculus: Green’s Theorem, Stokes’ Theorem, Divergence Theorem,
which will be presented in the next chapters.
1.1 Scalar Fields
1.1.1 Definition of Scalar Field
Definition 1.1: A scalar is defined to be any real number c which is a
quantity that has a magnitude alone. A scalar field is an assignment of
a scalar to each point in region in the space.
Figure 1.1. Temperatures at all points on the earth surface
For example, the temperatures at all points on the earth is a scalar
field. The Figure 1.1 shows the image of daytime temperatures of
Earth's surface on August 18, 2021 at 11:00 AM. The coldest
temperature, shown in indigo, is −52.2°C, while the hottest
temperature, shown in purple, is 40.9° C. The view is centered on Ho
Chi Minh City (10N, 106E), which is 35.5° C. We can see very hot
Scalar and vector fields 3
climate regions in Middle East, Africa where is covered by warm
material, showing a sandy and rocky surface. In this map, warm
climate regions can be observed in Australia and Asia continentals.
While, cold regions on North (blue) and South Poles (indigo), which
are covered by ice during the year round so the temperatures are
extremely low.
In this map, the surface temperature is described by various colors.
However, this temperature picture is presented in only two-
dimensional surface. Therefore, it does not show how temperature
varies as a function of altitude. In practice, a scalar field provides
values not only on a two-dimensional surface in space but for every
point in space. So in principle, we need to create a three-dimensional
atmospheric volume element and color it to represent the temperature
variation.
A simply way to represent the temperature picture of an object is
using a mathematical function. For example, if we want to describe
the temperature of Earth we shall use spherical coordinates where the
origin is at the center of the Earth as shown in the Figure 1.2. The
temperature at any point is
characterized by a function
T r , , where r , , are
spherical coordinates. In other
words, the value of this function at
point r , , is the temperature
of Earth at that point. The
temperature function T r , , is
an example of a “scalar field”. In
this example, the term “scalar”
implies that temperature at any
point is a number rather than a Figure 1.2. Sphere coordinates
vector (a vector has both
magnitude and direction).
1.1.2 Representations of Scalar Fields
As stated, a scalar field is a function that has a different value at every
point in space, thus it can be represented by an explicit function. For
4 Mathematics for mechanical engineers
example, the following equation defines the value of the scalar
function f at arbitrary point x, y , z in the rectangular coordinates:
1 2
f x, y , z (1.1)
x2 y d z 2 x2 y d z2
2 2
However, if we just look at Equation 1.1, it is not easy to imagine how
the filed can be seen actually. Thus, we need to use other presentations
for scalar fields which can give a better way to imagine the scalar
fields. Below we give three possible representations.
1.1.2.1 Contour Maps
The first way to a presentation is to use the contour map. In this way
we need to fix one of independent variables (z, for example) and then
show a contour map for the two remaining dimensions. In contour
map, each curve represents a line of constant values of the function f.
We show such a contour map in the xy-plane at z = 0 for Equation 1.1.
Figure 1.3. Contour map
1 2
f x, y,0
x2 y d x2 y d
2 2
Scalar and vector fields 5
The Figure 1.3 shows a contour map consisting of 100 contour levels
with d =0.5. A series of these maps for various (fixed) values of z will
give a feel for the properties of the scalar function.
1.1.2.2 Color-Coding
Another way we can represent the values of the scalar field is by
color-coding in two dimensions for a fixed value of the third. This was
the scheme used for illustrating the temperature fields in the above.
Figure 1.4 presents a color map for the scalar field f x, y , 0 when z
is set to zero. Different values of f x, y , 0 are characterized by
different colors in the map.
Figure 1.4. Color coding map
1.1.2.3 Relief Map
A third way to represent a
scalar field is to fix one of the
dimensions, and then plot the
value of the function as a
height versus the remaining
spatial coordinates. Figure
1.5 shows such a relief map
for the same function
Figure 1.5. Relief map
6 Mathematics for mechanical engineers
f x, y , 0 where z is fixed to zero while x and y change.
1.2 Vectors
In engineering problems, vector is used to indicate a quantity that has
both magnitude and direction (such as acceleration, velocity,
displacement or force). A vector is in general represented by an arrow
or a directed line segment. The length of the arrow represents the
magnitude of the vector, while the arrow points in the direction of the
vector. In this text book, we denote a vector by either: printing a letter
in boldface, for example v, or putting an arrow above the letter, for
example v .
B
Figure 1.6 presents a vector which
describes the velocity vector v of an
object moving from initial point A
(the tail) to terminal point B (the tip).
A
We can write this vector as follows:
Figure 1.6. A vector
v AB
If the vector u CD has the same length and the same direction as v
even though it is in a different position then we say u and v are equal
and write u v as shown in Figure 1.7.
B D
A C
Figure 1.7. Two equal vectors
1.2.1 Adding Vectors
AM and then
Assume that the object moves from A to M with vector
changes the direction, moves from M to B with vector MB . The final
Scalar and vector fields 7
result of combination of these displacements is that the object has
moved from A to B as depicted in
Figure 1.8. B
The
resulting displacementvector
AB is called the sum of AM and
MB . We write:
M
A
AB AM MB
Figure 1.8. Adding vectors
1.2.2 Triangle Law
The above law for adding two vectors
is sometime called the Triangle Law. B
Figure 1.9 will explain the reason why
they call this name. As can be seen
from this picture vector AB
corresponding to the side AB of the A
triangle ABC has the same initial and M
terminal points with the combination of
two vectors AM MB which
Figure 1.9. Triangle Law
correspond to two other sides of the
triangle AM and MB.
1.2.3 Parallelogram Law
Presenting vectors u and v
and draw another copy of v
with the same initial point as
u and completing the
parallelogram as shown in
Figure 1.10, we see that:
u v vu
This is called the
Parallelogram Law. Figure 1.10. Parallelogram Law
8 Mathematics for mechanical engineers
1.2.4 Scalar
In this context, we call the real
number c a scalar to distinguish it
from a vector. For example, we
want to describe a vector 2v to be
2v
the same as vector v v , which has
the same direction as v but is twice v
as long. In general, we multiply a 0.5v
vector by a scalar as follows. If c is
a scalar and v is a vector, the scalar
-v -1.5v
multiple cv is the vector whose
length is |c| times the length of v
and whose direction is the same as
v if c 0 and is opposite to v if
c 0 . Look at Figure 1.11, we see
that real numbers work like scaling Figure 1.11. Scalar multiplies of v
factors here. That’s why we call
them scalars.
u
1.2.5 Subtracting Vectors v
u-v
The difference u – v of two vectors
implies that:
-v
u – v = u + (–v)
To describe this visually, we can Figure 1.12. Subtracting vectors
build up u – v by first plotting the
negative of v, –v, and then adding it to u by the Parallelogram Law as
shown in Figure 1.12.
1.2.6 Components of Vectors in 2- or 3-Dimensional
Let’s put the initial point of a vector a at the origin of a rectangular
coordinate system. Then, the coordinates of terminal point of a can be
written in the form a1 , a2 or a1 , a2 , a3 . This depends on whether the
coordinate system is two- or three-dimensional as illustrated in Figure
1.13.
Scalar and vector fields 9
y
z
(a1, a2)
a2
(a1, a2, a3)
a
a
O
O a1 x
x
y
a = 〈a1, a2〉 a = 〈a1, a2, a3〉
Figure 1.13. Components of
These coordinates are called the components of a and we write:
a a1 , a2 or a a1 , a2 , a3
The notation a1 , a2 is used to distinguish the ordered pair that refers
to a vector and the ordered pair a1 , a2 that refers to a point in the
plane. Now, consider vector a with the initial point is A x1 , y1 , z1 and
the terminal point is B x2 , y2 , z2 . Then, we have:
x1 a1 x2
y1 a2 y2
z1 a3 z2
Thus,
a1 x2 x1
a2 y2 y1
a3 z2 z1
10 Mathematics for mechanical engineers
Therefore, if we have the points A x1 , y1 , z1 and B x2 , y2 , z2 , the
vector a with representation AB is:
a x2 x1 , y2 y1 , z2 z1
1.2.7 Length of Vector
The magnitude or length of the vector v is the length of any of its
representations. It is denoted by the symbol v or v .
The length of the two-dimensional (2-D) vector a a1 , a2 is:
a a12 a22
The length of the three-dimensional (3-D) vector a a1 , a2 , a3 is:
a a12 a22 a32
1.2.8 Adding Algebraic Vectors
Figure 1.14 shows that, if
a a1 , a2 and b b1 , b2 , then y
(a1+b1, a2+b2)
the sum of two vectors is:
a+b a1 b1 , a2 b2 a+b b b2
b1
at least for the case where the a
components are positive. This a2 a2
means that, to add algebraic O a1 b1 x
vectors, we add their
components.
Figure 1.14. Adding Algebraic Vectors
Similarly, to subtract vectors, we subtract components.
a-b a1 b1 , a2 b2
Scalar and vector fields 11
1.2.9 Multiplying Algebraic Vectors
To multiply a vector by a scalar, we multiply each component by that
scalar.
If a a1 , a2 and c is a scalar then
ca ca1 , ca2
1.2.10 3-D Algebraic Vectors
For 3-D vectors, we have similar formulas:
a1 , a2 , a3 b1 , b2 , b3 a1 b1 , a2 b2 , a3 b3
a1 , a2 , a3 b1 , b2 , b3 a1 b1 , a2 b2 , a3 b3
c a1 , a2 , a3 ca1 , ca2 , ca3
Example 1.1: If a 1, 2, 3 and b 2, 0, 2 , find: a and the vectors
a b , a b , 2a, 2a 3b .
Solution: We have:
a 12 22 32 = 14
a + b = 1, 2, 3 + 2, 0, 2
= 1 + ( 2), 2 + 0, 3 + 2
= 1, 2, 5
a b = 1, 2, 3 2, 0, 2
= 1 ( 2), 2 0, 3 2
= 3, 2,1
2a = 21, 2, 3
= 2.(1), 2(2), 2(3)
= 2, 4, 6
12 Mathematics for mechanical engineers
2a 3b = 21, 2, 3 32, 0, 2
= 2, 4, 6 6, 0, 6
= 8, 4, 0
1.2.11 Components of n-Dimensional Vectors
From now on we denote V2 as the set of all 2-D vectors, V3 as the set
of all 3-D vectors. More generally, we will later need to consider the
set Vn of all n-dimensional vectors. An n-dimensional vector is
defined as:
a a1 , a2 , , an
where a1 , a2 ,, an are real numbers that are called the components of
a.
1.2.12 Properties of Vectors
If a, b, and c are vectors in Vn and c and d are scalars, then we have
the following properties of vectors.
1. a + b = b + a 2. a + (b + c) = (a + b) + c
3. a + 0 = a 4. a + ( a) = 0
5. c(a + b) = ca + cb 6. (c + d )a = ca + da
7. (cd )a = c (da) 8. 1a = a
1.2.13 Unit Vector
We define a unit vector as a vector whose length is 1. In general, if a ≠
0, then the unit vector that has the same direction as a is:
1 a
| u | a
|a| |a|
In order to verify this, we let c 1 / a , then u c a where c is a
positive scalar. Therefore, u has the same direction as a. Also:
Scalar and vector fields 13
1
| u || ca || c || a | | a | 1
|a|
In rectangular coordinates, we have i, j, and k are all unit vectors
along the x, y and z axes, respectively.
Example 1.2: Find the unit vector in the direction of the vector
i 2 j 2k
Solution: The given vector has length:
| i 2 j 2k | 12 (2) 2 ( 2) 2 9 3
So, the unit vector with the same direction is:
1 1 2 2
(i 2 j 2k ) i j k
3 3 3 3
1.2.14 Dot Product
Definition 1.2: If a a1 , a2 , a3 and b b1 , b2 , b3 , then the dot
product of a and b is the number a b given by:
a b a1b1 a1b2 a3b3
Although Definition 2 is given for three-dimensional (3-D) vectors,
the dot product of two- dimensional vectors is defined in a similar
fashion:
a b a1 , a2 b1 , b2 a1b1 a1b2
Example 1.3: Given: a 1,3 , b 3, 1 , c 1,3, 2 ,
d 2, 1,1 , e i j 2k and f i 2 j k . Find a b , c d and
ef .
Solution: Applying directly from Definition 2 we have:
14 Mathematics for mechanical engineers
a b 1,3 3, 1 1 3 3 1 0
c d 1,3, 2 2, 2,1 1 2 3 2 2 1 10
e f i j 2k i 2 j k 1 1 1 2 2 1 1
Theorem 1.1: If a, b, and c are vectors in V3 and c is a scalar, then
1. a a=|a|2
2. a b b a
3. a (b c) a b a c
4. (ca) b c(a b) a (cb)
5. 0 a 0
The dot product a b can be given a geometric interpretation in terms
of the angle θ between a and b. This is the angle between vectors a
and b that start at the origin, where 0 . In other words, θ is the
angle between the line segments OA and OB as depicted in Figure
1.15. Note that if a and b are parallel vectors, then 0 or .
B
a-b
b
θ A
O a
x
y
Figure 1.15. Angle between two vector
Theorem 1.2: If θ is the angle between the vectors a and b, then
a b a b cos
Proof: If we apply the Law of Cosines to triangle OAB here, we get:
Scalar and vector fields 15
2 2 2 2 2
AB OA OB 2 OA OB cos a b 2 a b cos
In addition:
AB a-b a b a b a a a b b a b b
2 2
2 2
= a 2a b b
Thus,
2a b 2 a b cos
or
a b a b cos
Corollary 1.1: If θ is the angle between the nonzero vectors a and b,
then
a b
cos
| a || b |
1.2.15 Direction Angles and Cosines
Using the above Corollary 1.1 with b replaced by i, we obtain:
ai a
cos 1 (1.2)
| a || i | | a |
This can also be seen directly from the Figure 1.16.
Similarly, we also have:
a2 a
cos , cos 3 (1.3)
|a| |a|
16 Mathematics for mechanical engineers
By squaring the expressions in Equations 1.2 and 1.3 and adding, we
see that:
a12 a22 a32
cos 2 cos 2 cos 2 1
a
a3
a
γ
β a2
a1
α y
x
Figure 1.16. Direction angles
We can also write:
a a1 , a2 , a3
a cos , a cos , a cos
a cos , cos , cos
Thus,
1
a cos , cos , cos
|a|
This states that the direction cosines of a are the components of the
unit vector in the direction of a.
1.2.16 Projections
Scalar and vector fields 17
Figure 17 depicts representations PQ and PR of two vectors a and b
with the same initial point P. Let S be the foot of the perpendicular
from R to the line containing PQ . Then, the vector with
representation PS is called the vector projection of b onto a and is
denoted by proja b. Then,
PS b cos
Therefore, the equation:
a b a b cos a b cos
shows that the dot product of a and b can be interpreted as the length
of a multiply by the scalar projection of b onto a, i.e.
a b a
| b | cos b
|a| |a|
This means that the component of b along a can be calculated by
taking the dot product of b with the unit vector in the direction of a.
θ S a
P
Q
|b|cos θ = compab
Figure 1.17. Projection of a vector
Example 1.4: Find the work done by the force of 100 N applied to pull
a wagon a distance of 150 m along a horizontal path, provided that the
handle of the wagon is held at an angle of 30° above the horizontal.
18 Mathematics for mechanical engineers
Solution: Assume F and D are the force and displacement vectors, as
shown in Figure 1.18. Then, the work done is the component of force
in the horizontal direction times the displacement:
W F D F D cos 30o
100 150 cos 30 12990 Nm 12990 J
30° F
30°
D
Figure 1.18. A force applied to pull a wagon
1.2.17 Cross Product
Now we learn that unlike the dot product, the cross product a b of
two vectors a and b is a vector. For this reason, it is also called the
vector product. Note that a b is defined only when a and b are three-
dimensional (3-D) vectors.
Definition 1.3: If a a1 , a2 , a3 and b b1 , b2 , b3 , then the cross
product of a and b is the vector
a b a2b3 a3b2 , a3b1 a1b3 , a1b2 a2b1
The reason for the particular form of this definition is that the cross
product defined in this way has many useful properties. In order to
make this definition easier to remember, we use the notation of
determinants. A determinant of order 2 is defined by:
a b
ad bc
c d
A determinant of order 3 can be defined in terms of second-order
determinants as follows:
Scalar and vector fields 19
a1 a2 a3
b b3 b b b b
b1 b2 b3 a1 2 a2 1 3 a3 1 2 (1.4)
c2 c3 c1 c3 c1 c2
c1 c2 c3
Now, let’s rewrite Definition 3 using second-order determinants and
the standard basis vectors i, j, and k. We see that the cross product of
the vectors a a1i a2 j a3k and b b1i b2 j b3k is:
a2 a3 a1 a3 a1 a2
ab i j k (1.5)
b2 b3 b1 b3 b1 b2
In view of the similarity between the Equations 1.4 and 1.5, we can
write:
i j k
a b a1 a2 a3
b1 b2 b3
Theorem 1.3: The vector a b is orthogonal to both a and b.
Proof: In order to show that a b is orthogonal to a, we calculate their
dot product as follows:
a2 a3 a a3 a a2
(a b ) a a1 1 a2 1 a3
b2 b3 b1 b3 b1 b2
a1 (a2b3 a3b2 ) a2 (a1b3 a3b1 ) a3 (a1b2 a2b1 )
a1a2b3 a1b2 a3 a1a2b3 b1a2 a3 a1b2 a3 b1a2 a3 0
We can use a similar calculation
to show that a b b 0 . This
implies that a b is orthogonal
to both a and b. To determine
the direction of vector a b we
use the right hand law as follows
(see Figure 1.19): If the fingers
of your right hand curl in the
Figure 1.19. Right hand law
20 Mathematics for mechanical engineers
direction of a rotation (through an angle less than 180°) from a to b,
then your thumb points in the direction of a b .
Theorem 1.4: If θ is the angle between a and b 0 , then:
a b a b sin
Corollary 1.2: Two nonzero vectors a and b are parallel if and only if:
ab 0
If we apply Theorems 1.3 and 1.4 to the standard basis vectors i, j, and
k using / 2 , we obtain:
i j k, j k i, k i j, j i k, k j i, i k j
Theorem 1.5: If a, b, and c are vectors and c is a scalar, then:
1. a b b a
2. ca b c a b a cb
3. a b c a b a c
4. a b c a c b c
5. a b c a b c
6. a b c a c b a b c
1.2.18 Cross Product in
Physics
Here we give an example of cross
product in physics. If we tighten a
bolt by applying a force to a wrench,
we produce a torque τ (relative to the
origin). The torque is defined to be
the cross product of the position
vector and force vector τ r F . It
presents the ability of the force to Figure 1.20. A bolt is
rotate the body about the origin. The tightened by a wrench
Scalar and vector fields 21
direction of the torque vector indicates the axis of rotation.
According to Theorem 1.5, the magnitude of the torque vector is:
τ r F r F sin
where θ is the angle between the position vector and force vector. As
can be seen from Figure 1.20, vector F can have two components: one
is in the direction of r and another is perpendicular to r. We can
observe that the only component of F that can cause a rotation is the
one perpendicular to r, that is, F sin . The magnitude of the torque
is equal to the area of the parallelogram determined by r and F.
Example 1.5: Find the magnitude of the torque about the center of a
bolt given that the bolt is tightened by applying a 60 N force to a 0.2
m wrench, as shown in Figure 1.21.
Solution: The torque vector is the
cross product of position vector and
force vector r F . Thus, the
magnitude of the torque vector is:
τ r F r F sin 70o
0.2 60 sin 70o Figure 1.21. A bolt is
12sin 70 11.28 N.m
o
tightened by a wrench
1.2.19 Equations of Lines and
z
Planes
P0(x0, y0, z0)
A line L in three-dimensional (3-D) L P(x,y,z)
a
space is determined when we know:
r
A point P0 x0 , y 0 , z0 on L and the r0
direction of L. In three dimensions, O v
the direction of a line is
conveniently presented by a vector. x y
Let v be a vector parallel to L. Let
P x, y , z be an arbitrary point on Figure 1.22. Vectors and
22 Mathematics for mechanical engineers
L. Let r0 and r be the position vectors of P0 and P. That is, we have
representations of two vectors OP0 and OP as shown in Figure 1.22.
Let a be a vector with representation P0 P , then applying the Triangle
Law for vector addition gives:
r = r0 + a
However, since a and v are parallel vectors, there is a scalar t such
that:
a = tv
Thus,
r = r0 + t v (1.6)
This is a vector equation of L.
1.2.20 Scalar Equations of a Line
First, we state that two vectors are equal if and only if corresponding
components are equal. Hence, we have the following three scalar
equations.
x x0 at
y y0 bt (1.7)
z z0 ct
c, t . These equations are called the parametric equations of the
line L through the point P0 x0 , y 0 , z0 and parallel to the vector
v a, b, c . Each value of the parameter t determines a point x, y , z
on L.
Another way of describing a line L is to eliminate the parameter t from
parametric equations. If none of a, b, or c is 0, we can solve each of
Scalar and vector fields 23
these equations for t, equate the results, then we can derive the
following equations:
x x0 y y0 z z0
(1.8)
a b c
These equations are called symmetric equations of L. We call numbers
a, b, c the direction numbers of L.
Example 1.6: Find parametric equations and symmetric equations of
the line that passes through the points A 2, 4, 3 and B 3, – 1, 1 .
Solution: Although a vector parallel to the
line is not given but we can
see that the vector v with representation AB is parallel to the line, so:
v 3 – 2, –1 – 4,1 – –3 1, –5, 4
Thus, direction numbers are:
a 1, b –5, c 4
Taking the point 2, 4, –3 as P0 , we see that the parametric equations
are:
x 2 t , y 4 – 5t , z –3 4t
Therefore, the symmetric equations are:
x2 y 4 z 3
1 5 4
We can see from Example 1.6 that direction numbers of the line L
through the points P0 x0 , y0 , z0 and P1 x1 , y1 , z1 are:
x1 – x0, y1 – y0, z1 – z0
Thus, symmetric equations of L can be written as:
24 Mathematics for mechanical engineers
x x0 y y0 z z0
x1 x0 y1 y0 z1 z0
Example 1.7:
a) Find parametric equations and symmetric equations of the
line that passes through the points A 3,5, –2 and B –1, 2,3 .
b) At what point does this line intersect the xy-plane?
Solution:
a) Similar to Example 1.6 we know that the vector v with
representation AB is parallel to the line as depicted in Figure 1.23 and
v 1 3, 2 5,3 2 4, 3,5
Thus, direction numbers are:
a 4, b 3, c 5
Taking the point (3, 5, –2) as P0 , we see that
Parametric equations are: x 3 – 4t , y 5 – 3t , z –2 5t
x3 y 5 z 2
Symmetric equations are:
4 3 5
b) The line intersects the xy-plane when z 0 . So, replace z 0 in the
symmetric equations and obtain:
x3 y 5 2
4 3 5
This gives x 7 / 5 and y 19 / 5 . Thus, the line intersects the xy-plane
7 19
at the point C , , 0 .
5 5
Scalar and vector fields 25
Figure 1.23. Vector AB
1.2.21 Equation of Line Segments
In many cases, we need to describe just a line segment, not an entire
line. So how could we describe the line segment? Assume that we
need a description of the line segment AB in Example 1.7. We will
find the way to describes it.
If we put t 0 in the parametric equations in Example 1.7a, we get
the point 3,5, –2 . If we put t 1 , we get –1, 2,3 . So, the line
segment AB is described by either:
The parametric equations: x 3 – 4t , y 5 – 3t , z –2 5t
where 0 t 1 .
The corresponding vector equation:
r t 3 – 4t , 5 – 3t , –2 5t where 0 t 1
We know from Equation 1.6 that the vector equation of a line through
the tip of the vector r0 in the direction of a vector v is r r0 tv . If the
line also passes through the tip of r1, then we can take v r1 – r0 . So,
its vector equation is:
r r0 t r1 – r0 1 – t r0 tr1
26 Mathematics for mechanical engineers
This means that the line segment from r0 to r1 is given by the
parameter interval 0 t 1 .
1.2.22 Equation of Planes
As we know that a line in space is described by a point and a
direction, but a plane in space is more difficult to present. A single
vector is not enough to determine the ‘direction’ of a plane whether it
is in the plane or parallel to the plane. However, a vector
perpendicular to the plane does completely specify its direction. Thus,
a plane in space is determined by: A point P0 x0 , y0 , z0 in the plane
and a vector n that is orthogonal to the plane. This orthogonal vector n
is called a normal vector of the plane as shown in Figure 1.23.
Let P x, y , z be an arbitrary point in the plane. Let r0 and r1 be the
position vectors of P0 and P. Then, the vector r – r0 is represented by
P0 P . The normal vector n is orthogonal to every vector in the given
plane. In particular, n is orthogonal to r – r0 . Thus, we have:
n r r0 0
That can also be written as:
z
n r n r0
Either one of these two n
P(x,y,z)
equations is called vector
equation of the plane. Now to r - r0
r
derive a scalar equation for
the plane, we write: r0
P0(x0, y0, z0)
n a , b, c x y
r x, y , z
r0 x0 , y0 , z0 Figure 1.23. normal
vector of the plane
Scalar and vector fields 27
Then, the vector equation becomes:
a, b, c x x0 , y y0 , z z0 0
or it can also be written as:
a x x0 b y y0 c z z0 0 (1.9)
This equation is the scalar equation of the plane which is through
P0 x0 , y0 , z0 with normal vector n a, b, c . We call Equation 1.9
the standard equation of a plane.
Example 1.8: Find an equation of the plane through the point 2,3,1
with normal vector n 3, 4, 6 . Find the intercepts and sketch the
plane.
Solution: Putting
a 3, b 4, c 6, x0 2, y0 3, z0 1
into Equation 1.9, we see that an equation of the plane is:
3 x – 2 4 y – 3 6 z – 1 0
or
z
3x 4 y 6 z 24 (0,0,4)
In order to find the x-intercept, we
set y z 0 in the plane equation,
then we obtain x 8 . Similarly, the (0,6,0)
y-intercept is 6 and the z-intercept
is 4. From these intercepts we can x y
sketch the portion of the plane that (8,0,0)
lies in the first octant as shown in
Figure 1.24. A portion of the plane
Figure 1.24.
28 Mathematics for mechanical engineers
By collecting terms in Equation 1.9 as we did in Example 1.8, the
equation of a plane can be rewritten as follows.
ax by cz d 0 (1.10)
where d ax0 by0 cz0 . This is called a linear equation in x, y,
and z.
Conversely, it can be shown that, if a, b, and c are not all 0, then the
linear Equation 1.10 represents a plane with normal vector a, b, c .
1.2.23 Parallel planes
We say two planes are parallel if their normal vectors are parallel. For
instance, the planes:
3x 4 y 6 z 24 and –6 x – 8 y –12 z 12
are parallel because their normal vectors are:
n1 3, 4, 6 and n 2 –6, –8, –12 , this means n 2 –2n1 .
1.2.24 Nonparallel
planes
n1
n2
If two planes are not parallel,
then they intersect in a θ
straight line. The angle
between the two planes is
defined as the acute angle θ
between their normal vectors
as illustrated in Figure 1.25
Figure 1.25. Nonparallel planes
Example 1.9: Find:
a) The angle between the planes x 2 y 3z 1 and
3x – y z 3
Scalar and vector fields 29
b) Symmetric equations for the line of intersection L of these
two planes.
Solution
a) From the given plane equations, the normal vectors of these planes
are:
n1 1, 2,3 , n 2 3, –1,1 Thus, if θ is the angle between the planes,
then Corollary 1.1 gives:
n1 n 2 1(3) 2(1) 3(1) 4
cos
n1 n 2 1 4 9 9 11 154
4
cos 1 71
o
154
b) From the symmetric equations for the line we see that the line L is
determined if we know a point on it and a vector parallel to it. So we
first need to find a point on L then find vector v parallel to L:
One way is to find the point where the line intersects the xy-
plane by setting z 0 in the equations of both planes.
This provides the equations x 2 y 1 and 3x – y 3
whose solution is x 1, y 0 .
So, the point 1, 0, 0 lies on L.
As L lies in both planes, it is perpendicular to both the normal vectors.
Thus, a vector v parallel to L is given by the cross product:
i j k
v n1 n 2 1 2 3 5 i 8 j 7 k
3 1 1
So, the symmetric equations of L can be written as:
x 1 y z
5 8 7
30 Mathematics for mechanical engineers
1.2.25 Distance from a Point to a Plane
To find a formula for the distance D from a point P1 x1 , y1 , z1 to the
plane ax by cz d 0 , we let P0 x0 , y0 , z0 be any point in the
plane, let b be the vector with representation P0 P1 , then:
b x1 x0 , y1 y0 , z1 z0
As can be seen from Figure 1.26, the
distance D from P1 is equal to the P1
absolute value of the scalar projection of
θ
b onto the normal vector n a, b, c . D
b n
Therefore,
P0
n b
D comp n b
n
a ( x1 x0 ) b( y1 y0 ) c( z1 z0 ) Figure 1.26. Distance
from a point to a plane
a 2 b2 c2
(ax1 by1 cz1 ) ( ax0 by0 cz0 )
a 2 b2 c2
Since P0 lies in the plane, its coordinates satisfy the equation of the
plane, hence:
ax0 by0 cz0 d 0
Thus,
ax1 by1 cz1 d
D (1.11)
a2 b2 c2
is the distance between the parallel planes.
Example 1.10: Given that two lines
Scalar and vector fields 31
L1 : x –1 2t , y 1 t , z 3 – t
L2 : x 2 – s, y 1 3s, z –3 2 s
are skew. Find the distance between them.
Solution: Since the two lines L1 and L2 are not intersected, they can
be viewed as lying on two parallel planes P1 and P2 as shown in
Figure 1.27. Thus, the distance between L1 and L2 is equal to the
distance between P1 and P2 . Thus, to solve the problem we will find a
point on P1 and an equation of P2 then apply Equation 1.11. We have
the direction of L1 and L2 are:
v1 2,1, –1 (direction of L1 )
n
v 2 –1,3, 2 (direction of L2 )
v2
L2
So the normal vector of the two lines
is:
n v1 v 2 v1
L1
i j k Figure 1.27. Parallel planes
2 1 1 5 i 3 j 7k
1 3 2
If we put s 0 in the equations of L2 , we get the point 2,1, –3 on
L2 . So, an equation for P2 is:
5 x – 2 – 3 y –1 7 z 3 0
or
5 x – 3 y 7 z 14 0
Now we set t 0 in the equations for L1 , then we get the point
–1,1,3 on P1 . So, the distance between L1 and L2 is the same as the
32 Mathematics for mechanical engineers
distance from –1,1,3 to plane 5 x – 3 y 7 z 14 0 . Thus, this
distance is:
5( 1) 3(1) 7(3) 14 27
D 2.96
5 (3) (7)
2 2 2
83
1.3 Vector Fields
1.3.1 Vector Functions
Definition 1.4: A vector function is a function whose domain is a set
of real numbers and whose range is a set of vectors.
In engineering problems, vector functions r whose values are three-
dimensional (3-D) vectors are most useful. Therefore, from now on,
for every number t in the domain of r, there is a unique vector in V3
denoted by r t . If f t , g t , and h t are the components of the
vector r t , then f, g, and h are real-valued functions called the
component functions of r. We can write:
r t f t , g t , h t f t i g t j h t k
1.3.2 Limit of a Vector Function
Definition 1.5: The limit of a vector function r is defined by taking the
limits of its component functions. If r t f t , g t , h t , then
lim r (t ) lim f (t ), lim g (t ), lim h(t )
t a t a t a t a
provided the limits of the component functions exist.
If we write limt a r (t ) L , then this means that the length and
direction of the vector r t approach the length and direction of the
vector L.
Scalar and vector fields 33
Example 1.11: Find limt 0 r (t ) , where
ln(1 t )
r (t ) (1 t )i sin t j k
t
Solution: From Definition 5, the limit of r is the vector whose
components are the limits of the component functions of r:
ln 1 t
lim t 0 r t lim t 0 1 t i lim t 0 sin t j lim t 0 k
t
ik
1.3.3 Continuous of Vector Function and Space Curve
Definition 1.6: A vector function r is
continuous at a if: z
limt 0 r(t ) r(a)
P(f(t), g(t), h(t))
According to Definition 6, we can
state that r is continuous at a if and
only if its component functions f, g, C
and h are continuous at a. Assume r(t) = 〈f(t), g(t), h(t)〉
that f, g, and h are continuous real- O y
valued functions on an interval I. x
Then, the set C of all points x, y , z
in space, where Figure 1.28. Space curve
x f t , y g t , z h t
and t varies throughout the interval I is called a space curve as shown
in Figure 1.28. These equations are called parametric equations of C.
Also, t is called a parameter. If we write function
r t f t , g t , h t , then it means that r t is the position
vector of the point P f t , g t , h t on C.
34 Mathematics for mechanical engineers
1.3.4 Derivatives of Vector Functions
Definition 1.7: The derivative r t of a vector function is defined in
much the same way as for real-valued functions.
dr r t h r t
r t lim h 0
dt h
if this limit exists.
Theorem 1.6: If r t f t , g t , h t f t i g t j h t k ,
where f, g, and h are differentiable functions, then:
r t f t , g t , h t f t i g t j h t k
Proof: We have
r t
1
lim t 0 [r t t r t ]
t
1
lim t 0 f t t , g t t , h t t f t , g t , h t
t
f t t f t g t t g t h t t h t
lim t 0 , ,
t t t
f t , g t , h t
Example 1.12:
a) Find the derivative of:
1
r t 1 t i sin t j e 2 t k
2
b) Find the unit tangent vector at the point where t 0 .
Solution:
Scalar and vector fields 35
a) According to Theorem 1.6, we differentiate each component of r:
r t i cos t j e 2 t k
1
b) Since r 0 i k and r 0 i j k , the unit tangent vector
2
at the point 1, 0,1 is:
r 0 i j k 1 1 1
T(0) i j k
| r 0 | 111 3 3 3
Theorem 1.7: Suppose that u and v are differentiable vector functions,
c is a scalar, f is a real-valued function, then
d
1. u t v t u t v t
dt
d
2. cu t cu t
dt
d
3. f t u t f t u t f t u t
dt
d
4. u t v t u t v t u t v t
dt
d
5. u t v t u t v t u t v t
dt
d
6. u f t f t u f t (Chain Rule)
dt
1.3.5 Integrals of Vector Functions
The integral of a continuous vector function r t can be defined in
much the same way as for real-valued functions, except that the
integral is a vector. It means that we can express the integral of r in
terms of the integrals of its component functions f, g, and h as follows.
36 Mathematics for mechanical engineers
b
r t dt
a
n
lim n r ti* t
i 1
n n n
lim n f ti* t i g ti* t j h ti* t k
i 1 i 1 i 1
Thus
a
b
r (t ) dt
a
b
f (t ) dt i
b
a h(t ) dt k
g (t ) dt j
b
1.3.6 Vector Fields
y F(x,y)
Definition 1.8: Let D be a set in
2 (a plane region). A vector
(x,y)
field on 2 is a vector function
F that assigns to each point
x, y in D a two-dimensional O x
(2-D) vector F x, y .
We can illustrate a vector field by Figure 1.29. A vector field
drawing the arrow representing
the vector F x, y starting at the point x, y . In fact, it’s impossible
to do this for all points x, y so we just draw some representative
vectors as shown in Figure 1.29.
1.3.7 Vector Fields on 2
Since F x, y is a 2-D vector, we can express it in terms of its
component functions P and Q as:
F x, y P x , y i Q x, y j P x, y , Q x, y
or, for short,
Scalar and vector fields 37
F Pi Qj
Here P and Q are scalar functions of two variables. They are called
scalar fields to distinguish them from vector fields.
Example 1.13: A vector field on 2 is defined by: F x, y yi xj .
Describe F by sketching some of the vectors F x, y .
Solution: Since F 1, 0 j , we
draw the vector j 0,1 starting 2
at the point 1, 0 . Since
1
F 0,1 i , we draw the vector
–1, 0 with starting point 0,1 . 0
Continuing in this way, we
-1
calculate several other
representative values of F x, y
-2
then we draw the corresponding
vectors to represent the vector -2 -1 0 1 2
field as shown in Figure 1.30. x
It can be noted that that each Figure 1.30. 2D vector field
arrow is tangent to a circle with
center the origin. To confirm this, we take the dot product of the
position vector x xi yj with the vector F x F x, y :
x F x xi yj – yi xj – xy yx 0
This points out that F x, y is perpendicular to the position vector
x, y and is therefore tangent to a circle with center the origin and
radius | x | x 2 y 2 . Notice also that:
| F ( x, y ) | ( y ) 2 x 2 x 2 y 2 | x |
38 Mathematics for mechanical engineers
This means that the magnitude of the vector F x, y is equal to the
radius of the circle.
1.3.8 Vector Fields on 3
Let E be a subset of 3 . A vector field on 3 is a function F that
assigns to each point x, y , z in E a three-dimensional (3-D) vector
F x, y , z . We can express it in terms of its component functions P,
Q, and R as:
F x, y , z P x, y , z i Q x, y , z j R x , y , z k
Example 1.14: Sketch the vector
field on 3 given by:
F x, y , z zk
Solution: Following the same
procedure as we did for the
vector field in 2 . The sketch is
shown in Figure 1.31 with noting
that all vectors are vertical and
point upward above the xy-plane
or downward below it. The
magnitude increases with the
distance from the xy-plane.
1.4 Velocity Field Figure 1.31. 3D vector field
To give an example of a vector field in practice, imagine a fluid
consisting of particles flowing steadily along a pipe. Let V x, y , z be
the velocity vector of a particle at point x, y , z . Then, V assigns a
vector to each point x, y , z in a certain domain E (the interior of the
pipe). So, V is a vector field on 3 called a velocity field. A fluid
velocity field in a tube is illustrated Figure 1.32. The speed of the
particle at any given point is indicated by the length of the arrow.
Scalar and vector fields 39
Figure 1.32. Velocity vector field
Velocity fields can be found in many other areas of physics. For
instance, the vector field in Example 1.14 could be used as the
velocity field describing the counterclockwise rotation of a wheel.
Figure 1.33. Wind field in North Vietnam
Another example is the wind field which is very important for weather
forecasting. Figure 1.33 presents the wind field in North Vietnam on
Monday 16 August 2021 at 4:50 PM. They indicate the wind speed
40 Mathematics for mechanical engineers
and direction at points 10 m above the surface. Associated with every
point in the air, we can imagine a wind velocity field.
1.5 Gravitational Field
Newton’s Law of Gravitation is very well known, which states that the
magnitude of the gravitational force between two objects with masses
m and M is:
mMG
| F |
r2
where r is the distance between the objects, G is the gravitational
constant. For instance, M is the mass of the earth and the origin would
be at its center. Let the position vector of an airplane with mass m be
x x, y , z . Then,
r x
So,
2
r2 x
The gravitational force exerted on the airplane acts toward the origin.
The unit vector in this direction is:
x
|x|
Thus, the gravitational force acting on the airplane at x x, y , z is:
mMG
F (x ) x
| x |3
Physicists often use the notation r instead of x for the position vector.
So, we can write:
Scalar and vector fields 41
mMG
F r (1.12)
r3
This function is called the gravitational field because it associates a
gravitational force vector F x with every point x in space. Formula
1.12 is a compact way of writing the gravitational field. However, we
can also write it in terms of its component functions. We do this by
using the facts that:
x xi yj zk and x x 2 y 2 z 2
F ( x, y , z )
mMGx mMGy mMGz
i 2 j 2 k
(x y z )
2 2 2 3/2
(x y z )
2 2 3/ 2
( x y 2 z 2 )3/ 2
The gravitational field F is pictured here by computer program.
0.5
-0.5
0.5
0.5
0
0
y -0.5 -0.5 x
Figure 1.34. Gravitational field
42 Mathematics for mechanical engineers
1.6 Conservative Vector Fields
A vector field F is called a conservative vector field if it is the
gradient of some scalar function, that is, if there exists a function f
such that F f . In this case, f is called a potential function for F. It
should be noted that, not all vector fields are conservative. Still, such
fields do arise frequently in physics.
The gravitational field F in the previous example is conservative.
Suppose we define:
mMG
f ( x, y , z )
x2 y2 z2
Then,
f ( x, y, z )
f f f
i j k
x y z
mMGx mMGy mMGz
i 2 j 2 k
(x y z )
2 2 2 3/2
(x y z )
2 2 3/ 2
( x y 2 z 2 )3/ 2
F ( x, y, z )
This means that if we give a vector field F:
F ( x, y , z )
mMGx mMGy mMGz
i 2 j 2 k
(x y z )
2 2 2 3/2
(x y z )
2 2 3/ 2
( x y 2 z 2 )3/ 2
we can find a function
mMG
f ( x, y , z )
x2 y2 z2
so that
Scalar and vector fields 43
f ( x, y, z ) F( x, y, z ) .
Therefore, the given vector field is conservative. In next chapters we
will learn how to tell whether or not a given vector field is
conservative.
44 Mathematics for mechanical engineers
EXCERCISES
I. Find the sum of the given vector and illustrate geometrically.
1. 0, 4 , 3, 0
2. 1, 3 , 3, 2
3. 1, 0, 2 , 2, 2, 0
4. 0, 0,3 , 1, 0, 0 .
II. Find a b, a 2b, 2a , 2a b
1. a 2, 4 , b 4, 2
2. a i 3 j, b 2i 5 j
3. a i 3j 2k , b 2i 2 j k
4. a 3i 2k , b 2 j 3k
5. a i 3j k , b 2i k .
III. Find a unit vector that has the same direction as the given vector
1. 3, 4
2. 5i 3j
3. 3, 4, 5
4. 2i 3j 4k .
IV. Find a.b
Scalar and vector fields 45
1. a 2, 4 , b 4, 2
2. a 1, 3 , b 3, 2
3. a i 3 j, b 2i 5 j
4. a 1, 0, 2 , b 2, 2, 0
5. a 0, 0,3 , b 1, 0, 0
6. a i 3j 2k , b 2i 2 j k
7. a 3i 2k , b 2 j 3k
8. a i 3j k , b 2i k .
V. Find the angle between the vectors.
1. a 2, 4 , b 4, 2
2. a 1, 3 , b 3, 2
3. a i 3 j, b 2i 5 j
4. a 1, 0, 2 , b 2, 2, 0
5. 0, 0,3 , 1, 0, 0
6. a i 3j 2k , b 2i 2 j k
7. a 3i 2k , b 2 j 3k
8. a i 3j k , b 2i k
VI. Find the cross product a b .
46 Mathematics for mechanical engineers
1. a 0, 0,3 , b 1, 0, 0
2. a 1,3, 9 , b= 1, 3, 9
3. a 3, 1, 2 , b 2, 1, 4
4. a 0, 0, 3 , b 2,1,5
5. a i 3j 2k , b 2i 2 j k
6. a 3i 2k , b 2 j 3k
7. a i 3j k , b 2i k .
VII. Find an equation of the plane
1. The plane through the point 1, 2,3 and with normal vector
2, 0, 3 .
2. The plane through the point 3, 2,3 and with normal vector
1,3, 2 .
3. The plane through the point 3, 2,3 and perpendicular to the
vector 1, 3,5 .
4. The plane through the point 3, 2,3 and perpendicular to the line
x t , y 3 2t , z 5 3t .
VIII. Find the distance from the point to the given plane
1. 3, 1, 2 , x 2 y 3z 0
2. 0, 3, 5 , 2x y 4z 2
Scalar and vector fields 47
3. 2, 1, 4 , 4 x y 3z 9
4. 3,1,9 , x 3 y 5z 9
IX. Find the distance between the given parallel planes.
1. x 2 y 3z 2, 2 x 4 y 6 z 5
2. P : 2 x y 5 z 1, Q :4 x 2 y 10 z 5 .
ANSWERS
I.
1. 0, 4 3, 0 3, 4
2. 1, 3 3, 2 2, 1
3. 1, 0, 2 2, 2, 0 1, 2, 2
4. 0, 0, 3 1, 0, 0 1, 0, 3
48 Mathematics for mechanical engineers
II.
2.
a i 3j, b 2i 5 j
a b 1, 3 2, 5 1, 2
a 2b 1, 3 2 2,5 1, 3 4,10 3, 7
2a 2 1, 3 2, 6 22 6 40 2 10
2
2a b 2 1, 3 2,5 2, 6 2,5
4, 11 42 11 137
2
4.
a 3i 2k , b 2 j 3k
a b 3, 0, 2 0, 2,3 3, 2,1
a 2b 3, 0, 2 0, 4, 6 3, 4,8
Scalar and vector fields 49
2a 6,0, 4 62 02 4 52
2
2a b 6,0, 4 0, 2,3 6, 2, 7
62 2 7 89
2 2
III. Find a unit vector that has the same direction as the given vector
5 3
2. v 5i 3 j v 52 32 34 u ,
34 34
2
2
4. v 2i 3j 4k v 32 42 29
2 3 4
u i j k
29 29 29
IV. Find a.b
2. a 1, 3 , b 3, 2 a.b 1 3 3 2 9
4. a 1, 0, 2 , b 2, 2, 0 a.b 1 2 0 2 2 0 2
6. a i 3 j 2k , b 2i 2 j k
a.b 1 2 3 2 2 1 10
8. a i 3 j k , b 2i k a.b 1.2 3.0 1 .1 1
V.
2. a 1, 3 , b 3, 2 a.b 1 3 3 2 9
a 10, b 13
50 Mathematics for mechanical engineers
a.b 9 9
cos a, b a, b 1420
a b 10 13 130
4. a 1,0, 2 , b 2, 2, 0 a.b 1 2 0 2 2 0 2
a 5, b 8
a.b 2 2
cos a, b a, b 1080 26'
a b 5 8 40
6. a i 3 j 2k , b 2i 2 j k
a.b 1 2 3 2 2 1 10, a 11, b 3
a.b 10
cos a, b a, b 10
a b 3 11
8. a i 3j k , b 2i k
a.b 1.2 3.0 1 .1 1, a 11, b 5
a.b 1 1
cos a, b a, b 82015'
a b 11 5 55
VI.
2. a 1, 3,9 , b 1,3, 9 a b 54, 0, 6
4. a 0, 0,3 , b 2,1, 5 a b 3, 6,0
6. a 3, 0, 2 , b 2, 0,1 a b 0, 7, 0
VII.
Scalar and vector fields 51
2. The equation of the plane is:
2 x 1 0 y 2 3 z 3 0 or 2 x 3 z 11
4. The normal vector n 1, 2, 3 . Therefore, the equation of the
plane is:
1 x 3 2 y 2 3 z 3 0
or x 2 y 3z 16
VIII.
2.0 1. 3 4.5 2 25
2. D
2 12 4 21
2 2
1.3 3.1 5.9 9 36
4. D
12 3 52 35
2
IX. Find the distance between the given parallel planes.
2. The planes are parallel because their normal vectors
2,1, 5 , 4, 2,10 are parallel.
1
The point 0, 0, is in the plane Q : 4 x 2 y 10 z 5 .
2
The distance between P and Q is the same as the distance from
1
0, 0, to P. This distance is
2
1
5. 1
2 7
D .
2 12 5 2 30
2 2
52 Mathematics for mechanical engineers
Gradient, divergence, curl and laplacian operators 53
CHAPTER 2. GRADIENT, DIVERGENCE, CURL
AND LAPLACIAN OPERATORS
In this section we define four operators that can be performed on
scalar and vector fields. These operators play a basic role in the
applications of vector calculus to engineering, physics, and computer
sciences, in general, but particularly solid mechanics, aerodynamics,
aeronautics, fluid flow, heat flow, electrostatics, quantum physics,
laser technology, robotics as well as other areas have applications that
require an understanding of vector calculus.
2.1 Gradient operator
2.1.1 Partial Derivative
Since the gradient operator associates with the partial derivative then
the partial derivative is presented here first. The partial derivative of a
function of several variables is its derivative with respect to one of
those variables, with the others held constant. We define the partial
derivative here. If z f x, y , then the partial derivatives f x and f y
are defined as:
f ( x0 h, y0 ) f ( x0 , y0 )
f x ( x0 , y0 ) lim
h0 h
f ( x0 , y0 h) f ( x0 , y0 )
f y ( x0 , y0 ) lim
h 0 h
There are many alternative notations for partial derivatives. For
instance, instead of f x , we can write f1 or D1 f (to indicate
f
differentiation with respect to the first variable) or . If z f x, y ,
x
we write:
f z
f x ( x, y ) f x f ( x, y ) f1 D1 f Dx f
x x x
f z
f y ( x, y ) f y f ( x, y ) f 2 D2 f Dy f
y y y
54 Mathematics for mechanical engineers
Example 2.1: If f x, y x 3 3 x 3 y 2 2 y 4 , find f x 1, 2 and f1 1, 2
.
Solution: Holding y constant and differentiating with respect to x, we
get:
f x x, y 3 x 2 9 x 2 y 2
Thus,
f x 1, 2 3.12 9.12.22 33
Holding y constant and differentiating with respect to x, we get:
f y x, y 6 x3 y 8 y 3
Then,
f y 1, 2 6.13.2 8.23 52
Definition 2.1 (Tangent plane equations): Suppose C1 and C2 are the
curves obtained by intersecting the vertical planes y y0 and x x0
with the surface S as shown in
Figure 2.1. Suppose C1 and C2
intersect at point P. If T1 and T2
are the tangent lines to the
curves C1 and C2 at the point P.
Then, the tangent plane to the
surface S at the point P is the
plane that contains both tangent
lines T1 and T2.
Recall from Chapter 1 that any
plane passing through the point
P x0 , y0 , z0 has an equation
Figure 2.1. Tangent plane
of the form
Gradient, divergence, curl and laplacian operators 55
A x x0 B y y0 C z z0 0
This equation can be rewritten as:
z z0 a x x0 b y y0 (2.1)
Suppose that Equation 2.1 represents the tangent plane at P, then this
tangent plane will intersect the plane y y0 by the tangent line T1 .
Setting y y0 in Equation 2.1 gives:
z z0 a x x0
or
z ax ax0 z0
This is the equation of a T1 with the slop a f x x0 , y0 . Similarly,
putting x x0 in Equation 2.1, we get: z z0 b y y0 . This must
represent the tangent line T2 . Thus, b f y x0 , y0 . Therefore, if f has
continuous partial derivatives, then from Equation 1, an equation of
the tangent plane to the surface z f x, y at the point P x0 , y0 , z0
is:
z z0 f x x0 , y0 x x0 f y x0 , y0 y y0 (2.2)
Theorem 2.1 (Chain Rule): Suppose that z f x, y is a
differentiable function of x and y, where x g s and y h s are
both differentiable functions of s. Then, z is a differentiable function
of s and:
dz f dx f dy
ds x ds y ds
Example 2.2: If z x 2 y 3 xy 4 , where x sin 2s and y cos s , find
dz / ds when s 0 .
56 Mathematics for mechanical engineers
Solution: The Chain Rule gives:
dz z dx z dy
(2 xy 3 y 4 )(2 cos 2 s) ( x 2 12 xy 3 )( sin s )
ds x ds y ds
In this case we simply observe that, when s 0 , we have x sin 0 0
and y cos 0 1 so we don't need to substitute the expressions for x
and y in terms of s then,
dz
(0 3)(2.1) (0 0)( 0) 6
ds s 0
Having presented the definition of partial derivative of a function of
several variables, now we can define the gradient operator as follows.
2.1.2 Gradient Operator
Definition 2.2: We introduce the vector differential operator (“del”)
as:
i j k
x y z
This is called the gradient operator. For a function f of two variables,
the gradient vector, denoted by f or grad f, is:
f f
f ( x, y) f x ( x, y ), f y ( x, y) i j
x x
Example 2.3: If f ( x, y ) x 2 y sin y 2 , then
f ( x, y ) f x , f y 2 xy , x 2 2 y cos y 2
f (1, 0) 0,1
In general, using scalar fields instead of vector fields is of a
considerable advantage because scalar fields are easier to use than
vector fields. The gradient operator associates the scalar fields and it
Gradient, divergence, curl and laplacian operators 57
allows us to obtain vector fields from scalar fields. The gradient is of
great practical importance to the engineer. We can see some uses of
the gradient as follows.
2.1.3 Directional Derivative
The directional derivative is a type of derivative that enables us to find
the rate of change of a function of two or more variables in any
direction. The directional derivative of f at x0 , y0 in the direction of a
unit vector u a, b is defined as follows:
f ( x0 ha, y0 hb) f ( x0 , y0 )
Du f ( x0 , y0 ) limh0
h
Theorem 2.2: With the notation for the gradient vector, we can express
the directional derivative as:
Du f ( x, y ) f x ( x, y )a f y ( x, y )b
f x ( x, y ), f y ( x, y ) a, b
f ( x , y ) u
This expresses the directional derivative in the direction of the unit
vector u as the scalar projection of the gradient vector onto u.
Proof: Assume that g h is defined as a function of h as
g h f x0 ha, y0 hb , then:
g (h) g (0)
g (0) lim
h 0 h
f ( x0 ha, y0 hb) f ( x0 , y0 )
lim (2.3)
h 0 h
Du f ( x0 , y0 )
Put x x0 ha, y y0 hb , we have:
58 Mathematics for mechanical engineers
f x f y
g (h) f x ( x, y )a f y ( x, y )b (2.4)
x h y h
If we put h 0 into Equation 2.4 and compare with Equation 2.3, then
Du f ( x0 , y0 ) f x ( x0 , y0 ) a f y ( x0 , y0 )b
This equation is true for any ( x0 , y0 ) so we can replace ( x0 , y0 ) by
( x, y) in this equation to establish the theorem:
Du f ( x, y ) f x ( x, y ) a f y ( x, y )b
Example 2.4: Find the directional derivative of the function
f x , y xy 3 4 x 2 at the point 1, –2 in the direction of the vector
v i 3j .
Solution: We first compute the gradient vector at 1, –2 :
f ( x, y ) y 3 8 x i 3xy 2 j
f (1, 2) 16 i 12 j
Note that v is not a unit vector so we need to find the unit vector in the
direction of v. Since | v | 10 , the unit vector in the direction of v is:
v 1 3
u i j
|v| 10 10
Therefore,
1 3
Du f (1, 2) f (1, 2) u (16i 12 j) i j
10 10
16 1 12 3 52
10 10
Gradient, divergence, curl and laplacian operators 59
Directional derivatives of functions of three variables can be define in
a similar manner.
Definition 2.3: For a function f of three variables, the gradient vector,
denoted by f or grad f, is:
f ( x, y , z ) f x ( x, y , z ), f y ( x, y , z , ), f z ( x, y , z )
or,
f f f
f f x , f y , f z i j k
x y z
Example 2.4: If f x, y, z xye xz find:
a) The gradient of f
b) The directional derivative of f at 1, 2, 0 in the direction of
v 3i 2 j k .
Solution:
a) The gradient of f is:
f ( x, y , z ) f x ( x, y , z ), f y ( x, y, z ), f z ( x, y, z )
ye xz xyze xz , xe xz , x 2 ye xz
b) At 1, 2, 0 , we have:
f (1, 2, 0) 2,1, 2
The unit vector in the direction of v 3i 2 j k is:
3 2 1
u i j k
14 14 14
60 Mathematics for mechanical engineers
Therefore, we find the directional derivative of f at 1, 2, 0 as:
Du f (1, 2, 0) f (1, 2, 0) u
3 2 1 6
2i j 2k i j k
14 14 14 14
2.1.4 Tangent Planes to a Level Surfaces
Let S be a surface with equation F x, y , z k as presented in Figure
2.2. That is, this equation describes a level surface of a function F of
three variables. Then, let P x0 , y0 , z0 be a point on S. And, let C be
any curve that lies on the surface S and passes through the point P.
Recall that the curve C is described
by a continuous vector function:
r t x t , y t , z t
Let t0 be the parameter value
corresponding to point
P x0 , y0 , z0 , it means:
r t0 x0 , y0 , z0
Then, any point x t , y t , z t
needs to satisfy the equation of S Figure 2.2. Level surface S
because C lies on S. That is,
F x t , y t , z t k
Suppose F is differentiable and x, y, and z are also differentiable
functions of t, then we can differentiate both sides of the previous
equation by using the Chain Rule to obtain:
F dx F dy F dz
0 (2.5)
x dt y dt x dt
Gradient, divergence, curl and laplacian operators 61
However, as we have defined:
F Fx , Fy , Fz
r(t ) x t , y t , z t
Equation 2.5 can be written in terms of a dot product of vectors F
and r t 0 as:
F r t 0
In particular, when t t0 , we have:
r t0 x0 , y0 , z0
So,
F x0 , y0 , z0 r t0 0
This equation says that the gradient vector at P x0 , y0 , z0 ,
F x0 , y0 , z0 , is perpendicular to the tangent vector r t0 to any
curve C on S that passes through P. If F ( x0 , y0 , z0 ) 0 , the tangent
plane to the level surface F x, y, z k at P x0 , y0 , z0 is thus defined
as the plane that passes through P and has normal vector
F ( x0 , y0 , z0 ) . Using the standard equation of a plane, we can write
the equation of this tangent plane as:
Fx ( x0 , y0 , z0 )( x x0 ) Fy ( x0 , y0 , z0 )( y y0 ) Fz ( x0 , y0 , z0 )( z z0 ) 0
2.1.5 Normal Line
The normal line to the surface S at point P is the line that passes
through P and is perpendicular to the tangent plane of S at P. Thus, the
direction of the normal line is given by the gradient vector
F ( x0 , y0 , z0 ) . So its symmetric equations are:
62 Mathematics for mechanical engineers
x x0 y y0 z z0
Fx ( x0 , y0 , z0 ) Fy ( x0 , y0 , z0 ) Fz ( x0 , y0 , z0 )
2.2. Curl Operator
Definition 2.4: Suppose that F Pi Qj Rk is a vector field on 3
and partial derivatives of P, Q, and R all exist. Then, the curl of F is
the vector field on 3 defined by:
R Q P R Q P
curl F i j k
y z z x x y
To remember this equation, let’s rewrite it using operator notation. We
introduce the vector differential operator (“del”) as:
i j k
x y z
It produces the gradient of a function when it operates on a scalar
function f:
f f f f f f
f i j k i j k
x y z x y z
Let consider as a vector with components / x , / y , and / z ,
we can also consider the formal cross product of with the vector
field F as follows:
i j k
F
x y z
P Q R
R Q P R Q P
i j k
y z z x x y
curl F
Gradient, divergence, curl and laplacian operators 63
R Q P R Q P
i j k
y z z x x y
curl F
Thus, the easiest way to remember Definition 2.4 is by means of the
symbolic expression:
curl F F (2.6)
Example 2.5: If F x, y , z x 2 i xyz 2 j zy 2k , find curl F.
Solution: Using Equation 2.6, we have the following result:
curl F F
i j k
x y z
x2 xyz 2 y2 z
y 2 z xyz 2 i y 2 z x 2 j
y z x z
xyz 2 x 2 k
x y
2 yz 2 xyz i 0 0 j yz 2 0 k
2 yz 1 x i yz 2 k
Theorem 2.3: If f is a function of three variables that has continuous
second-order partial derivatives, then:
curl f 0
Proof: From Equation 6 we have:
64 Mathematics for mechanical engineers
curl f f
i j k
x y z
f f f
x y z
2 f 2 f 2 f 2 f 2 f 2 f
i j k
y z z y z x x z x y y x
0i 0 j 0k 0
2.2.1 Conservative Vector Fields
A conservative vector field is defined as one for which F f where f
is a scalar function. It means that, if F is a vector function and there
exists a scalar function f so that F f then F is the conservative
vector field. So, Theorem 2.3 can be rephrased as: If F is conservative,
then curl F 0 . This gives us a very useful way to check if a vector
field is not conservative.
Example 2.6: Show that the vector field:
F x, y , z x 2 i xyz 2 j zy 2k
is not conservative.
Solution: In Example 2.5, we showed that:
curl F 2 yz 1 x i yz 2 k
This shows that curl F 0 . So, by Theorem 2.3, F is not conservative.
It is noted that the converse of this theorem is not true in general. The
following theorem, though, says that it is true if F is defined
everywhere. More generally, it is true if the domain is simply
connected, that is, it “has no hole.”.
Gradient, divergence, curl and laplacian operators 65
Theorem 2.4: If F is a vector field defined on all of 3 whose
component functions have continuous partial derivatives and curl
F 0 , then F is a conservative vector field.
Example 2.7:
a) Show that F x, y , z 2 xyzi x 2 zj x 2 yk is a conservative
vector field.
b) Find a function f such that F f .
Solution:
a) We have:
curl F F
i j k
x y z
2 xyz x2 z x2 y
x 2 x 2 i 2 xy 2 xy j 2 xz 2 xz k 0
As curl F 0 and the domain of F is 3 , F is a conservative vector
field by Theorem 2.4.
b) Since F is a conservative vector field then F f . That is:
f x x, y , z 2 xyz
f y x, y , z x 2 z (2.7)
f z x, y , z x 2 y
Integrating f x x, y , z with respect to x, we obtain:
f x , y , z x 2 yz g y , z
66 Mathematics for mechanical engineers
Where g y , z is the integrating constant. Differentiating f with
respect to y, we get
f y x, y , z x 2 z g y y , z ,
Thus, comparing this equation with the second equation in (2.7) gives:
g y y, z 0
Now, integrating this equation we have:
g y, z h z
Thus:
f x , y , z x 2 yz h z
Differentiating this equation with respect to z and compare to the third
equation in (2.7), we obtain:
f z x, y , z x 2 y
h z 0
Thus,
hz C
Therefore:
f x , y , z x 2 yz C (C – constant).
2.2.1 Physical Meaning of Curl
In physics, the curl vector is associated with rotations. For example,
particles around point x, y, z in the fluid tend to rotate about the axis
that points in the direction of curl F x, y , z . The length of this curl
Gradient, divergence, curl and laplacian operators 67
vector describes velocity of the
particles moving around the axis
as shown in Figure 2.3. curl F(x,y,z)
If curl F 0 at a point P, the fluid
is not rotated at P and F is called
irrotational at P. That is, there is (x,y,z)
no whirlpool or eddy at P. If
curl F 0 , there will be a tiny
paddle wheel moves with the fluid
but doesn’t rotate about its axis. If
curl F 0 , the paddle wheel Figure 2.3. curl F
rotates about its axis.
2.3. Divergence Operator
Definition 2.5: If F Pi Qj Rk is a vector field on 3 and P / x ,
Q / y , and R / z exist, the divergence of F is the function of three
variables defined by:
P Q R
div F
x y z
We observe that curl F is a vector field and div F is a scalar field. In
terms of the gradient operator:
i j k
x y z
the divergence of F can be written symbolically as the dot product of
and F:
div F F
Example 2.8: If F x, y , z x 2 i xyz 2 j zy 2k , Find div F.
Solution: By Definition 2.5 of divergence we have:
68 Mathematics for mechanical engineers
2
div F F
x
x xyz 2 zy 2 2 x xz 2 y 2
y z
If F is a vector field on 3 , then curl F is also a vector field on 3 .
As such, we can compute its divergence. The next theorem shows that
the divergence of curl F is always zero.
Theorem 2.5: If F Pi Qj Rk is a vector field on 3 and P, Q, and
R have continuous second order partial derivatives, then
div curl F 0
Proof: By the Definitions 4 and 5 of divergence and curl, we can see
div curl F F
R Q P R Q P
x y z y z x z x y
2 R 2Q 2 P 2 R 2Q 2 P
0
x y x z y z y x z x z y
The terms cancel in pairs by Clairaut’s Theorem. Note the analogy
with the scalar triple product we have:
a a b 0
This establish the theorem.
2.4 Laplace Operator
The Laplace operator is named after the French mathematician Pierre-
Simon de Laplace (1749–1827), who first applied the operator to the
study of celestial mechanics, where the operator gives a constant
multiple of the mass density when it is applied to the gravitational
potential due to the mass distribution with that given density. The
Laplacian occurs in differential equations that describe many physical
phenomena, such as electric and gravitational potentials, the diffusion
Gradient, divergence, curl and laplacian operators 69
equation for heat and fluid flow, wave propagation, and quantum
mechanics.
Definition 2.6: If f is a function of three variables, we have:
2 f 2 f 2 f
div f f
x 2 y 2 z 2
This equation occurs so often in physics that we abbreviate it as 2 f .
The operator 2 is called the Laplace operator due to its
relation to Laplace’s equation:
2 f 2 f 2 f
2 f 0
x 2 y 2 z 2
We can also apply the Laplace operator to a vector field
F Pi Qj Rk
in terms of its components:
2 F 2 P i 2Q j 2 R k
The curl and divergence operators allow us to rewrite Green’s
Theorem in versions that will be useful in our later work.
70 Mathematics for mechanical engineers
EXERCISES
I. Calculate div F , div F , curl F and 2F :
1) F y 2 i x 2 j z 2k
2) F x y 2 i xyj zk
3) F x 2 yzi xy 2 zj xyz 2k
4) F ln yi e xy j yxk
5) F e x i e y j e z k
II. Which function satisfies the Laplace equation?
1) f x, y , z 4 x 4 y 2 z
2) f x, y , z 3 x 2 2 y 2 2 z 2
3) f x, y, z x ln y 2 z 2
4) f x , y , z x e y z
5) f x, y , z 5 x 2 5 y 2 x y 4 z
ANSWERS
I.
2. F x y 2 i xyj zk
div F=.F=
x
x y 2 xy z 2 x
y z
Gradient, divergence, curl and laplacian operators 71
div F 1, 0, 0
i j k
curl F= F= 3 yk
x y z
x y2 xy z
2 F= 0
4. F ln yi e xy j yxk
div F xe xy
div F 1 xy e xy i x 2 e xy j
or div F e xy 1 xy , x 2 e xy , 0
1
curl F x, y, ye xy
y
2 F= x 2e xy
II.
2. f x, y , z 3 x 2 2 y 2 2 z 2
2 f 2 f 2 f
f 2 2 2 3 2 2 7 0
2
x y z
So, this function doesn’t satisfy the Laplace equation.
4. f x , y , z x e y z
72 Mathematics for mechanical engineers
2 f 2 f 2 f
2 f 0 e y z e y z 2e y z 0 .
x 2 y 2 z 2
So, this function doesn’t satisfy the Laplace equation.
5. f x, y , z 5 x 2 5 y 2 x y 4 z
2 f 2 f 2 f
2 f 10 10 0 0
x 2 y 2 z 2
So, this function satisfies the Laplace equation.
Multiple integrals 73
CHAPTER 3. MULTIPLE INTEGRALS
The multiple integral is a type of definite integral extended to
functions of more than one real variable. This chapter presents the
double and triple integrals and their applications for computing
volumes, masses, and centroids of more general regions than we were
able to consider using the single integral. Depending on types of
regions, polar coordinates are useful in computing double integrals.
Meanwhile, cylindrical coordinates and spherical coordinates are
suitable in calculating triple integrals over specific regions in three-
dimensional space.
3.1. Double Integrals Over
Rectangles
Let’s begin with the volume problem.
Figure 3.1 presents a surface above the
xy-plane which is described by a
function z f x, y defined in a
closed rectangle:
R a , b c, d
x, y | a x b, c y d
Let S be the solid that lies above R and Figure 3.1. Surface
under the graph of f, that is, above a solid
S x, y , z 3 | 0 z f x, y , x, y
We want to find the volume of S. Extending the idea in calculating the
area of a region, we divide the solid region into small thin rectangular
columns parallel to the z-axis so that their bottom surfaces lie on the
xy-plane and their tops intersect the surface z f x, y . Then, the
volume of S is approximated by sum of the small rectangular columns.
To do that, we first divide the rectangle R into subrectangles as shown
in Figure 3.2:
74 Mathematics for mechanical engineers
We divide the interval a, b into m subintervals xi –1 , xi of
equal width x b – a / m .
Then, we divide c, d into n subintervals y j –1 , y j of
equal width y (d – c) / n .
Then, we draw lines parallel to the coordinate axes through the
endpoints of these subintervals.
Figure 3.2. Rectangle R is divided into subrectangles
As a result, we obtain the subrectangles
Rij xi –1 , xi y j –1 , y j x, y | xi –1 x xi , y j –1 y y j
each with area
A xy
In each rectangle Rij we choose a sample point xij* , yij* . Then, we can
approximate the part of S that lies above each Rij by a thin rectangular
column with base Rij and height f xij* , yij* . The volume of this
Multiple integrals 75
column is the height of the column times the area of the base
rectangle:
f xij* , yij* A
We follow this procedure for all the
rectangles and add the volumes of
the corresponding column. Finally,
the volume of S is approximated by
rectangles as depicted in Figure 3.3.
Thus, we form an approximation to
the total volume of S:
Figure 3.3. Volume of S is
m n approximated by rectangles
V f ( xij* , yij* ) A
i 1 j 1
This double sum is called a double Riemann sum. We can see that as
the rectangular becomes smaller, i.e as m and n become larger the
approximation given in this equation becomes better. So, we would
expect that:
m n
V lim
m , n
f ( x , y ) A
i 1 j 1
*
ij
*
ij (3.1)
We use the limit in Equation 3.1 to define the double integral as
follows.
Definition 3.1: The double integral of f over the rectangle R is:
m n
f ( x, y ) dA lim
m , n
f ( x , y
i 1 j 1
*
ij
*
ij ) A
R
if this limit exists. If f is bounded, (that is f x, y M for all x, y
in R where M is constant) and f is continuous in R, except on a finite
number of smooth curves, then f is integrable over R. Since the sample
point xij* , yij* is arbitrary in the subrectangle Rij* , so for simplicity
76 Mathematics for mechanical engineers
reason here, we choose it to be the upper right-hand corner of Rij and
call its xi , y j . Then, the expression for the double integral becomes
simpler:
m n
f ( x, y ) dA lim
m , n
( x , y ) A
i 1 j 1
i i (3.2)
R
From Equations 3.1 and 3.2, the volume V of the solid that lies above
the rectangle R and below the surface z f x, y is:
V f ( x, y ) dA
R
Example 3.1: Approximate the
volume of the solid that lies above
the square R 0, 4 0, 4 and
below the elliptic paraboloid
z 100 x3 2 y 3 . by dividing R
into four equal squares and choosing
the sample point to be the upper
right corner of each square Rij .
Sketch the solid and the Figure 3.4. Approximating
approximating rectangular columns. rectangular boxes
Solution: Following the definition of the double integral we
approximate the volume by the Riemann sum with m n 4 , then we
have:
4 4
V f ( xi , y j ) A
i 1 j 1
f (1,1) A f (1, 2) A f (1,3) A f (1, 4) A
f (2,1) A f (2, 2) A f (2,3) A f (2, 4) A
f (3,1) A f (3, 2) A f (3,3) A f (3, 4) A
f (4,1) A f (4, 2) A f (4,3) A f (4, 4) A
2800
Multiple integrals 77
Figure 3.4 shows the approximating rectangular boxes and the graph
of the elliptic paraboloid.
As observed, we obtain better result for estimating the volume in
Example 3.1 if we increase the number of rectangular columns. Figure
3.5 shows the approximation of solid when we use 16 and 64 columns.
We can see that the more columns are made, the better the solid looks
and more accurate the approximate volume becomes.
Figure 3.5. Approximation with different number of rectangular
boxes
Properties of Double Integrals
3.1. [ f ( x, y ) g ( x, y )] dA f ( x, y ) dA g ( x, y ) dA
R R R
3.2. cf ( x, y) dA c f ( x, y) dA
R R
where c is a constant. This is called the linearity property of the
integral.
If f x, y g x, y for all x, y in R, then
3.3. f ( x, y) dA g ( x, y) dA
R R
78 Mathematics for mechanical engineers
3.2 Iterated Integrals
As shown in Example 3.1, it is complicated to compute the double
integral by using directly the definition of double integral. Moreover,
look at Definition 1, to compute the double integral we need to
calculate the limit of the Riemann double sum which is not a simple
task. Here we present a theorem that gives the way for calculating the
double integral which avoids computing the limit of Riemann double
sum.
Theorem 3.1 (Fubuni’s Theorem): If f is continuous on the rectangle
R = {(x, y) |a ≤ x ≤ b, c ≤ y ≤ d
then
b d d b
f ( x, y ) dA
R
a c
f ( x, y ) dy dx
c
a
f ( x, y ) dx dy
More generally, this is true if we assume that:
f is bounded on R.
f is discontinuous only on a finite number
of smooth curves.
The iterated integrals exist.
The theorem is named after the Italian mathematician Guido Fubini
(1879–1943), who proved a very general version of this theorem in
1907. However, the version for continuous functions was known to
the French mathematician Augustin-Louis Cauchy almost a century
earlier.
x 2 y dA , where
2
Example 3.2: Evaluate the double integral
R
R x, y | 0 x 1, 0 y 2
Solution: Applying Fubini’s Theorem we have:
Multiple integrals 79
1 2
( x 2 y ) dA ( x 2 2 y ) dy dx
2
0 0
R
1 y 2
x 2 y y 2 dx
0 y 0
1
1 2 x3 10
(2 x 4) dx 2
4x
0
3 0 3
We can first integrate with respect to x:
2 1
( x 2 y ) dA (x 2 y ) dx dy
2 2
0 0
R
x 1
2 x3
3 2 xy dy
0
x0
2
2 1 1 2 10
2 y dy y y
1
3 3 1 3
The results are the same whether we first integrate with respect to x or
y. It is noted that the result in this example is negative. However,
nothing is wrong with that because we do not calculate a volume, we
just calculate a given integral with non-positive function f.
3.3 Double Integrals Over General Regions
In the previous section we learned how to compute a double integral
over rectangle regions. A question raising now is: what will be
happens if the integrating region D is not rectangle? For this question,
we suppose that D is a bounded region so that D can be enclosed in a
rectangular region R as shown in Figure 3.6.
In this case we define a new function F with domain R by:
f ( x, y ) if ( x, y ) is in D
F ( x, y )
0 if ( x, y ) is in R but not in D
If F is integrable over R, then we define the double integral of f over D
by:
80 Mathematics for mechanical engineers
f ( x, y) dA F ( x, y) dA
D R
y
Now the integral of F over the non- D
rectangle region D is equal to the
integral of F over the rectangle region R
R which has been previously defined.
This means that we can choose any x
rectangle region R we want as long as Figure 3.6. A general
it contains D. region
There are different types of plane regions that give different formulas
for computing double integrals over these regions. We will classify
them into two types as follows.
3.3.1 Type I Region
Figure 3.7. Type I regions
The type I region is defined as a region which lies between the graphs
of two continuous functions of x, that is,
Multiple integrals 81
D x, y | a x b, g1 x y g 2 x
where g1 and g2 are continuous on a, b . We can see some pictures
illustrating type I regions in Figure 3.7.
To compute f ( x, y ) dA
D
over a region D of type I, we choose a
rectangle R a, b c, d that contains D and define the function F
so that F agrees with f on D and F is 0 outside D as presented in
previous. Using Fubini’s Theorem we compute the double integral as
an iterated integral:
b d
f ( x, y) dA F ( x, y) dA
D R
a c
F ( x, y ) dy dx
The explanation for this formula is addressed here. From Figure 3.8
we see that F x, y 0 if y g1 x or y g 2 x because x, y
then lies outside D. Thus,
d g2 ( x ) g2 ( x )
c
F ( x, y ) dy
g1 ( x )
F ( x, y ) dy
g1 ( x )
f ( x, y ) dy
because F f when
g1 x y g 2 x
From this result, we derive the following
formula that enables us to evaluate the
double integral as an iterated integral: If f
is continuous on a type I region D such
that
D x, y | a x b, g1 x y g 2 x Figure 3.8. Type I region
then
82 Mathematics for mechanical engineers
b g2 ( x )
f ( x, y) dA
D
a g1 ( x )
f ( x, y ) dy dx
The iterated integral in this formula is similar to the ones we have
presented. However, is should be noted that, in the inner integral, x is
considered as constant not only in f x, y but also in the limits of
integration, g1 x and g 2 x .
3.3.2 Type II region
The plane regions of type II can be defined as:
D x, y | c y d , h1 y x h2 y
where h1 and h2 are continuous. Two such regions are illustrated in
Figure 3.9.
y y
d d
𝑥 = ℎ (𝑥) D
D 𝑥 = ℎ (𝑥)
𝑥 = ℎ (𝑥) 𝑥 = ℎ (𝑥)
c c
O x O x
Figure 3.9. Type II regions
Using the same procedures described in previous, we have:
d h2 ( y )
D
f ( x, y ) dA
c
h1 ( y )
f ( x, y ) dx dy
where D is a type II region given as these figures.
Multiple integrals 83
Example 3.3: Evaluate ( x 2 y) dA , where D is the region bounded
D
by the parabol y x 2 and the line y x 6 .
Solution: The parabol and line intersect when x 2 x 6 , that is,
x 2 x – 6 0. Thus, x –2 and x 3 . Observe from Figure 3.10 that
the region D is a type I region but not a type II region. So, we can
write:
D x, y | –2 x 3, x 2 y x 6
So, by Fubini’s Theorem for type I
region:
( x 2 y) dA
D
3 x6
( x 2 y ) dy dx
2 x 2
1
[ xy y 2 ] yy xx2 6 dx
1
1
( x 4 x3 6 x 36) dx
1
3
x5 x 4
3x 2 36 x
5 4 2
625
Figure 3.10. Type I region
4
Note that, it is helpful to draw an arrow to determine the limits of
integration for the inner integral when we set up a double integral as
in Example 3.3. The arrow starts at the lower boundary y g1 x ,
which gives the lower limit in the integral. The arrow ends at the
upper boundary y g 2 x , which gives the upper limit of integration.
For a type II region, we draw arrow horizontally from the left
boundary x h1 y , which gives the lower limit in the integral, to the
right boundary y h2 y , which gives the upper limit of integral.
84 Mathematics for mechanical engineers
Example 3.4: Find the volume of the solid that lies under the plane
x 2 y z 0 and above the region D in the xy–plane bounded by the
line y 2 x and the parabola y x 2 .
Solution: From Figure 3.11 we can choose D as a type I region, so
D x, y | 0 x 2, x 2 y 2 x
Figure 3.11. Type I region
Then, the volume under z x 2 y 2 and above D is computed as
follows.
V ( x 2 y ) dA
D
2 2x
( x 2 y ) dy dx
0 x2
2 y 2 x
xy y 2 2 dx
0 yx
2
x5 x 4 28
x x 6 x dx 2 x3
2
4 3 2
0 5 4 0 5
However, the region D can also be written as a type II region, thus:
1
D x, y | 0 y 4, y x y
2
Multiple integrals 85
So, another expression for V is as follows.
V ( x 2 y ) dA
D
4 y
1
( x 2 y ) dx dy
0 2
y
x y
4 x2
2 2 xy dy
0
x 12 y
4 y y2
2 y 3/2 y 2 dy
0
2 8
4
3 4 1 28
y 3 y 5/2 y 2
8 5 4 0 5
This result of this example can be interpreted as the volume of solid,
which lies above the xy-plane, below the plane z x 2 y , and
between the plane y 2 x and the parabolic cylinder y x 2 .
Example 3.5: Find the volume of the tetrahedron bounded by the
planes: x 2 y z 3 , x 2 y , x 0 , z 0 .
Figure 3.12. The given tetrahedron
86 Mathematics for mechanical engineers
Solution: We first draw the tetrahedron T bounded by the coordinate
planes x 0 , z 0 , the vertical plane x 2 y , and the plane
x 2 y z 2 as shown in Figure 3.12.
We see that the plane x 2 y z 3 intersects the xy-plane (whose
equation is z 0 ) in the line x 2 y 3 , T lies above the triangular
region D in the xy-plane within the lines x 2 y , x 2 y 3 , x 0 .
The plane x 2 y z 3 can be written as z 3 – x – 2 y . Therefore,
the required volume lies under the graph of the function z 3 – x – 2 y
and above
D x, y | 0 x 1.5, x / 2 y 3 – x / 2 .
So,
V (3 x 2 y ) dA
D
3/ 2 3 x / 2
(3 x 2 y ) dy dx
0 x /2
3/ 2 y 3 x / 2
3 y xy y 2 y x / 2 dx
0
1 3/ 2
4 0
4 x 2 12 x 9 dx
3/ 2
1 4 x3 9
6 x 2 9 x
4 3 0 8
Example 3.6: Evaluate the iterated
sin x dx dy .
2 2
2
integral
0 y
Solution: The trouble we get here is that
it’s impossible to compute the inner
integral since sin x 2 dx is not an
elementary function. Hence, we must try
to change the order of integration. To do
that we first convert the given iterated Figure 3.13. Type I
integral into a double integral. Then we region
Multiple integrals 87
need to draw the region D and consider it as the type I region as
shown in Figure 3.13.
Using definition of the double integral backward, we have:
sin x 2 dx dy sin x 2 dA
2 2
0 y
D
where D x, y | 0 y 2, y x 2
From the graph of the region D, we see that an alternative description
of D is:
D x, y | 0 x 2, 0 y x
Since D is type I region, the double integral is expressed as an iterated
integral as follows.
sin x dA sin x 2 dy dx
2 x
2
0 0
D
y sin x 2 y 0 dx
2 yx
x sin x 2 dx
2
0
2
1 1
cos x 2 (1 cos 4)
2 0 2
Properties of Double Integrals (Continued)
With the assumption that all the following integrals exist, the first
three properties of double integrals over region D follow immediately
from the previous definition and properties.
3.4. f x, y g x, y dA f x, y dA g x, y dA
D D D
3.5 cf x, y dA c f x, y dA
D D
88 Mathematics for mechanical engineers
If f x, y g x, y for all x, y in y D
D, then:
3.6. f ( x, y) dA g ( x, y) dA
D D
D1
D2
If D D1 D2 , where D1 and D2 x
don’t overlap except perhaps on their
boundaries as shown in Figure 3.14, Figure 14. Union regions
then:
3.7. f x, y dA f x, y dA f x, y dA
D D1 D2
a) D is not the neither type I nor b) D D1 D2 , D1 is type I, D2
type II is type II
Figure 3.15. D is a union of regions of type I or type II
Property 3.7 is useful for computing double integrals over regions D
that are neither type I nor type II but can be described as a union of
regions of type I or type II as illustrated in Figure 3.15.
If f x, y 1 , then the double integration of f x, y over a region
D is equal to the area of D:
Multiple integrals 89
3.8. 1 dA A D
D
To explains this property we can look
at Figure 3.16 to see that the solid
cylinder whose base is D and whose
height is 1 has volume
V A D 1 A D . We know that
this volume can be expressed as a
double integral:
V 1 dA Figure 3.16. The solid cylinder
D
If m f x, y M for all x, y in D, then
3.9. mA( D ) f x, y dA MA D
D
Property 3.9 can be easily proved from Properties 3.5, 3.6, and 8.
3.4 Double Integrals in Polar Coordinates
When the region R in the double integral is complicated to be
described in term of rectangular coordinates but it is convenient to be
described by polar coordinates, we use the polar to evaluate the double
integral. Suppose that we want to evaluate a double integral
f ( x, y ) dA , where R is one of the regions as shown in Figure 3.17.
R
We can see in either case, the description of R in terms of rectangular
coordinates is rather complicated but R is easily described by polar
coordinates. Recall the equations relates the rectangular coordinates
x, y of a point to the polar coordinates (r , ) :
r 2 x 2 y 2 , x r cos , y r sin
The regions in the first figure are special cases of a polar rectangle as
illustrated in Figure 3.18.
90 Mathematics for mechanical engineers
R r , | 0 r a , 0 2 R r , | 0 , a r b
a0 ba0
Figure 3.17. Regions that can be described in polar coordinates
R r , | a r b,
Figure 3.18. Polar rectangles
To evaluate the double integral f ( x, y ) dA
R
with R is a polar
rectangle. We consider r and θ as the rectangle coordinates, then we
will use the same methods as presented in previous to form the double
integral. First, we divide:
The interval a, b into m subintervals ri –1 , ri of equal
width r = b – a / m .
The interval [ , ] into n subintervals [ j –1 , i ] of equal
width ( – ) / n .
Multiple integrals 91
By this way, the circles r ri and the rays i divide the polar
rectangle R into the small polar rectangles shown in Figure 3.19. Here,
the “center” of the polar subrectangle:
Rij ( r , ) | ri –1 r ri , j –1 i
has polar coordinates
ri* 1
2 ri –1 ri
*j 12 j –1 j
Then, we compute the area of Rij
using the fact that the area of a
sector of a circle with radius r and
central angle θ is 12 r 2 . We find
Figure 3.19.
that the area of Rij by subtracting
the areas of two such sectors, each of which has central angle
j – j –1 .
Ai 12 ri 2 12 ri 21
12 (ri 2 ri 21 )
12 (ri ri 1 )(ri ri 1 )
ri* r
The rectangular coordinates of the center of Rij are ri*cos *j , ri* sin *j .
So, a typical Riemann sum is:
m n
f (r
i 1 j 1
i
*
cos *j , ri* sin *j ) Ai
m n
f (ri* cos *j , ri* sin *j ) ri* r
i 1 j 1
92 Mathematics for mechanical engineers
If we want to express the Riemann sum in term of rθ- coordinates, we
put g r , rf r cos , r sin , the Riemann sum can be written as:
m n
g (r ,
i 1 j 1
i
* *
j ) r
This is a Riemann sum for the double integral in rθ- coordinates
b
a
g (r , ) dr d
Therefore, we derive:
m n
f ( x, y ) dA lim
m , n
f (r
i 1 j 1
i
*
cos *j , ri* sin *j ) Ai
R
m n
lim
m , n
g (r ,
i 1 j 1
i
* *
j ) r
b
g ( r , ) dr d
a
b
f ( r cos , r sin ) r dr d
a
So, if f is continuous on a polar rectangle R given by
0 a r b,
where 0 – 2 , then
b
f ( x, y ) dA
R
a
f (r cos , r sin )r dr d (3.3)
Formula 3.3 says that we convert from rectangular to polar
coordinates in a double integral by putting x r cos and y r sin ,
replacing dA by rdrd and using the appropriate limits of integration
for r and θ.
Multiple integrals 93
( x 2 y ) dA , where R is the region in the
2
Example 3.7: Evaluate
R
upper half-plane bounded by the circles x 2 y 2 1 and x 2 y 2 4 .
Solution: The region R can be expressed as:
R x, y | y 0, 1 x 2 y 2 4
This region is the half-ring as shown in Figure 3.20. In polar
coordinates, we have:
1 r 2, 0
R r , |1 r 2, 0
Hence, we compute the double
integral by using Fubini's Theorem in
polar coordinates:
( x 2 y ) dA
2
Figure 3.20. Half ring
R
2
(r 2 cos 2 2r sin ) r dr d
0 1
2
(r 3 cos2 2r 2 sin ) dr d
0 1
2
r4
2
cos 2 r 3 sin d
0
4 3 1
15 14
cos 2 sin d
0
4 3
15 14
1 cos 2 sin d
0
8 3
15 28
8 3
Example 3.8: Evaluate the volume of the solid bounded by: the plane
z 0 and the paraboloid z 4 – x 2 – y 2 as shown in Figure 3.21.
94 Mathematics for mechanical engineers
Solution: Let put z 0 in the equation of the paraboloid, we get
x 2 y 2 4 . This implies that the plane z 0 and the paraboloid
intersect in the circle x 2 y 2 4 . So, we need to find the volume of
solid lying under the paraboloid and above the circular disk R given
by x 2 y 2 4 . In polar coordinates, D is expressed as:
R r , | 0 r 2, 0 2
and the paraboloid is described by:
4 – x2 – y 2 4 – r 2
Figure 21. The given paraboloid
Thus, the required volume is:
V (4 x 2 y 2 ) dA
D
2 1
(4 r
2
) r dr d
0 0
1
2 1 r4 7
d (4r r ) dr 2 2r 2
3
0 0
4 0 2
Had we applied rectangular coordinates instead, we would have
obtained:
Multiple integrals 95
2 4 x2
V (4 x 2 y 2 ) dA (4 x 2 y 2 ) dy dx
2 4 x 2
D
This leads us to evaluate the
integral (4 x 2 )3/ 2 dx which is not
an elementary function.
What we have done so far can be
extended to the more complicated
type of region shown in Figure
3.22. It’s similar to the type II
rectangular regions previously
considered. If f is continuous on a
polar region of the form Figure 3.22. Complicated region
D ( r , ) | , h1 ( ) r h2 ( )
then
h2 ( )
f ( x, y) dA
D
h1 ( )
f (r cos , r sin ) r dr d
Example 3.9: Use a double integral to find the area enclosed by a half
of the cardioid r 1– cos .
Solution: From the sketch of the given curve as presented in Figure
3.23, we see that a loop is given by the region
D r , | 0 , 0 r 1 – cos
So, the area is computed as the double
integral:
Figure 3.23.
A half of the
cardioid
96 Mathematics for mechanical engineers
A( D) dA
D
1 cos
r dr d
0 0
[ 12 r 2 ]10cos d
0
1 cos
2
1
2 d
0
1
4 0
(3 cos cos 2 ) d 43
Example 3.10: Calculate the volume of the solid which is bounded by
the paraboloid z x 2 y 2 and z 4 3x 2 3 y 2 .
Solution: The paraboloid z x 2 y 2 intersects the paraboloid
z 4 3x 2 3 y 2 in the circle x 2 y 2 1.
Figure 3.24 shows that the solid lies above the paraboloid z x 2 y 2
and under the paraboloid z 4 3x 2 3 y 2 . The projection of solid
onto xy-plane is the disk D x 2 y 2 1 . In polar coordinates, we have:
Figure 3.24. Illustration of the given solid
Multiple integrals 97
x2 y2 r 2
4 3x 2 3 y 2 4 3r 2
The boundary circle becomes r 1 . Thus, the disk D is given by:
D {(r , ) | 0 2 , 0 r 1}
So, we have:
V 4 (1 x 2 y 2 ) dA
D
1 r r dr d
2 1
4 2
0 0
1
r2 r4
8 2
2 4 0
3.5 Triple Integrals
3.5.1 Definition of Triple Integrals
We just have defined double integrals for functions of two variables.
Now we can define triple integrals for functions of three variables. Let
us consider the simplest case where f is defined on a rectangular box
as shown in Figure 3.25:
Figure 3.25. Simplest region of a triple integral
98 Mathematics for mechanical engineers
B x, y , z a x b, c y d , r z s
We first divide B into sub-boxes, by dividing:
The interval a, b into l subintervals xi 1 , xi of equal width
x .
c, d into m subintervals of width y .
r , s into n subintervals of width z .
The planes through the endpoints of these subintervals parallel to the
coordinate planes divide the box B into lmn sub-boxes:
Bijk xi 1 , xi y j 1 , y j zk 1 , z k
The volume of each sub-box is V xyz (see Figure 3.26). Then,
the triple Riemann sum is computed as:
l m n
f x
i 1 j 1 k 1
*
ijk
*
, yijk *
, zijk V
where the sample point xijk
* *
, yijk *
, zijk is in Bijk .
Bij
z
x x
y
Figure 3.26. B is divided into sub-boxes
Multiple integrals 99
Similar to the definition of a double integral, we define the triple
integral as the limit of the triple Riemann sums.
Definition 3.2: The triple integral of f over the box B is:
l m n
f x, y, z dV lim
l , m, n
f x
i 1 j 1 k 1
*
ijk
*
, yijk *
, zijk V
B
if this limit exists.
We can choose the sample point to be any point in the sub-box. For
simplicity reason, we choose it to be the point xi , y j , zk that we get a
simpler-looking expression:
l m n
f x, y, z dV lim
l , m , n
f x , y , z V
i 1 j 1 k 1
i j k
B
The same for double integrals, we evaluate triple integrals by
expressing them as iterated integrals.
Theorem 3.2 (Fubini’s Theorem For Triple Integrals): If f is
continuous on the rectangular box B a, b c, d r , s , then
s d b
f x, y, z dV f x, y, z dx dy dz
B
r c a
This formula requires us to integrate in the following order:
1. With respect to x (considering y and z as constants)
2. With respect to y (considering z as a constant)
3. With respect to z
We can integrate the triple integrals by using five other possible
orders that give the same value. For example, if we can integrate with
respect to y, then z, and then x, we have:
b s d
f x, y, z dV f x, y, z dy dz dx
B
a r c
100 Mathematics for mechanical engineers
Example 3.11: Evaluate the triple integral xy 1 z dV , where B is
B
the rectangular box
B x, y , z 0 x 1, 1 y 2, 0 z 1
Solution: Let's integrate with respect to x, then y, and then z, we have:
1 2 1
xyz dV xy 1 z dx dy dz
2
0 1 0
B
x 1
1 x 2 y 1 z
2
dy dz
0 1
2 x 0
1 1 2
y 1 z dy dz
2 0 1
y 2
1 1 3 1
y 2 1 z dz 1 z dz
4 0 y 1 4 0
1
3 z 2 3
z
4 2 0 8
3.5.2 Integral Over a General Bounded Region
If the region E is not as simple as rectangular one, we need to define
the triple integral over a general bounded region E in three-
dimensional space (a solid). We can do that by much the same way
that we did for double integrals. That means, we enclose E in a box B,
then we define a function F so that it agrees with f on E but is 0 for
points in B that are outside E. Then
f x, y, z dV F x, y, z dV
E B
This integral exists if f is continuous and the boundary of E is
“reasonably smooth.”. The properties of the triple integral are
basically the same as properties of the double integral. In this text
book we just consider continuous functions f and certain simple types
of regions for triple integrals.
Multiple integrals 101
3.5.2.1 Type I Region
The solid region of type I it the region which lies between the graphs
of two continuous functions of x and y. That is,
E x, y, z x, y D, u x, y z u
1 2 x, y
where D is the projection of E onto the xy-plane.
Figure 3.27. Type I region
We notice from Figure 3.27 that the upper boundary of the solid E is
the surface with equation z u2 x, y and the lower boundary is the
surface z u1 x, y . Using the same procedures to derive the formula
for double integrals in region I, it can be shown that, if E is a type I
region, then:
u2 x , y
f x, y, z dz dA
f x, y, z dV
E D
u1 x , y
The meaning of the inner integral on the right side of Equation 3.6 is
that x and y are held fixed. Therefore, u1 x, y and u2 x, y are
regarded as constants. f x, y , z is integrated with respect to z. In
particular, if the projection D of E onto the xy-plane is a type I plane
region, then
102 Mathematics for mechanical engineers
E x, y , z a x b, g1 ( x ) y g 2 ( x ), u1 ( x, y ) z u2 ( x, y )
Therefore, the triple integral can be computed as:
b g2 ( x ) u2 ( x , y )
f x, y, z dV
E
a g1 ( x )
u1 ( x , y )
f x, y , z dz dy dx
If D is a type II plane region as shown in Figure 3.28, then
D
x
Figure 3.28. D is a type II plane region
E x, y , z c y d , h1 ( y ) x h2 ( y ), u1 ( x, y ) z u2 ( x, y )
So,
d h2 ( y ) u2 ( x , y )
f x, y, z dV
E
c h1 ( y )
u1 ( x , y )
f x, y , z dz dx dy
Example 3.12: Evaluate x dV , where E is the solid tetrahedron
E
bounded by the four planes: x 0, y 0, z 0, x 2 y z 3 .
Solution: Similar to set up a double integral, to evaluate the triple
integral it’s wise to draw two diagrams: the solid region E and its
projection D on the xy-plane.
Multiple integrals 103
Figure 3.29 shows that the upper boundary is the plane x 2 y z 3
or z 3 – x – 2 y and the lower boundary of the tetrahedron is the plane
z 0 . So, let u1 x, y 0 and u2 x, y 3 – x – 2 y .
Observe that the planes x 2 y z 3 and z 0 intersect in the line
3 x
x 2 y 3 or y in the xy-plane. Thus, the projection of E is
2
the triangular region as shown, the following description of the region
E is obtained:
E x, y, z | 0 x 3,
3 x
0 y ,
2
0 z 3 x 2 y
This description of E as a type I region
allows us to evaluate the integral as follows.
3 x
3 3 x 2 y
x dV
E
0 0
2
0
x dz dy dx
3 x
xz
3 z 3 x 2 y
2
z 0
dy dx
0 0
3 x
3
2
(3 x x 2 2 xy )dy dx
0 0
3 x
3 y
3 xy x 2 y xy 2
2
dx
0 y 0
Figure 3.29. The given
1 3 3
4 0
x 6 x 2 9 x dx solid tetrahedron
3
1 x4 9 27
2 x3 x 2
4 4 2 0 16
3.5.2.2 Type II Region
The solid region E of type II is the region which is described as:
104 Mathematics for mechanical engineers
E x, y, z y, z D, u ( y, z ) x u ( y, z )
1 2
Here, D is the projection of E onto the yz-plane as presented in Figure
3.30.
The back surface is determined by x u1 y , z and the front surface is
determined by x u2 y, z . Thus, we derive:
f x, y, z dV f x, y, z dx dA
u2 ( y , z )
(3.4)
u1 ( y , z )
E D
x E
y
Figure 3.30. Type II region
3.5.2.3 Type III Region
Finally, a type 3 region shown described in Figure 3.31 is expressed in
the form
E x, y, z x, z D, u ( x, z) y u
1 2 x, z
where:
D is the projection of E onto the xz-plane.
y u1 x, z is the left surface.
Multiple integrals 105
y u2 x, z is the right surface.
D
y
Figure 3.31. Type III region
This type of region leads to:
f x, y, z dV f x, y, z dy dA
u 2( x , z )
(3.5)
u1 ( x , z )
E D
Notice that there might be two possible expressions for the integral
depending on whether D is a type I or type II plane region using each
of Equations 3.4 and 3.5.
Example 3.13: Calculate
E
x 2 z 2 dV , where E is the region
bounded by the paraboloid y x 2 z 2 and the plane y 2 .
Solution: We draw the solid E as shown in Figure 3.32. If we consider
it as a type I region, then its projection onto the xy-plane is D1 . That is
the parabolic region shown here. The surface y x 2 z 2 projected in
the plane z 0 is the lines y x .
Solving equation y x 2 z 2 for z, yields: z y 2 x 2 .
106 Mathematics for mechanical engineers
Figure 3.32. The given region
So, the lower boundary surface of E is: z y 2 x 2 .
The upper surface is: z y2 x2 .
Thus, the region E described as a type I region is:
E x, y, z 0 y 2, y x y, y 2 x2 z y2 x2
Therefore, we have:
x z 2 dz dx dy
2 y y 2 x2
x 2 y 2 dV
2
0 y y 2 x2
E
Though this expression is correct, it is extremely difficult to evaluate.
So, instead of considering E as a type I region, we consider it as a type
III region. As such, its projection D2 onto the xz-plane is the disk
x 2 z 2 4 . We observe that the left boundary of E is the cone
y x 2 z 2 , while the right boundary is the plane y 2 . So, letting
u1 x, z x 2 z 2 and u2 x, z 2 in Equation 3.5, we obtain:
Multiple integrals 107
y 2 dV x 2 z 2 dy dA
2
x
2
E
D2
x2 z 2
D2
2 x 2 z 2 x z dA
2 2
This integral can be expressed in term of iterated integral as:
2 4 x2
2 4 x 2 2 x2 z 2 x z dz dx
2 2
However, it is still complicated to evaluate this integral in the
rectangular coordinates, but it’s easier to evaluate in polar coordinates.
Let's convert it to polar coordinates in the xz-plane:
x r cos , z r sin
That gives:
x
E
2
D2
z 2 dV 2 x 2 z 2 x z dA
2 2
2 2
2 r r r dr d
2
0 0
d 2r r dr
2 2
3 4
0 0
2
r 4 r 5 16
2
2 5 0 5
3.5.3 Applications of Triple Integrals
b
Recall that if f x 0 , then the single integral f ( x) dx describes the
a
area under the curve y f x from a to b. If f x, y 0 , the double
integral f ( x, y) dA
D
represents the volume under the surface
z f x, y and above D. However, it is difficult to interpret the
108 Mathematics for mechanical engineers
meaning of a triple integral f ( x, y, z ) dV , where f x, y , z 0 .
E
The “hypervolume” of a four-dimensional (4-D) object might be
suitable for the description of the triple integral, but we cannot sketch
it.
Nonetheless, in specific cases, the triple integral has physical
meanings which depends on the physical interpretations of x, y, z and
f x, y , z . Let’s see the case where f x, y , z 1 for all points in E.
Then, the triple integral represents the volume of E:
V E dV
E
To know the reason why it is true, we consider this in the case of a
type I region by putting f x, y , z 1 in the formula of integral:
1 dV dz dA u 2 ( x, y ) u1 ( x, y ) dA
u2 ( x , y )
u1 ( x , y )
E D D
Obviously, this presents the volume between two surfaces u1 x, y
and u2 x, y .
Example 3.14: Use a triple integral to find the volume of the
tetrahedron T bounded by the planes: x 0, y 0, z 0, x 2 y z 3
Figure 3.33. The given tetrahedron
Multiple integrals 109
Solution: The tetrahedron T and its projection D on the xy-plane are
shown in Figure 3.33. We can choose the region T as type I region and
region D as type I region. Thus, the lower boundary of T is the plane
z 0 , and the upper boundary is the plane x 2 y z 3 , that is,
z 3 – x – 2 y . So, we have:
V T dV
T
3 3 x / 2 3 x 2 y
dz dy dx
0 0 0
3 3 x / 2 9
3 x 2 y dy dx
0 0 4
Of course, since we can use the double integral to evaluate the
volumes so in practice it is not necessary to use triple integrals. This is
just an example to describe the procedure for setting up the calculation
of triple integrals.
The applications of triple integrals can be found in engineering
problems. We can see a physical application of the triple integral here.
Suppose we need to evaluate the mass of a solid object in with the
density is not a constant, but it is distributed in the object. We also
want to determine the mass, the inertia moments about three
coordinate planes, and the center of mass of the object.
We will solve the problem by using triple integral as follow. Assume
the density function of the solid object that occupies the region E is
x, y , z in units of mass per unit volume, at any given point
x, y , z . Then, its mass is:
m x, y, z dV
E
The moments of the object about the three coordinate planes can be
expressed as:
110 Mathematics for mechanical engineers
M yz x x, y, z dV
E
M xz y x, y, z dV
E
M xy z x, y, z dV
E
The center of mass is located at the point x , y , z , where:
M yz M xz M
x , y , z xy
m m m
If the density is constant, the center of mass of the solid is called the
centroid of E. The moments of inertia about the three coordinate axes
are:
I x y 2 z 2 x, y , z dV
E
I y x 2 z 2 x, y , z dV
E
I z x 2 y 2 x, y , z dV
E
3.6 Triple Integrals in Cylindrical Coordinates
Recall that in the cylindrical coordinate system, a point P in three-
dimensional (3-D) space is represented by the ordered triple (r , , z) ,
where r and θ are polar coordinates of the projection of P onto the xy–
plane, z is the directed distance from the xy-plane to P as shown in
Figure 3.34.
The equations to change from cylindrical to rectangular coordinates
are:
Multiple integrals 111
x r cos z
y r sin
zz
z
To convert from rectangular to
cylindrical coordinates the
y
following equation are used: r
r 2 x2 y2 x
tan y / x
Figure 3.34. Cylindrical coordinates
zz
Suppose that E is a type I region whose projection D on the xy-plane is
conveniently described in polar coordinates as shown in Figure 3.35.
In particular, suppose that f is continuous and
D
x
Figure 3.35. E is described in polar coordinates
E x, y , z | x, y D, u1 z u2 x, y
where D is given in polar coordinates by:
D r , | , h1 r h2
We have known that:
112 Mathematics for mechanical engineers
f ( x, y, z ) dV f x, y, z dz dA
u2 ( x , y )
E D
u1 ( x , y )
Applying the Formula 3.3 for double integrals in polar coordinates, we
have:
f x, y, z dV
E
h2 ( ) u2 r cos , r sin
f r cos , r sin , z r dz dr d
h1 ( ) u1 r cos , r sin
This is the formula for triple integration in cylindrical coordinates. It
says that to convert a triple integral from rectangular to cylindrical
coordinates we implement the flowing steps:
Writing x r cos , y r sin .
Leaving z as it is.
Using the appropriate limits of integration for z, r, and θ.
Replacing dV by rdzdrd .
Example 3.15: Figure 3.36 presents a solid which lies within: The
cylinder x 2 y 2 1, below the plane z 2 , above the cone
z x 2 y 2 . The density at any point is proportional to its distance
from the axis of the cylinder with a proportionality constant K. Find
the mass of E.
Solution: In cylindrical coordinates, the cylinder is r 1 and the cone
is z r . So, we can write:
E r , , z | 0 2 , 0 r 1, r z 2
The distance from a point (x, y, z) to the z-axis is:
f x, y , z x 2 y 2 r
Since the density is proportional to this distance with the
proportionality constant K, then the density function is:
Multiple integrals 113
f x, y, z Kr
We know that the mass of E is
calculated by the triple integral:
m x, y, z dV
E
Thus,
m K x 2 y 2 dV
E Figure 36. The given solid
2 1 2
( Kr ) r dz dr d
0 0 r
2 1
Kr 2 r dr d
2
0 0
K 2 2r r dr
1
2 3
0
1
2 r 4 5 K
2 K r 3
3 4 0 6
2 4 x2 4
Example 3.16: Evaluate
2 4 x 2
x2 y 2
x 2 dz dy dx
Solution: We can see that this iterated
integral is a triple integral, but it might
be complicated as being evaluated in the
rectangular coordinates, while it seems
to be simpler in the cylinder
coordinates. So let's sketch the graph of
the solid region. To do that we need to
convert the function and the limits of
the iterated integrals to cylinder
coordinates.
In the cylinder coordinates, we know
that x 2 y 2 r , then in the inner
Figure 3.37. The given solid
integral we have r 2 z 4 . The limits
114 Mathematics for mechanical engineers
of remaining integrals give x 2 y 2 22 , 4 x 2 y 4 x 2 ,
0 x 2 then the projection of the solid region onto the xy-plane
should be a disk x 2 y 2 4 . Thus, the solid and its projection onto
the xy-plane is sketched as shown in Figure 3.37.
Therefore, we the iterated integral is regarded as a triple integral over
the solid region:
E
{ x, y, z | 2 x 2, 4 x 2 y 4 x 2 , x 2 y 2 z 4}
This describes the solid region E in which the projection of E onto the
xy-plane is the disk x 2 y 2 4 , the upper surface of E is the
paraboloid z x 2 y 2 and the lower surface is the plane z 4 . Using
the cylindrical coordinates, we have:
E (r , , z ) | 0 2 , 0 r 2, r 2 z 4
Thus, the iterated integral can be computed as the triple integral as
follows:
2 4 x2 4
2 4 x 2 x2 y 2
x 2 dz dy dx x 2 dV
E
2 2 4
r cos 2 r dz dr d
0 0 r2
2 2 4 2 2 4
2
cos .r 2 dz dr d 2 r dz dr d
0 0 r 0 0 r2
2
2 r4
d r 4 r dr 4 2r 2 16
2
2 2
0 0
4 0
3.7 Triple Integrals in Spherical Coordinates
The spherical coordinate system is very useful coordinate system in
three dimensions and it has many applications in engineering
problems. In term of triple integral, it simplifies the evaluation of
Multiple integrals 115
triple integrals over regions bounded by spheres or cones. The
spherical coordinates ( , , ) of a point P in space consist of:
OP is the distance from the origin to P,
θ is the same angle as in cylindrical coordinates,
is the angle between the positive z-axis and the line
segment OP,
as shown in the Figure 3.39. Note that:
0
0 z
The spherical coordinate system is
especially helpful in problems where
there is symmetry about a point, and
the origin is placed at this point.
z
The equations describe the relation
between the rectangular coordinates y
and the spherical coordinates can be r
derived easily from Figure 3.38 and
is written as: x
x sin cos
Figure 3.38. Spherical coordinate
y sin sin
z cos
2 x2 y2 z 2
So the triple integral in the spherical coordinates is presented in the
following form:
f x, y, z dV
E
d b
f sin cos , sin sin , cos sin d d d
2
c a
where E is a spherical wedge given by:
116 Mathematics for mechanical engineers
E , , a b, , c d
y dV , where B lies between the spheres
2
Example 3.17: Evaluate
B
x y z 1 and x y 2 z 2 4 in the first octant as shown in
2 2 2 2
Figure 3.39.
Figure 3.39. Sketch of the given solid
Solution: Since the boundary of the solid region B is a sphere, we use
spherical coordinates:
B , , |1 2, 0 , 0
2 2
In addition, spherical coordinates are appropriate because:
x2 y 2 z 2 2
Using the iterated integrals in spherical coordinates, we have:
2
y dV 2 sin 2 sin 2 2 sin d d d
2 2 2
0 0 1
B
2
2 sin 3 d 2 sin 2 d 4 d
0 0 1
Multiple integrals 117
2
1 sin 2 2 5
cos cos
3 2
0 2
2 0 5 1
2 31 31
34 5 30
Example 3.18: Use spherical
coordinates to find the volume of
the solid that lies within the sphere
x 2 y 2 z 2 4 as presented in
Figure 3.40:
Above the xy – plane
Below the cone
z x2 y2
Solution: We write its equation in
spherical coordinates as: 2 . Figure 3.40. The given solid
The equation of the cone can be
written as:
cos 2 sin 2 cos 2 2 sin 2 sin 2 sin
This gives: sin cos or .
4
Thus, the description of the solid E in spherical coordinates is:
E , , 0 2 , , 0 2
4 2
So, the volume of E is:
V ( E ) dV
E
2 /2 2
2 sin d d d
0 /4 0
118 Mathematics for mechanical engineers
2
2 /2 3
d sin d
0 /4
3 0
16 / 2
3 /4
sin d
16
cos /4
/2
3
8 2
3
3.8. Change of Variables in Multiple Integrals
In one-dimensional calculus we often use a change of variable to
simplify an integral as:
b d
a
f ( x) dx f ( g (u)) g '(u) du
c
or
b d dx
a
f ( x) dx f ( x(u ))
c du
du
where x g u and a g c , b g d .
The change of variables can also be useful in multiple integrals. We
will see what happens when we change the variables in multiple
integrals.
3.8.1 Change of Variables in Double Integrals
We have already seen one example of this: conversion to polar
coordinates. The variables r and θ in the polar coordinates are
converted to the variables x and y in the rectangular coordinates by the
equations:
x r cos , y r sin
Multiple integrals 119
Then the change of variables formula can be written as:
f ( x, y) dA f (r cos , r sin ) r dr d
R S
where S is the region in the rθ-plane that corresponds to the region R
in the xy-plane.
Now we consider a change of variables in general that is given by a
transformation T from the uv-plane to the xy-plane:
T u , v x, y
where x and y are related to u and v by relationships:
x g u, v , y h u, v (3.6)
These can be written as:
x x u, v , y y u, v
v y
T (x1, y1)
S
(u1, v1)
R
T-1
u x
Figure 3.41. Effect of a transformation T on a region S in the uv-plane
Assume that T is a C1 transformation so that g and h have continuous
first-order partial derivatives. Therefore, a transformation T is really
just a function whose domain and range are both subsets of 2 . If we
120 Mathematics for mechanical engineers
convert a point u1 , v1 from the uv-plane to point x1 , y1 in the xy-
plane by a transformation T u1 , v1 x1 , y1 , then we call point
x1 , y1
the image of the point u1 , v1 . If no two points have the same
image, T is called one-to-one transformation. Figure 3.41 shows the
effect of a transformation T on a region S in the uv-plane. We see that
T transforms S into a region R in the xy-plane. We call R the image of
S which consists of the images of all points in S.
Let T be a one-to-one transformation, it has an inverse transformation
T–1 from the xy-plane to the uv-plane. Thus, we can solve Equations
3.6 for u and v in terms of x and y :
u G x, y , v H x, y
Example 3.19: Given a transformation is defined by:
u v 2u v
x , y
3 3
Find the image of the square S u , v | 0 u 4, 3 v 3
Solution: The transformation maps the boundary of S into the
boundary of the image as shown in Figure 3.42. So we need to find the
image of the sides of S.
In the first side S1 we have:
v 3, 0 u 4
From the given equations, we have:
u 3 2u 3
x , y .
3 3
Eliminating u, we obtain: Figure 3.42. S region
y 2 x 3, 1 x 7 / 3
Multiple integrals 121
7 5
which is the line segment from (1, –1) to , in the xy-plane.
3 3
The second side, S2, is:
u 4, 3 v 3
Putting u = 4 in the given equations, we get:
4v 8v
x , y
3 3
Eliminating v, we obtain:
y x 4, 1/ 3 x 7 / 3
7 5 1 11
which is the line segment from , to , in the xy-plane.
3 3 3 3
Similarly, S3 is given by:
v 3, 0 u 4
Its image is the line segment y 2 x 3, 1 x 1 / 3 from 1,1 to
1 11
, in the xy-plane.
3 3
Finally, S4 is given by:
u 0, –3 v 3
whose image is line segment y x, 1 x 1 from 1,1 to
1, 1 in the xy-plane.
The image of S is the region R bounded by four line segments which
shown in Figure 3.43.
122 Mathematics for mechanical engineers
Figure 3.43. S and its image R
Now, we will see how a change of variables influences a double
integral. Let's consider a small rectangle S in the uv-plane whose
lower left corner is the point u0 , v0 and whose dimensions are u
and v as illustrated in Figure 3.44. You can see that the image of S is
a region R in the xy-plane, one of whose boundary points is:
Figure 3.44. A region S and its image R
x0 , y0 T u0 , v0
In the xy-plane, the vector r u , v g u , v i h u , v j is the position
vector of the image of the point u , v .
Multiple integrals 123
Look at the lower side of S we see
that its equation is v v0 and its
image curve is given by the vector
function r u , v0 . The tangent vector
at x0 , y0 to this image curve is:
ru gu (u0 , v0 ) i hu (u0 , v0 ) j
x y
i j
u u Figure 3.45. Image R of S
Similarly, the tangent vector at x0 , y0 to the image curve of the left
side of S where u u0 is:
rv gv (u0 , v0 ) i hv (u0 , v0 ) j
x y
i j
v v
The image region R T S can be approximated by a parallelogram
determined by the secant vectors as shown in Figure 3.45:
a r (u0 u , v0 ) r (u0 , v0 )
b r (u0 , v0 v) r (u0 , v0 )
Remember that,
r (u0 u , v0 ) r (u0 , v0 )
ru lim
u 0 u
So,
r(u0 u, v0 ) r(u0 , v0 ) u ru
Figure 3.46. Approximation of R
Similarly, we get:
r(u0 , v0 v) r(u0 , v0 ) v rv
124 Mathematics for mechanical engineers
That is, we can approximate R by a parallelogram determined by the
vectors u ru and v rv as demonstrated in Figure 3.46. Thus, the area
of R can be approximated by the area of this parallelogram:
uru vrv ru rv uv
Computing the cross product, we get:
i j k x y x x
x y u u u v
ru rv 0 k k
u u x y y y
x y v u u v
0
v u
The determinant that arises in this calculation is called the Jacobian of
the transformation. We give the definition of the Jacobian of the
transformation T as follows.
Definition 3.3: The Jacobian of the transformation T given by
x g u , v and y h u , v is:
x x
( x, y ) u v x y x y
(u, v) y y u v v u
u v
With this notation, we can give an approximation to the area A of R:
( x, y )
A u v (3.7)
(u , v )
where the Jacobian is evaluated at u0 , v0 . Next, we divide a region S
in the uv-plane into small rectangles Sij and call their images in the xy-
plane Rij . From Equation 3.7 we have:
Multiple integrals 125
( x, y )
Rij Sij
(u , v)
So we can approximate the double integral of f over R as follows.
f ( x, y ) dA
R
m n
f ( xi , y j ) A
i 1 j 1
m n
( x, y )
f ( g (ui , v j ), h(ui , v j )) u v
i 1 j 1 (u, v)
where the Jacobian is evaluated at ui , v j . Notice that this double sum
is a Riemann sum for the integral:
( x, y )
f ( g (u, v), h(u, v)) (u, v) du dv
S
Theorem 3.3: Suppose that:
T is a C1 transformation whose Jacobian is nonzero and that
maps a region S in the uv-plane onto a region R in the xy-
plane.
f is continuous on R and that R and S are type I or type II
plane regions.
T is one-to-one, except perhaps on the boundary of S.
Then,
( x, y )
f ( x, y ) dA f ( x(u, v), y(u, v)) (u, v) du dv
R S
This theorem tells us that if we change from an double integral in xy-
plane to an integral in uv-plane the variables x and y are expressed in
terms of u and v and:
126 Mathematics for mechanical engineers
( x, y )
dA du dv
(u , v)
Recall the single integration:
b d dx
a
f ( x) dx f ( x(u ))
c du
du
We notice the similarity between the single and the double integrals:
dx
Instead of the derivative , we have the absolute value of the
du
( x, y )
Jacobian .
(u , v)
Figure 3.47. Transformation T from the xy-plane to the rθ-plane
Looking back at the double integral in the polar coordinates, we see
that the formula for integration in polar coordinates is just a special
case. Using Theorem 3.3 we can derive the same formula for the
double integral in the polar coordinates as presented in Figure 3.47.
Indeed, the transformation T from the rθ-plane to the xy-plane is given
by:
x g r , r cos
y h r , r sin
Multiple integrals 127
The geometry of the transformation is shown here. T maps an ordinary
rectangle in the rθ-plane to a polar rectangle in the xy-plane. The
Jacobian of T is:
x x
( x, y ) r
(r , ) y x
r
cos r sin
r cos 2 r sin 2 r 0
sin r cos
So, from the Theorem 3.3,
( x, y )
f ( x, y) dx dy f (r cos , r sin ) (u, v) dr d
R S
b
f (r cos , r sin ) r dr d
a
This is the same as the formula for double integral in polar coordinate.
Example 3.20: Use the change of variables x u 2 – v, y =uv to
evaluate the integral y dA , where R is the region bounded by:
R
The x-axis.
The parabolic y 2 4 x 8 and the line segment
y 3x 27 , y 0 .
Solution:
Figure 3.48 shows that R is not an easy region for evaluating the
integral. Thus, we need to change variables to evaluate the given
integral. We knew that T S R where S is the rectangular
0, 3 0, 2 . So, we see that S is a much simpler region than R. To
derive the formula of the double integral uv-plane we compute
Jacobian of the variable change:
128 Mathematics for mechanical engineers
Figure 3.48. Sketch of the given region
x x
( x, y ) u v 2u 1
2u 2 v 0, v 0
(u , v ) y y v u
u v
Thus, by Theorem 3.3,
( x, y )
y dA uv (u, v) dA
R S
2 3
(uv)(2u 2 v) du dv
0 0
2 3
(2u 3v uv 2 ) du dv
0 0
u 3
1
2 1
u 4 v u 2v 2 dv
0 2 2
u 0
2 81 9
v v 2 dv
0
2 2
2
81 3
v 2 v 3 93
4 2 0
This example show that the derived integral was not very difficult to
solve as we were given a suitable change of variables. However, if we
are not supplied with a transformation, we need to think of an
appropriate change of variables.
Multiple integrals 129
Example 3.21: Evaluate the integral x y dA , where R
R
is the
parallelogram region bounded by lines y 2 x 3 , y 2 x 3 , y x
and y x 4 .
Solution: The region R is pictured
in Figure 3.49. We make a change
of variables suggested by the form
of this function:
u x y, v – 2 x y
These equations define a
transformation T–1 from the xy-
plane to the uv-plane. The
theorem talks about a
transformation T from the uv-
plane to the xy-plane. It is Figure 3.49. The given region
obtained by solving the above
equations for x and y:
uv 2u v
x , y
3 3
The Jacobian of T is:
x x 1 1
( x, y ) u v 3 3 1
(u , v ) y y 2 1 3
u v 3 3
To find the region S in the uv-plane corresponding to R, we note that:
The sides of R lie on the lines: 2 x y 3 , 2 x y 3 ,
x y 0 , x y 4 .
From either Equations 3.10 or Equations 3.11, the image
lines in the uv-plane are: v 3, v 3, u 0, u 4 .
130 Mathematics for mechanical engineers
Thus, the region S is the rectangular region with vertices 0, –3 ,
4, –3 , 4,3 , 0,3 as presented in Figure 3.50.
S u , v | –3 v 3, 0 u 4
Figure 3.50. The image of R is a rectangular region
So, by Theorem 3.3,
( x, y )
x y dA u
R S
(u, v)
du dv
1 3 4
3 3 0
u du dv
1 3 4
dv udu
3 3 0
4
1 u2
3 3 16
3 2 0
3.8.2 Change of Variables in Triple Integrals
We can extend the change of variables for double integrals to triple
integrals. Suppose T is a transformation that maps a region S in uvw-
space onto a region R in xyz-space by the following equations:
Multiple integrals 131
x g u , v , w , y h u , v, w , z k u , v , w
The Jacobian of T is the 3 x 3 determinant:
x x x
u v w
( x, y , z ) y y y
(u, v, w) u v w
z z z
u v w
Under hypotheses similar to those in Theorem 3.3, we have the
formula for triple integrals:
f ( x, y, z ) dV
R
( x, y , z )
f ( x (u , v, w), y (u, v, w), z (u, v, w)) du dv dw
S
(u , v , w )
Now we derive the formula for triple integration in spherical
coordinates by using this formula. Let's change the variables given by
equations:
x sin cos
y sin sin
z cos
We compute the Jacobian as follows:
sin cos sin sin cos cos
( x, y, z )
sin sin sin cos cos sin
( , , )
cos 0 sin
sin sin cos cos
cos
sin cos cos sin
132 Mathematics for mechanical engineers
sin cos sin sin
sin
sin sin sin cos
cos ( sin cos sin 2 2 sin cos cos 2 )
2
sin ( sin 2 cos 2 sin 2 sin 2 )
2 sin cos 2 2 sin sin 2
2 sin
As 0 , we have sin 0 . Thus,
( x, y, z )
2 sin 2 sin
( , , )
This leads to:
f ( x, y, z) dV
R
f ( sin cos , sin sin , cos ) 2 sin d d d
S
Obviously, we can see that this is the same with the formula for triple
integrals in spherical coordinates as presented in previous.
Multiple integrals 133
EXERCISE
I. Calculate the iterated integral:
2 1
1. 1 2xy dxdy
1 0
2x xy dxdy
1 3
3
2.
1 0
x y 2 dxdy
2 2
2
3.
0 0
1
4.
0 0
x sin ydxdy
ln 2 ln 5
5.
0 0
e 2 x y dxdy
II. Calculate the double integral:
5 x y 2 2 y 3 dA , R x, y | 0 x 2, 0 y 2
2
1.
R
2. sin x y dA , R x, y | 0 x , 0 y
R
xy 2
3. R x2 1 dA , R x, y | 0 x 1, 0 y 5
1 2x2
4. R 2 y 2 dA , R x, y | 0 x 2, 0 y 2
xye xy dA , R x, y | 0 x 2, 0 y 1
2
5.
R
6. x 2 y 3 dA , R x, y | 0 x 2, 0 y x
R
2y
7. dA , R x, y | 0 x 2 y , 0 y 1
R x 2 3
134 Mathematics for mechanical engineers
8) e 2 x / y dA , R x, y | y x y 3 , 0 y 1
R
III. Find the volume of the solid bounded by the surfaces:
1. z x x 2 y 2 and the planes x 0, x 2, y 0, y 2 .
2. z 2 e y sin x and the planes x 1, y 1 .
3. z 1 x 2 y 2 and the planes x 0, x 1, y 0, y 3 .
4. Under the paraboloid z x 2 y 2 and above the disk x 2 y 2 4 .
5. Inside the sphere x 2 y 2 z 2 4 outside the cylinder x 2 y 2 1.
IV. Sketch the region of integration and change the order of
integration:
4 x
1. f x, y dydx
0 0
1 x
2. f x, y dydx
0 0
1 x
3. f x, y dydx
0 x2
1 x2
4. f x, y dydx
0 x4
V. Evaluate the triple integral:
1. E
xdV , E x, y, z |0 y 2, 0 x
1 y2 , 0 z y
2. xydV , E x, y , z |0 x 1, 0 y x, x z 3 x
E
3. yz cos x 3 dV , E x, y , z |0 x 1, 0 y x, 0 z y 2
E
Multiple integrals 135
4. E
xy 2 dV , where E lies under the plane z 2 x y and above
the region in the xy-plane bounded by the curves y x , y 0 and
x 1.
VI. Use a triple integral to find the volume of the given solid E:
1. The tetrahedron enclosed by the coordinate planes and the plane
2x 2 y z 3 .
2. The solid bounded by the cylinder y x 2 and the planes
z 0, z 1, y 4 .
3. The solid bounded by paraboloids z x 2 y 2 and z 6 2 x 2 2 y 2
VII. Use spherical coordinates to:
1. Find the volume of the solid E that lies above the cone
z x 2 y 2 and below the sphere x 2 y 2 z 2 4 .
2. Evaluate E
zdV , where E lies above the paraboloid z x 2 y 2
and below the plane z 2 y .
VIII. Evaluate the integral by changing to spherical coordinates:
2 4 y2 4 x2 y 2
1.
2 4 y 2 0
z x 2 y 2 z 2 dzdxdy
x y 2 z 2 dzdxdy
1 9 y 2 16 x 2 y 2
2
2.
0 9 y2 x2 y 2
ANSWERS
I.
2. 81
136 Mathematics for mechanical engineers
1
4. 2 sin 2
2
II.
2. sin x y dA 0
R
1 2 x2
4.
11 2
R 2 y 2 dA 3 arctan 2
32
6. R
x 2 y 3 dA
7
1
8. R
e2 x / y dA
8
1 e2
III.
2. V 8
4. V 8
IV.
1 1
2. f x, y dxdy
0 y
1 4 y
4. f x, y dxdy
0 y
V.
1
2. E
xydV
2
1 1 2 x y 989
4. E
xy 2 dV
0
y2 0
xy 2 dzdxdy
3024
Multiple integrals 137
VI. Hint: Use the formula V 1dV .
E
VII. 1. Hint: Use the formula V 1dV .
E
2y
zdV 2 zdzdA , where D is the disk x 2 y 1 1
2
2.
E x y2
D
VIII. 1. Hint: The integral domain in rectangular coordinates is:
E x, y, z |
2 y 2, 4 y 2 x 4 y 2 , 0 z 4 x 2 y 2
This describes the upper half of a sphere with radius of 2. Thus the
domain in the spherical coordinates will be:
E , , |0 2, 0 , 0 2 .
2
138 Mathematics for mechanical engineers
Differential forms 139
CHAPTER 4. DIFFERENTIAL FORMS
The calculus of differential forms gives an alternative to vector
calculus which is ultimately simpler and more flexible. In this chapter,
a brief supplement to explain how to work with them is presented.
4.1 1-forms
Definition 4.1: A differential 1-form on an open subset of 2 is an
expression:
F x, y, z dx G x, y, z dy
where F, G are -valued functions on the open set.
A very important example of a differential is given as follows: If
f x, y is C1 -valued function on an open set U, then its total
differential (or exterior derivative) is:
f f
df dx dy
x y
In a similar manner, a differential 1-form on an open subset of 3 is
an expression:
F x, y , z dx G x, y , z dy H x, y , z dz
where F, G, H are -valued functions on the open set. If f x, y , z is
a C1 function on this set, then its total differential is:
f f f
df dx dy dz (4.1)
x y z
It is noted that, a differential form is very similar to a vector field. In
fact, we can set up a correspondence as follows:
Fi Gj Hk Fdx Gdy Hdz
140 Mathematics for mechanical engineers
Where i, j, k are the standard unit vectors along the x, y, z axes. Under
this set up, the gradient f corresponds to df. The differential notation
is very suggestive and ultimately quite powerful. We can see that we
can write the previous concepts using by using differential notations.
For example, we can see how to derive Equation 4.1 just by using
differential notations. Suppose that that variables x, y, z depend on
some parameter t, and f depends on x, y, z, then the chain rule says:
f f x f y f z
t x t y t z t
Thus, the formula for df can be obtained by canceling the notation dt:
f f f
df dx dy dz
x y z
This formula is the same with Equation 4.1.
4.2 Exactness in 2
Definition 4.2: Suppose that, Fdx Gdy is a differential form with C1
coefficients. We say that it is exact if one can find a C2 function
f x, y , i.e. function f x, y has second derivative, so that:
df Fdx Gdy (4.2)
However, most differential forms are not exact. To see why, we notice
that Equation 4.2 is equivalent to:
f f
F ,G
x y
Therefore, if f exists then
F 2 f 2 f G
(4.3)
y yx xy x
Differential forms 141
But Equation 4.3 would fail for most examples such as x3 ydx e y dy .
Indeed, we can see that:
F x3 y , G e y
This lead to:
F G
x3 , 0
y x
Thus:
F G
y x
Therefore, the 1-form differential x3 ydx e y dy is not exact.
F G
Definition 4.3: We call a differential Fdx Gdy closed if and
y x
are equal.
Obviously, we have just shown that if a differential is to be exact, then
it had better be closed. Exactness is a very important concept. You
may already have seen it in differential equations. For example, if we
are given a differential equation:
dy F x, y
dx G x, y
so we can express this in 1-form as:
Fdx Gdy 0
If Fdx Gdy is exact then there exists a function f so that:
df Fdx Gdy (4.4)
142 Mathematics for mechanical engineers
This is equivalent to df 0 , so the solution of Equation 4.4 should be
curves:
f x df C
where C is integral constant.
Another example is in physics. Suppose a vector field F F1i F2 j
represents a force field, if there exists a function P x, y such that
F P , then P x, y is called the potential energy. The force is
called conservative if it has a potential energy function. In terms of
differential forms, vector field F is conservative precisely when
F1dx F2 dy is exact.
Theorem 4.1: If F x, y dx G x, y dy is a closed form on all of 2
with C1 coefficients, then it is exact.
This is the converse statement that we have stated: a differential form
is exact then it is closed.
Proof: To prove Theorem 4.1 we would need to find a function
f x, y so that:
df Fdx Gdy
In other words, we need to solve this differential equation which
should clearly involve some kind of integration process. To do this,
we first have to choose a parametric C1 curve C. Then this lead to the
definition of the line integrations.
Line Integrals
Definition 4.4: The line integration in 2 is
b dx dy
Fdx Gdy F x t , y t G x t , y t dt
C a
dt dt
Differential forms 143
Lemma 4.1:
C
Fdx Gdy Fdx Gdy
C
The line integral in 3 is defined in similar manner.
Definition 4.5: Suppose that F x, y dx G x, y dy is a differential
form with C1 coefficients. Let C : a, b 3 be a piecewise C1
parametric curve, then the line integration in 3 is:
C
Fdx Gdy Hdz
dx dy
b
a
F x t , y t , z t G x t , y t , z t
dt dt
dz
H x t , y t , z t dt
dt
The exactness extends naturally to 3 : a form is exact if it is equal to
df for a C 2 function. One of the most important properties of exactness
is its path independence.
Proposition 4.1: If a 1- form Fdx Gdy is exact and C1 and C2
are two parametrized curves with the same endpoints (or more
accurately the same starting point and ending point), then:
C1
C2
This can be easily proved using the definition of the line integral.
Indeed, if df and C1 : a, b 2 then:
b df
C1
df
a dt
dt
The Fundamental Theorem of Calculus shows that:
144 Mathematics for mechanical engineers
df
dt f C1 b f C1 a
b
a dt
A similar setting up for the line integral over C2 gives same answer:
df
a
b
df dt f C 2 b f C 2
C2 a dt
since the two curves C1 and C2 have the same endpoints.
If the C is closed as shown in Figure 4.1, which means that the starting
point coincides the endpoint, then we have:
b
C1
df df
C2 C1
f C 2 b f C 2 a
f C1 b f C 1 a 0 a C2
This leads to the following corollary.
Figure 4.1. C is a closed curve
Corollary 4.1: If is exact and C is
closed, then 0 .
C
Now back to the theorem: if Fdx Gdy is closed form on 2 , we
choose the curve C as seen in Figure 4.1
and set:
f x, y Fdx Gdy
C
We parameterize both line segments
separately as: x t , y y0 and
x constant, y t as depicted in Figure
4.2. Thus, Figure 4.2. Two line segments
x y
f x, y F t , y0 dt G x, t dt
x0 y0
Differential forms 145
We need to prove that:
df Fdx Gdy
or
f f
F, G
x y
Using the Fundamental Theorem of Calculus, we have:
f y
G x, y dy G x, y
y y 0
On the other hand:
f x y
F t , y0 dt G x, t dt
x x 0 x x y0
y G x, t
F x , y0 dt
y0 x
Since Fdx Gdy is closed then:
G x, y F x, y F x, t
x y t
Thus,
f y F x, t
F x, y0 dt
x y0 t
F x, y0 F x, y F x, y0 F x, y
So,
f f
df dx dy F x, y dx G x, y dy
x y
146 Mathematics for mechanical engineers
Theorem 4.1 is proved.
Line integrals have many important applications in engineering fields
involving: forces, electricity, magnetism, fluid flow. One very direct
example is the work done by a force when it moves an object over a
distance. For instance, when you pick up a rock off the ground it takes
energy. If you move up the rock in straight line over a distance
represented by a vector d 0, 0, z by a vertical force F 0, 0, F3
as presented in the left picture of Figure 4.3. Then your work done is
simply expressed as:
F d F3z
Figure 4.3. Straight and parabolic trajectory
However, consider the case that a ball is kicked into the sky. Under
the gravitation field, the ball will finally fall toward the ground. If the
trajectory of the ball is not straight, but it is a parabol C as shown in
the right picture of Figure 4.3. The gravitation field acting on the ball
presented by a vector F F1 , F2 , F3 is not constant, then the work
done by the gravitational field would be the integral line:
C
F1dx F2 dy F3 dz
or
Differential forms 147
C
F ds
Here we can think of ds as the vector dx, dy, dz .
From this example, we can see that the differences between the
differential forms and vectors: To understand the nature of physics,
equations are expressed by coordinates, while the general principles
behind are often differential forms. The differential form represents a
coordinate free way of dealing with differentiation.
Example 4.1: Evaluate work done by the gravitational field as
mMGx mMGz
presented in Chapter 1, Fx ( x, z ) 2 , Fz ( x, z ) 2 2 3/ 2
(x z )
2 3/2
(x z )
when a ball moves on the parabola trajectory z 1 x 2 , 1 x 1 as
shown in the Figure 4.3.
Solution: We parametrize x t , z 1 t 2 , dx dt , dz 2tdt then
using the line integral formula for the work done, we have:
mMGt mMG 1 t 2
2t dt
1
C Fx dx Fz dz 1 2
t 1 t t 1 t
3/ 2 3/ 2
2 2 2
mMGt 2tmMG 1 t 2 dt
1
1
1
mMGt mMG 2t 3 dt
1
1 1
mMGt 2 mMG 2t 4
0
2 1
4 1
The work done by the gravitational field on the ball is zero can be
explained by as follows. We can see that the trajectory is symmetric
with respect to the z-axis. The ball moves up in the first half trajectory
so the work done by the gravitational field is negative. While, the ball
moves down in the second half trajectory so the work done is positive.
Thus, the works done in the first half of trajectory is equal to that of
the second trajectory but with opposite sign, so that the total works in
both half of trajectory is zero.
148 Mathematics for mechanical engineers
4.3 2-form
The cross product of vectors u v is a very useful operation in 3
dimensional geometry. It determines the area of the parallelogram
containing these vectors and the plane containing it. While there is no
direct analogue of the cross product in higher dimensions. The wedge
product is presented to deal with this problem. Also, the wedge
product can be used to express the dot products, cross products, ect.,
in a simple way. The wedge product together with gradient, curl,
divergence operators, etc., give elegant formulations of topics such as
Jacobian of a transformation and the integral theorems.
Definition 4.6 (Wedge Product): Given (row) vectors u, v n , the
wedge product of is u and v is the matrix:
1 T
u v
2
u v vT u
In 3 we have:
0 ae bd af cd
1
a, b, c d , e, f ae bd 0 bf ce
2
af cd bf ce 0
We can see that the nonzero entries are basically the same as for the
cross product:
a, b, c d , e, f bf ce i af cd j ae bd k
So these two operations are in some sense equivalent. The big
difference is, of course, that the wedge product produces a matrix, but
not just any matrix. Recall that a matrix A aij is a skew symmetric
if AT A , i.e. a ji aij we can see that the wedge product is a skew
symmetric matrix.
Theorem 4.2: The wedge product of two vectors lies in 2 n , where
2 n denotes the space of all skew symmetric n n real matrices.
Differential forms 149
When n 3 we can use standard calculations from linear algebra to
obtain the theorem as:
1 T 1
u v u v vT u vT uTT uT vTT
T T
2 2
1 T 1
v u uT v uT v vT u u v
2 2
Also, the following properties of wedge product can be verified
directly from Definition 4.1:
u v v u
uu 0
c u v cu v u cv
u v w u v u w
These rules are very useful to work with wedge products rather than
the Definition 1.
Now we will use the wedge product to define the 2-form which gives
a convenient way to deal with essential geometric features of the cross
product and surface integrals. Suppose F1dx G1dy H1dz and
F2 dx G2 dy H 2 dz are two 1-forms. We express the wedge product
of these two 1-forms with the use of wedge product properties as
follows:
F1dx G1dy H1dz F2 dx G2 dy H 2 dz
F1 F2 dx dx G1G2 dy dy H1 H 2 dz dz
F1G2 dx dy G1 H 2 dy dz H 1 F2 dz dx
G1 F2 dy dx H1G2 dz dy F1 H 2 dx dz
150 Mathematics for mechanical engineers
F1G2 dx dy G1 H 2 dy dz H1 F2 dz dx u u 0
G1 F2 dx dy H1G2 dy dz F1 H 2 dz dx u v v u
F1G2 G1 F dx dy G1 H 2 H1G2 dy dz
H1 F2 F1 H 2 dz dx
F x, y, z dx dy G x, y, z dy dz H x, y, z dz dx
This leads to the following definition for 2-form differential.
Definition 4.7: A 2-form is an expression which is built using wedge
products of pairs of 1-form. On R3, it is expresses as:
F x, y , z dx dy G x, y , z dy dz H x, y , z dz dx
Where F, G, H are functions defined on an open subset of R3. Any
wedge production of two 1-form can be put in this format. For
example, using the above rules, we can see that:
x dx ydy dx 2dy
3
x3dx dx 2 x3 dx dy ydy dx 2 ydy dy
2 x 3 y dx dy
4.4. Exactness in 3 and conservation of energy
A C1 1-form Fdx Gdy Hdz is called exact if there is a C 2
function f (called a potential) such that df . A 1-form is called
closed if d 0 , or equivalently if
Fy Gx ; Fz H x ; Gz H y
These equations must hold when
F fx ; G f y ; H fz
Theorem 4.3: Exact 1-forms are closed.
Differential forms 151
Theorem 4.4: If Fdx Gdy Hdz is a closed form on 3 with C1
coecients, then is exact. In fact if f x0 , y0 , z0 , where C is
C
any piecewise C 1
curve connecting 0, 0, 0 to x0 , y0 , z0 , then
df .
This can be interpreted in vector fields as follows. A C1 force field
F Fi Gj Hk is called conservative if there is a C 2 real valued
function P, called potential energy, such that F P . This implies
that a force F, which is C1 on all of 3 is conservative if and only if
F 0 . P x, y , z is the work done in moving an object of unit
mass along a path connecting 0, 0, 0 to x, y , z .
4.4.1 Convert a Vector Field to a 2-Form
In previous we knew how to convert the expression of a vector field to
a 1-forms:
F1i F2 j F3k F1dx F2 dy F3dz
Now, we can also associate the expression of a vector field to 2-forms
as follows:
F1i F2 j F3k F1dy dz F2 dz dx F3dx dy
Here we relate i to dy dz , j to dz dx , k to dx dy . We can convert a
2-form to a vector by replacing dx, dy, dz by i, j, k and taking their
cross products.
F1dy dz F2 dz dx F3dx dy
F1 j k F2k i F3i j=F1i F2 j F3k
As a memory aid to interchange 1-forms and 2-forms, the Hodge star
operator is presented here:
152 Mathematics for mechanical engineers
F1dx F2 dy F3 dz F1dy dz F2 dz dx F3dx dy
F1dy dz F2 dz dx F3dx dy F1dx F2 dy F3 dz
4.4.2 Derivative of 1-Forms
Consider a 1-form:
F x, y , z dx G x, y , z dy H x, y , z dz
If we want to define its derivative d which will be a 2-form, we
suppose that and are 1-forms and f is a function, then we use the
following rules:
d d d
d f df fd
d dx d dy d dz 0
Recall that:
df f x dx f y dy f z dz
So
d Fdx Gdy Hdz
dF dx Fd dx dG dy Gd dy dH dz Hd dz
Fx dx Fy dy Fz dz dx Gx dx Gy dy Gz dz dy
H x dx H y dy H z dz dz
Gx Fy dx dy H y Gz dy dz Fz H x dz dx
4.4.3 Curl and Derivative of 1-Forms
Let's see the relationship between curl and derivative of 1-forms. We
start with a vector field V Fi Gj Hk , then replace it by 1-form
Fdx Gdy Hdz and apply the derivative d:
Differential forms 153
d Fdx Gdy Hdz
Gx Fy dx dy H y Gz dy dz Fz H x dz dx
Convert this to vector field, we end up with the curl of V as presented
in Chapter 2.
G x Fy k H y Gz i Fz H x j= V
Thus, the derivative d of a 1-form is equivalent to curl of a vector.
4.4.4 Derivative of 2-Forms
Earlier we defined wedge product of pairs of vectors. Now we extend
it to triple vectors:
u v w v u w v w u ...
Notice that:
u vw 0
if any two of the vectors are equal. In addition, we suppose the
following laws:
u1 u 2 v w u1 v w u 2 v w
d dx dy d dy dz d dz dx 0
dx dy dz dy dx dz dy dz dx ...
We call any expression f x, y , z dx dy dz a 3-form. We now take
the derivative of a 2-form Fdy dz Gdz dx Hdx dy to obtain a
3-form as follows:
d Fdy dz Gdz dx Hdx dy
Fx dx Fy dy Fz dz dy dz
Gx dx Gy dy Gz dz dz dx H x dx H y dy H z dz dx dy
154 Mathematics for mechanical engineers
Fx dx dy dz Fy dy dy dz Fz dz dy dz
Gx dx dz dx Gy dy dz dx Gz dz dz dx
H x dx dx dy H y dy dx dy H z dz dx dy
Fx dx dy dz Gy dx dy dz H z dx dy dz
Fx Gy H z dx dy dz
This says that the derivative of 2-forms gives 3-forms.
4.4.5 Surface Integrals
Suppose S is a smooth parameterized surface:
x f u, v
y g u, v
z h u, v
u, v D
with orientation corresponding to the ordering u, v. The symbols dx,
dy etc. can be converted to the new coordinates as follows:
x x
dx u v
u v
y y
dy u v
u v
Thus,
x x y y
dx dy u v u v
u v u v
x y x y
du dv
u v v u
x, y
du dv
u, v
Differential forms 155
x, y
Notice that is Jacobian of the variable change from the xy-
u, v
coordinates to uv-coordinates. Similarly, we have the following
equations:
y z y z
dy dz du dv
u v v u
y, z
du dv
u, v
z x z x
dz dx du dv
u v v u
z, x
du dv
u, v
Definition 4.8: The surface integral of a scalar function on S is given
by:
f x, y, z dS f x, y, z T T
S D u v dudv
Where D is u-v domain.
Figure 4.4. Tangent vectors of a surface
156 Mathematics for mechanical engineers
Recall from Chapter 3 that the tangent vectors Tu , Tv of a surface
corresponding to parameters u and v at point P are (See Figure 4.4):
x y z
Tu i j k
u u u
x y z
Tv i j k
v v v
Thus,
x y z x y z
Tu Tv i j k i j k
u u u v v v
i j k
x y z
u u u
x y z
v v v
y z z y z x x z x y y x
i j k
u v u v u v u v u v u v
y , z z , x x, y
= i j k
u, v u, v u, v
So
2 2 2
y , z z , x x, y
Tu Tv
u, v u, v u, v
Therefore, the surface integral of over a surface S is computed as:
f x, y, z dS
S
2 2
y , z z , x x, y
2 (4.5)
f x, y, z
dudv
u, v u, v u, v
D
Differential forms 157
Example 4.2: Evaluate ydS where
S
S is the surface z x y 2 , 0 x 1,
0 y 2 as shown in Figure 4.5.
Solution: We regard x, y as
parameters, then:
y y
y, z x y z y Figure 4.5. Graph of S
1
x, y z z x y
x y
z z
z , x x y x z
2 y
x, y x x x y
x y
x x
x, y x y x y
1
x, y y y x y
x y
Let D be the domain of parameters x, y. So, using Formula 4.5, we
have:
y dS y 1 2 y 1
2 2 2
dxdy
S D
1 2
y 1 1 4 y 2 dy dx
0 0
1 2
dx 2 y 1 2 y 2 dy
0 0
2
2 14 23 (1 2 y 2 )3/ 2
0
13 2
3
158 Mathematics for mechanical engineers
There are many situations arising in physics which needs to integrate a
vector field F F1 x, y, z i F2 x, y, z j F3 x, y, z k over a surface
S. For instance, when we need to estimate the rate at which a fluid
with density x, y , z and velocity field v passes through a surface S
is given by the flux integral
S
v dS .
Definition 4.9: The integral of a vector field over a surface S is called
the flux integral and is presented in the form:
F dS
S S
F1dy dz F2 dz dx F3 dx dy
Therefore, it is probably easier to remember the way to compute the
surface integral as a two - step process. First, convert F to a 2-form:
F1i F2 j F3k F1dy dz F2 dz dx F3dx dy
then integrate it by using Definition 4.9.
Theorem 4.5: Let F be a vector field, S be a smooth surface, then
F dS F ndS
S S
Proof: We know that the normal vector n of a smooth surface can be
calculated from the cross product of tangent vectors Tu , Tv as:
Tu Tv
n
Tu Tv
Assume F is a vector field:
F x, y , z F1 x, y , z i F2 x, y, z j F3 x, y , z k
then,
Differential forms 159
x, y y, z z, x
S F dS D u, v u, v u, v dudv
F1 F2 F3
x, y y , z z , x
F dudv
u, v u , v u, v
D
F Tu Tv dudv
D
Tu Tv
F Tu Tv dudv
D Tu Tv
F ndS
S
For a closed oriented surface S, its orientation is defined such that the
normal vector points outward. That is, the positive orientation of S
coincides the direction of the outward normal vector.
Example 4.3: Evaluate FdS , where
S
F xi zk and S consists of the
paraboloid S1 : x y 2 z 2 ,0 x 1 and
the disk S2 : y 2 z 2 1, x 1.
Solution: As shown in Figure 4.6. we see
that S consists of top surface S1 and
bottom surface S2 .
For S1, we parameterize it by x and y, so: Figure 4.6. Graph of S
x y 2 z 2 , y y, z z
Convert the vector field F to a 2-form:
F xi zk xdy dz zdx dy
Thus, the surface integral on S1 is:
160 Mathematics for mechanical engineers
y, z x, y
S1 F d S S1 y, z y, z dydz
x z (4.4)
We have:
y y
y, z y z y z
1
y, z z z y z
y z
z z
z , x y z z x
2 y
y, z x x z z
y z
x x
x, y y z x y
2 z
y, z y y z y
y z
The direction of normal vector n of S1 is the direction of Tu Tv ,
where
y , z z , x x, y
Tu Tv = i j k
u, v u, v u, v
= 1, 2 y, 2 z
As observed from Figure 4.4, the normal vector of S1 that points
outward should have a negative x component. However, the x
component of Tu Tv is positive. Thus, the positive orientation of S1
is - Tu Tv or -n. Substituting the above Jacobians into Equation 4.4
regarding the orientation of S1 , one gets:
Differential forms 161
F dS F n dS
S1 S1
y S1
2
z 2 2 z 2 dydz
y 2
3 z 2 dydz
S1
Using the polar coordinates: y r cos , z r sin , we have:
r cos 2 3r 2 sin 2 rdrd
2 1
F dS 2
S1 0 0
cos 3sin r dr
2 1
2 2 3
0 0
2 1
0
2 cos 2 d 0 r 3dr
For S2 , x 0 , then we have:
y, z x, y
1, 0
y, z y, z
Therefore,
S2
F dS xdydz 0dydz 0
S2 S2
Finally,
F dS
S S1
F dS F dS
S2
162 Mathematics for mechanical engineers
EXERCISE
I. Determine which of the following 1-forms on 2 are exact. Express
the exact 1-forms in the form df.
1. 3 ydx 2 xdy
2. 3 ydx 3xdy
3. ln ydx x 2 dy
4. e y dx e x dy
II. Check the following statements by direct calculation:
1. The area of a rectangle enclosed by C is
C
xdy
2. The area of a circle C of radius r centered at 0 oriented
counterclockwise enclosed by C is xdy
C
III Calculate:
1. C
dx dy , where C is a circle r sin , 0
2. C
ydx xdy , where C is a circle 0 r R, 0 2
IV. Let xdx ydy zdz, xdx zdy ydz, xydz . Calculate:
1.
2.
3.
4. d
Differential forms 163
5.
D
where D is the square 0 x 1, 0 y 1, z 1 oriented
upward normal.
6.
D
where D is the square 0 x 1, 0 y 1, z 1 oriented
upward normal.
ANSWERS
I.
3 y 3x
2. df 3 ydx 3xdy is exact 1-form since
y x
II.
1. The area of a rectangle enclosed by C is C
xdy
C
xdy xdy xdy xdy xdy
S1 S2 S3 S4
On segments S1 and S3 we have
y
dy 0 , on S4 we have x 0 so the
integrals on these segments are b
equal to zero. Thus, S3
S4 S2
b
S1
C
xdy xdy ady ab
S2 0 0 a x
2. The area of a circle C of radius r centered at 0 oriented
counterclockwise enclosed by C is xdy
C
2 2
xdy r cos d r sin r 2 cos 2 d
C 0 0
4
2 r 2 1 cos 2 r 2 1 sin u
d 2 r2
0 4 4 0
164 Mathematics for mechanical engineers
IV.
1.
xdx ydy zdz xdx zdy ydz
x 2 dx dx xzdx dy xydx dz
yxdy dx yzdy dy y 2 dy dz
zxdz dx z 2 dz dy zydz dz
xz yx dx dy xy zx dx dz y 2 z 2 dy dz
5. We have
xz yx dx dy y 2 z 2 dy dz xy zx dz dx
D D
x, y y, z z, x
xz yx y2 z2 xy zx dxdy
D
x, y x, y x, y
where:
x x y y
x, y x y x y y, z x y
1, 0
x, y y y x y x, y z z
x y x y
z z
z , x x y
0
x, y x x
x y
Thus,
1 1
xz yx dxdy x yx dydx
D D 0 0
1
1 xy 2 1x 1
x dx dx
0
2 0 0 2 4
Line integrals and circulation of a field 165
CHAPTER 5. LINE INTEGRALS AND
CIRCULATION OF A FIELD
5.1 Line Integrals
In Chapter 4 we defined the line integral which use the differential
form to deal with general principles. The differential form presents a
coordinate free way for calculating line integrals. However, in many
physical circumstances, equations need to be expressed on
coordinates, that is, in these cases equations had better be expressed
by using vectors.
5.1.1 Line Integrals in 2-D Coordinates
We define the line integral as an integral over a curve C. In practice,
the "line integral" does not mean an integral over a straight line, but
any curve, so “curve integrals” would be better terminology. First we
parameterize a plane curve C by the parametric equations:
x x t , y y t , a t b (5.1)
y
P0
Pi
P1 C
Pi-1 Pn
P2
x
O
a ti-1 ti b t
Figure 5.1. Parameter interval is divided into n subintervals
The curve C can be expressed by the vector equation:
166 Mathematics for mechanical engineers
r t x t i y t j
Suppose that C is a smooth curve, which means that r t is
continuous and r t 0 . We divide the parameter interval a, b into
n subintervals ti 1 , ti of equal width as shown in Figure 5.1. Put:
xi x ti , yi y ti
We see that points Pi xi , yi corresponding to ti divide the curve C
into n subarcs with lengths s1 , s2 , …, sn . Assume that any point
Pi * xi* , yi * in the ith subarc corresponds to a point ti * in ti 1 , ti , we
have the following definition.
Definition 5.1: If f any function of two variables defined on a smooth
curve C given by Equations 5.1, the line integral of f along C is:
f x, y ds lim f xi* , yi* si (5.2)
C n
i 1
if this limit exists.
n
Notice that the sum f xi* , yi* si is similar to a Riemann sum. We
i 1
have known that if f is a continuous function, then the limit in
Definition 5.1 always exists. We have also known that the length of C
is:
2 2
L ds
b dx dy dt
C a dt dt
Then, this formula can be used to evaluate the line integral:
2 2
dx dy
f x, y ds f x t , y t dt
b
C a dt dt
(5.3)
Line integrals and circulation of a field 167
We call this the line integral with respect to arc length.
5.1.2 Interpretation of a Line Integral
The physical interpretation of a
line integral f x, y ds depends
C
on the physical interpretation of
the function f. To illustrate an
application of the line integral, we
suppose f x, y 0 is the height
of a fence whose base is the curve
C, then f x, y ds represents the
C Figure 5.2. Area of fence
area of the fence shown in Figure
5.2.
y
Example 5.1: Evaluate
C
4 ds
x
where C is the arc of the parabola
x2 1
y from 1, to (2, 2) as
2 2
shown in Figure 5.3.
Solution: Since C is a function of
x. So, we can choose x as the Figure 5.3. An arc of the
parameter. We can parameterize parabola
the curve C as:
t2
x t, y , 1 t 2
2
Applying Formula 5.3, we have:
2
y 2 1t
C x 1 t 2 1 t dt 1 2t 1 t dt 1 1 t d 1 t
2 2
2 1/ 2
4 ds 4 2 2 2
2 2
3
3/ 2 2
1 3
1 t 2 5 5 2 2
168 Mathematics for mechanical engineers
5.1.3 Piecewise-smooth Curve
We just have defined the line integral over a smooth curve C. We now
defined the line integral when C is a piecewise-smooth curve. Suppose
that C is a piecewise-smooth which consists of a finite number of
smooth curves C1 , C2 , …, Cn , where the initial point of Ci 1 is the
terminal point of Ci . Then, the integral of f over C is evaluated as the
sum of the integrals of f over each of the smooth curve Ci :
C
f x, y ds f x, y ds f x, y ds ... f x, y ds
C1 C2 Cn
y
Example 5.2: Evaluate C
4 ds , where C consists of the arc
x
x2
C1 of the parabola y followed
2
by the vertical line segment C2 from
2, 0 to 2, 2 .
Solution: The given line integral is
the sum of the integral of f over two
curves C1 and C2 as depicted in
Figure 5.4.
Figure 5.4. The given
On C1 , from Example 5.1 we have: parabola
y 2
C1
4 ds 5 5 2 2
x 3
On C2 , x is fixed at 2 so we choose y as the parameter. Thus, the
equations of C2 are:
x 2, y t , 0 t 2
Line integrals and circulation of a field 169
and
2 2
y 2 t dx dt 2
C2 4 x ds 0 4 2 dt dt dt 0 2 tdt 4
Thus,
y y y 2
C
4 ds 4 ds 4 ds 5 5 2 2 4
x C1 x C 2 x 3
5.1.4 Mass
Consider a metal wire with shape like
a curve C that has non-uniform cross
section as shown in Figure 5.5. That is,
the linear density x, y of the wire
varies over the length of the ring.
Then, the mass of the segment of the
wire from Pi 1 to Pi in this figure can
be approximated by xi* , yi* si . So,
the total mass of the wire is
approximately xi* , yi* si . By Figure 5.5. Metal wire
increasing the number of segments of
the wire, we obtain the mass m of the metal wire as the limiting value
of these approximations:
n
m lim xi* , yi* si x, y ds
n C
i 1
y
Recall function f x, y 4
in Example 5.2. If this function
x
describes the density of a wire, then the integral in Example 5.2 would
represent the mass of the wire. We now define that the center of mass
of the wire with density function ρ is the point x , y , where:
170 Mathematics for mechanical engineers
1
m C
x x x, y ds
(5.4)
1
y y x, y ds
m C
Example 5.3: Consider a wire that it has the shape of a line segment
y x, and its cross section increases from the base 0, 0 to the top
2, 2 as presented in Figure 5.6. Suppose that the linear density at
any point is proportional to its distance from the line y 0 by the
proportional factor k. Find the center of mass of the wire.
Solution: We use the parametric equations:
x t , y t , ds 2dt , 0 t 2
The linear density is described as:
x, y ky
The mass of the line segment is the
line integral:
m ky ds
C
2
2ktdt
0
2
t2 Figure 5.6. Line segment with
2k 2 2k
2 0 non-uniform cross section
The center of mass is computed as follows:
1 1
x
m C
x x, y ds xkyds
m C
2
1 2 1 t3 4
2 2k
0 ktt 2 dt
2 3 0 3
Line integrals and circulation of a field 171
and,
1
m C
y y x, y ds
1
2 2k C
y ky ds
2
1 2 1 t3 4
2t dt
2
2 2 0 2 3 0 3
4 4
Thus, the center of mass is: ,
Figure 5.7. Centre of mass
3 3
as shown in Figure 5.7. In special case, where si are replaced by
either:
xi xi xi 1
yi yi yi 1
Then, we have two integrals called the line integrals of f along C with
respect to x and y:
f x, y dx lim f xi* , yi* xi
C n
i 1
n
(5.5)
f x, y dy lim f xi , yi yi
* *
C n
i 1
In many circumstances that line integrals with respect to x and y occur
together. When this happens, we often combine them together by
writing:
C
P x, y dx Q x, y dy P x, y dx Q x, y dy
C C
(5.6)
This formula is exactly the same with the line integral using
differential form as presented in Definition 4.4 in Chapter 4. To
evaluate this line integral, we can parameterize the curve whose
geometric description is given as presented in previous. However, if
we need to evaluate the line integral over a line segment, it is useful to
172 Mathematics for mechanical engineers
parameterize the line segment by using vector representation. Recall
that the vector equation describing the line segment that starts at r0
and ends at r1 is given is:
r t 1 t r0 tr1 0 t 1 (5.7)
Example 5.4: Evaluate C
ydx 2 x dy , where C is shown in Figure
5.8, and:
a) C C1 is the line segment from
1
1, to 2, 2 .
2
b) C C2 is the arc of the parabola
x2 1
y from 1, to 2, 2 .
2 2
Solution: Figure 5.8. Graph of C
a) From the graph of C1 presented in Figure 5.8 we parameterize the
1
line segment by using Equation 5.7 with r0 1, and r1 2, 2
2
as:
1
r t 1 t 1, t 2, 2 , 0 t 1
2
Then,
x t 1 t 1 2t 3t 1
1 3 1
y t 1 t 2t t
2 2 2
Line integrals and circulation of a field 173
3
and dx 3dt , dy dt . Thus, the parametric representation for the
2
line segment is:
3 1
x 3t 1, y t , 0 t 1
2 2
Therefore, Formula 5.6 gives:
1 3 1 3
C1 ydx x dy 0 2 t 2 3 dt 2 3t 1 2 dt
3 1
9t 1 dt
2 0
1
3 9t 2 21
t
2 2 0 4
b) We see that the parabola is given as a function of x. So, we put
x t as the parameter. However, we can use directly x as the
parameter, then C2 is parameterized as:
x2
x x, y , 1 x 2
2
and, dy xdx and, by using Formula 5.6, we have:
x2
2
C2
ydx 2 x dy
1 2
dx 2 x.xdx
1 2
5 x 2 dx
2 1
2
1 5 3 15
x
2 3 1 2
5.1.5 Curve Orientation
If a curve C is described by parametric equations
174 Mathematics for mechanical engineers
x x t
y y t
at b
then its orientation is determined by the positive direction
corresponding to increasing values of the parameter t.
For instance, as shown in Figure 5.9, the initial point A corresponds to
the parameter value a. The terminal point B corresponds to t b .
A C
a b t
B
A
–C
Figure 5.9. Positive and negative directions of C
We denote –C as the curve which has the same end points as C but
with the opposite orientation (from initial point B to terminal point A),
we have:
C
f x, y dx f x, y dx
C
C
f x, y dy f x, y dy
C
However, if we integrate with respect to arc length, the value of the
line integral does not change when we reverse the orientation of the
curve:
C
f x, y ds f x, y ds
C
Line integrals and circulation of a field 175
This is because si is always positive, whereas xi and yi change
sign when we reverse the orientation of C.
5.1.6 Line Integrals in 3-D Coordinates
If C is a smooth space curve given by a vector equation:
r t x t i y t j z t k
or the by parametric equations
x x t , y y t , z z t , a t b
Assume that f is a function of three variables which is continuous on
some region containing C. Then, similar to plane curves, we define the
line integral of f along C as:
f x, y, z ds lim f xi* , yi * , zi* si
C n
i 1
The formula to evaluate it is also similar to Formula 5.3:
C
f x, y, z ds
2 2 2 (5.8)
dx dy dz
f x t , y t , z t
b
a dt dt dt
In engineering, the integrals in both Formulas 5.3 and 5.8 can be
written in the more compact vector notation:
b
a
f r t r t dt
For the special case f x, y , z 1 , we get:
b
C
ds r t dt L
a
176 Mathematics for mechanical engineers
where L is the length of the curve C.
Line integrals along C with respect to x, y, and z can also be defined
similar to that for line integrals in plane. For example:
f x, y, z dx lim f xi* , yi* , zi* xi
C n
i 1
b
f x t , y t , z t x ' t dt
a
Thus, line integrals are computed by similar formula for evaluating
line integrals in the plane:
C
P x, y, z dx Q x, y, z dy R x, y, z dz (5.9)
by expressing everything x, y , z , dx, dy , dz in terms of the parameter
t.
Example 5.5: Evaluate C
x 2 ds , where C is the circular helix presented
in Figure 5.10 given by the equations:
t
x cos t , y sin t , z , 0 t 2
2
Solution: Applying Formula 5.8, we have:
C
x 2 ds
2 2 2
dx dy dz
cos 2 t dt
2
0 dt dt dt
1 2 1
2 0
1 cos 2t sin 2 t cos 2 t dt
2
Figure 5.10. Graph of C
6 2 6
4 0
1 cos 2t dt
2
Line integrals and circulation of a field 177
Example 5.6: Evaluate C
2 y dx zdy xdz , where C consists of the
line segment C1 from 2, 2, 0 to 4,6,5 , followed by the vertical line
segment C2 from 4, 6, 5 to 4,6, 0 as shown in Figure 5.11.
Solution: Using Equation 5.7, C1 is described by vector equation:
Figure 5.11. C consists of the two segments C1 and C2
r t 1 – t 2, 2, 0 t 4,6,5
2 2t , 2 4t ,5t
Or we can parameterize C1 as:
x 2 2t , y 2 4t , z 5t
0 t 1
Thus,
1
C1
2 y dx z dy x dz 2 2 4t 2dt 5t 4 dt 2 2t 5 dt
0
1 1
18 6t dt 18t 3t 2 0 21
0
Likewise, we write C2 in the form
178 Mathematics for mechanical engineers
r t 1 – t 4, 6, 5 t 4, 6, 0 4, 6,5 – 5t
or
x 4, y 6, z 5 – 5t , 0 t 1
Then, dx 0 dy . Thus,
1
C2
2 y dx z dy x dz 4 5 dt 20
0
Finally, the given integral is:
C
2 y dx z dy x dz 21 20 1
5.1.7 Line Integrals of Vector Fields
We have known that the work done by a constant force f in moving an
object on a straight line segment d is f d , and the work done by a
variable force f x in moving an object from a to b along the x-axis
b
is W f x dx . We also learned from Chapter 1 that the work done
a
by a constant force F in moving an object along a displacement vector
D PQ in space is W F D .
We now want to find the work done by a force field F x, y , z in 3
in moving an object along an arbitrary smooth trajectory C in space.
Assume that:
F x, y , z P x, y , z i Q x, y , z j R x , y , z k
is a continuous force field on 3 such as the gravitational field in
Chapter 1. So, to compute the work done by this force field in moving
an object along a smooth curve C, we parameterize C and divide it
into subarcs Pi 1 Pi with lengths si corresponding to equal
subintervals of the parameter interval a, b .
Line integrals and circulation of a field 179
P0
Pi
P1
Pi-1 Pn
P2
O x
a ti-1 ti b t
Pi-1 Pi
O Pn
x P0 y
Figure 12. Two-dimensional and three-dimensional curves
The first graph in Figure 5.12 shows the two-dimensional case. The
second shows the three-dimensional one. Suppose a point
Pi * xi* , yi* , zi* on the ith subarc corresponding to ti* . If si from Pi 1 to
Pi is small, we can approximate si as a vector with direction the
same with the unit tangent vector T ti* at Pi* . Therefore, the work
done by the force F to move the object from Pi 1 to Pi is
approximately
180 Mathematics for mechanical engineers
F xi* , yi* , zi* si T ti* F xi* , yi* , zi* T ti* si
Then, the total work done in moving the object along C is
approximately
F( x
i 1
i
*
, yi* , zi* ) T( xi* , yi* , zi* ) si (5.10)
where T( xi* , yi* , zi* ) is the unit tangent vector at the point x, y , z on
C. We can see that Obviously, when n becomes larger these
approximations will become better. Thus, the work W done by the
force field F is computed as the limit of the Riemann sums in Formula
5.9:
W F x, y, z T x, y, z ds F Tds (5.11)
C C
We know that F.T is the component of F on the direction of T. Thus,
this formula means that the work done in moving an object along a
smooth curve C is the line integral with respect to arc length of the
tangential component of the force.
If the curve C is given by the vector equation
r t x t i y t j z t k , then:
r t
T t
r t
Remember that:
ds x t y t z t dt r t dt
2 2 2
So, using Equation 5.8, the work done expressed in Equation 5.11 is
now rewritten:
b r t b
W F r t r t dt F r t r t dt
a
r t a
Line integrals and circulation of a field 181
In physics, this integral is often written in a short form
C
F dr . We
definite the line integral of any continuous vector field as follow.
Definition 5.2: If F is a continuous vector field defined on a smooth
curve C given by a vector function r t a t b , then the line
integral of F along C is:
b
C
F dr F r t r t dt F Tds
a C
Example 5.7: Find the work done by the force field:
F x, y x 4 i xyj to move an object along the quarter-circle
r t ti t 2 j , 0 t 2 .
Solution: Since x t and y t 2 ,
we have:
F r t t 4i t 3 j
y
and
r t i 2tj
Therefore, the work done is:
2 Figure 5.13. The force field
C
F dr F r t r ' t dt
0
2
3 96
3t dt t 5
2
4
0 5 0 5
Figure 5.13 shows the force field and the curve in the example. The
work done is negative because the field impedes movement along the
curve.
Notice that C
F dr F dr because the unit tangent vector T in
C
dr Tds is replaced by its negative when C is replaced by –C.
182 Mathematics for mechanical engineers
Example 5.8: Evaluate F dr , where: F x, y , z – yi xj 4 zk , C
C
t
is the circular helix given by x cos t , y sin t , z , 0 t 2 .
2
Solution: We have:
t
r t cos t i sin t j k
2
1
r ' t sin t i cos t j k
z
2
F r t sin t i cos t j 2t k
Thus,
C
F dr
2 Figure 5.14. The circular helix
F r t r t dt
0
2
2 t2
tdt 2 2
0 2 0
Figure 5.14 shows the circular helix in this example and vectors acting
at points on C.
5.1.8 Line Integral in Differential Forms and Vector
Fields
Finally, we note the connection between line integrals of vector fields
and line integrals of scalar fields. Suppose the vector field F on 3 is
given in component form by:
F Pi Qj Rk
We use Definition 5.2 to compute its line integral along C, as follows.
Line integrals and circulation of a field 183
b
C
F dr F r t r t dt
a
b
P i Q j R k x t i y t j z t k dt
a
P x t , y t , z t x t
b
Q x t , y t , z t y t dt
a
R x t , y t , z t z t
Pdx Qdy Rdz
C
Observe that last integral is precisely the line integral presented in
Equation 5.9 and it is also exactly the same with the line integral in
differential form as presented in Definition 4.5 in Chapter 4. Hence,
we have:
C
F dr P dx Q dy R dz
C
For example, the integral y dx x dy 4 z dz in Example 5.8 could
C
be expressed as F dr where F x, y , z – yi xj 4 zk .
C
5.2. Fundamental Theorem for Line Integrals
We have learned that the Fundamental Theorem of Calculus (FTC)
can be expressed as:
b
a
F ( x) dx F (b) F (a) (5.12)
where F is continuous on a, b .
Equation 5.12 is also called the Net Change Theorem: The integral of
a rate of change is the net change. Similar to the Fundamental
Theorem of Calculus, the following theorem can be regarded as a
version of the Fundamental Theorem for line integrals.
184 Mathematics for mechanical engineers
Theorem 5.1: Suppose C is a smooth curve described by the vector
function r t , a t b . Suppose f is a differentiable function of two
or three variables whose gradient vector f is continuous on C. Then,
f dr f r b f r a
C
Here, we can think of the gradient vector f of a function f of two or
three variables as a sort of derivative of f. If F f is a vector field
then this vector field is conservative (that is, there exists a function f
so that F f ). Thus, Theorem 5.1 gives us a very simple way to
evaluate the line integral of a conservative vector field if we know the
value of f at the endpoints of C. Theorem 5.1 also implies that the line
integral of f is the net change in f.
Proof: We prove Theorem 5.1 using the definition of line integrals of
vector fields:
f dr f r t r t dt
b
C a
b f dx f dy f dz
dt
a
x dt y dt z dt
b d
f r t dt f r b f r a
a dt
z
y B(x2, y2, z2)
C
B(x2, C
y)
A(x1, y1) O y
x
O x A(x1, y1, z1)
Figure 5.15. Plane curve and space curve
Line integrals and circulation of a field 185
The last step follows from the FTC. This theorem is still true for
piecewise-smooth curves. You can prove it by subdividing C into a
finite number of smooth curves and adding the resulting integrals.
If f is a function of two variables and C is a plane curve with end
points A x1 , y1 and B x2 , y2 as presented in Figure 5.15, then
f dr f x2 , y2 f x1 , y1
C
If f is a function of three variables and C is a space curve with end
points A x1 , y1 , z1 and B x2 , y2 , z2 , then
f dr f x2 , y2 , z2 f x1 , y1 , z1
C
Example 5.9: Find the work done by the gravitational field
mMG
F (x) 3
x in moving a particle with mass m from the point
x
5,10,10 to the point 2, 2,1 along a piecewise-smooth curve C.
Solution: We have known that the gravitational field F is a
conservative vector field, F f , where:
mMG
f x, y , z
x2 y2 z2
So, the work done is easily evaluate by using Theorem 5.1:
W F dr f dr
C C
f 2, 2,1 f 5,10,10
mMG mMG
22 22 12 52 102 102
1 1 4
mMG mMG
3 15 15
186 Mathematics for mechanical engineers
5.3 Line Integrals as Circulation
In general, the line integrals have the same initial and terminal points
does not give the same results (see Example 5.4), i.e:
C1
F dr F dr
C2
So the question is raised: in what conditions the line integrals over
curves with the same end points give the same results? Let C1 and C2
be two piecewise-smooth curves, which are called paths, that have the
same end points A and B. Let f be continuous, i.e F f is
conservative vector field, then from Theorem 5.1 we have:
C1
F dr F dr
C2
This says that the line integral of a conservative vector field depends
on only end points of the path, or we can say that the line integral is
independent of path.
Let consider the line integral in a mechanical circumstance. Suppose
C is an oriented closed curve and v represents the velocity field in
fluid flow. Consider the line integral v dr v T ds and recall
C C
that v T is the component of v in the direction of the unit tangent
vector T. This means that the closer the direction of v is to the
direction of T, the larger the value of v T . Thus, v dr is a
C
measure of the tendency of the fluid to move around C and is called
the circulation of v around C.
The notation
C
vdr or C
vdr is sometimes used to indicate that the
line integral is calculated using the positive orientation of the closed
curve C. Another notation for the positively oriented boundary curve
of D is D . So the circulation can be presented by the notation
vdr . The circulation can be positive or negative, depending on the
D
orientation of C compared to the flow of F as demonstrated in Figure
5.16.
Line integrals and circulation of a field 187
C C
T
T v
a)
C
vd r 0 b)
C
vd r 0
Figure 5.16. Circulations in different velocity fields
Example 5.10: Evaluate C
F dr if
F x, y – y , x
and C is the circle oriented counterclockwise:
x2 y 2 9
Solution: As shown in Figure 5.17, the vector field appears to
circulate in the counterclockwise direction, tending to point in the
same direction of the orientation of the curve. We expect the
circulation F dr to be positive.
C
We can compute the circulation
by parameterizing C by
c(t ) 3cos t ,3sin t
0 t 2
y
Since
c(t ) 3sin t ,3cos t
the line integral is:
Figure 5.17. The vector field
188 Mathematics for mechanical engineers
F dr F c t c t dt
2
C 0
2
F 3cos t ,3sin t 3sin t ,3cos t dt
0
2
0
3sin t ,3cos t 3sin t ,3cos t dt
9sin t 9 cos 2 t dt 9dt 18
2 2
2
0 0
Example 5.11: Evaluate
C
F dr if F x, y 0, 2 x and C is the
ellipse oriented clockwise , as pictured in Figure 5.18:
y2
x2 1
9
Solution: In this case, there is
no obvious circulation of F.
However, if you look closely
at the alignment of the vector
field, you will see that it tends
to align with the direction of
the curve. The circulation of F
around C is negative.
We verify this by calculating
directly the circulation. Figure 5.18. The given velocity field
Parameterizing the ellipse by
c(t ) cos t ,3sin t , 0 t 2
the circulation is:
F dr F c t c t dt
2
C 0
2
F cos t ,sin t sin t ,3cos t dt
0
2
0
0, 2 cos t sin t ,3cos t dt
2 2
6 cos 2 tdt 3
0 0
1 cos 2t dt 6
Line integrals and circulation of a field 189
Suppose that C
F dr is independent of path in D, C is any closed
path in D. We can choose any two points A and B on C and regard C
as a composed of the path C1 from A to B followed by the path C2
from B to A as depicted in Figure 5.19, then,
C
F dr F dr F dr F d r
C1 C2 C1 C2
F dr 0
This is because C1 and –C2 have C2
the same initial and terminal
points.
A B
C1
Conversely, if
C
F dr 0 where
C is a closed path in D, then we Figure 5.19. Closed path C
demonstrate independence of path
as follows. Take any two paths C1 and C2 from A to B in D and define
C to be the curve consisting of C1 followed by –C2 . Therefore,
C
F dr 0 F dr
C1 C2
F dr F dr F dr
C1 C2
Hence,
C1
F dr F dr
C2
Theorem 5.2: C
F dr is independent of path in D if and only if:
C
F dr 0 for every closed path C in D.
Obviously, the line integral of any conservative vector field F is
independent of path. It follows that F dr 0 for any closed path.
C
From Theorem 5.2 we can deduce that the work done by a
conservative force field (such as the gravitational in Chapter 1) as it
moves an object around a closed path is zero.
We love to know if there are other vector fields that are conservative?
The following theorem says that the only vector fields that are
190 Mathematics for mechanical engineers
independent of path are conservative. Suppose that D is open, that is
for every point P in D, there is a disk with center P that lies entirely in
D. This means that D doesn’t contain any of its boundary points. In
addition, we assume that D is connected, i.e. any two points in D can
be connected by a path that lies in D.
Theorem 5.3: Let F be a vector field that is continuous on an open,
connected region D. If F dr is independent of path in D, then F is
C
a conservative vector field on D. That is, there exists a function f such
that f F .
Proof: We choose a point A a, b in D then construct the desired
potential function f by defining y
x, y
f ( x, y ) F dr (a, b)
a ,b
(x1, y)
for any point (x, y) in D. (x, y) C1
C2
D
As
C
F dr is independent of
path, it does not matter which
O x
path C from a, b to x, y is
used to evaluate f x, y . Since D Figure 5.20. Conservative field
is open, there exists a disk
contained in D with center x, y as shown in Figure 5.20. Let get any
point x1 , y in the disk with x1 x . Then, let C consist of any path C1
from a, b to x1 , y followed by the horizontal line segment C2 from
x1 , y to x, y . Then,
x1 , y
f ( x , y ) F dr F d r F d r F dr
C1 C2 a ,b C2
Obviously the first of these integrals does not depend on x. Hence,
f ( x, y ) 0 F dr Pdx Qdy
x x C2 x C2
Line integrals and circulation of a field 191
Here we assume that F Pi Qj .
On C2 , y is constant; so, dy 0 . Using t as the parameter, where
x1 t x , we have:
x
f ( x, y ) P dx Q dy P t , y dt P x, y
x x C2 x x1
A similar argument, using a vertical line segment, shows that:
y
f ( x, y ) P dx Q dy Q x, t dt Q x, y
y y C2 y y1
Thus,
f f
F Pi Q j i j f
x y
This says that F is conservative. We can see the similarity between
this Theorem 4.3 and Theorem 4.1 in Chapter 4.
We may want to know how is it possible to determine whether or not a
vector field is conservative? Assume we know that F Pi Qj is
conservative, where P and Q have continuous first-order partial
derivatives. Then, there is a function f such that F f , that is,
f f
P and Q
x y
Therefore, by Clairaut’s Theorem,
P 2 f 2 f Q
y yx xy x
Theorem 5.4: If F x, y P x, y i Q x, y j is a conservative vector
field, where P and Q have continuous first-order partial derivatives on
a domain, then, throughout D, we have:
192 Mathematics for mechanical engineers
P Q
y x
The converse of Theorem 5.4 is true only for a special type of region
as you can see in the following section.
5.4 Physical Interpretation of Line Integrals -
Conservative Vector Fields
In order to give a convenient method for verifying that a vector field
on 2 is conservative we present some concepts as follows.
simple, not closed not simple, not
closed
simple, closed not simple,
closed
simple – connected region
Region that are not simply-connected
Figure 5.21. Some types of curves and regions
Definition 5.3 (Simple curve): A simple curve is a curve that doesn’t
intersect itself anywhere between its endpoints.
Line integrals and circulation of a field 193
Definition 5.4 (Simply-connected region): A simply-connected region
in the plane is a connected region D such that every simple closed
curve in D encloses only points in D. Intuitively, it contains no hole
and can’t consist of two separate pieces.
Figure 5.21 shows some examples for different types of curves and
regions.
Theorem 5.5: Let F Pi Qj be a vector field on an open simply-
connected region D. Suppose that P and Q have continuous first-order
derivatives and P Q throughout D. Then, F is conservative.
y x
Example 5.12: Determine whether or not the vector field
F x, y 2 xi x y j is conservative.
Solution: Let P x, y 2 x and
Q x, y x y . Then,
P Q
0, 1
y x
P Q
As , F is not
y x
conservative by Theorem 5.5.
If we choose a curve C as
shown in Figure 5.22. The
vectors in the figure that start
on the closed curve C all
appear to point in roughly the
same direction as C. Thus, it Figure 5.22. Non-conservative field
looks as if F d r 0 and so
C
F is not conservative. The calculation in this example confirms this
impression.
Example 5.13: Determine whether or not the vector field
194 Mathematics for mechanical engineers
F x, y y 2 2 x i 2 xy – y j
is conservative.
Solution: Let
P x, y y 2 2 x
Q x, y – y 3 xy
then:
P Q
2y
y x
We observe that the domain
of F is the entire plane (
D 2 ), which is open and
simply-connected. Applying
Theorem 5.5 we see that F
is conservative. Let's choose
y
arbitrary curves C1 and C2
as presented in Figure 5.23.
Some vectors near the
curves C1 and C2 in the
figure point in
approximately the same
direction as the curves, Figure 5.23. Conservative field
whereas others point in the
opposite direction. So, it appears plausible that line integrals around
all closed paths are 0. This example shows that F is indeed
conservative.
However, Theorem 5.5 only help us to know if the vector field is
conservative. i.e. there exist a (potential) function f such that F f
but it does not tell us how to find the function f . Now we can see how
to find it using “partial integration” as in the next example.
Line integrals and circulation of a field 195
Example 5.14:
a) If F x, y y 3 2 x i – y 3 xy 2 j , find a function f such that
F f
b) Evaluate the line integral C
F d r , where C is the curve given by
r t 2 cos t i 3sin t j, 0 t .
2
Solution:
a) Do the same way as presented in Example 5.13, we know that F is
conservative. So, there exists a function f such that F f . Thus,
f x x, y y3 2 x (5.13)
f y x, y y 3 xy 2 (5.14)
Integrating Equation 5.13 with respect to x, we get:
f x, y xy3 x 2 g y (5.15)
The constant of integration is a constant with respect to x, i.e. it is not
a constant or a function of x, but it is a function of y which is denoted
by g y . Next, we differentiate both sides of Equation 5.15 with
respect to y:
f y x, y 3 xy 2 g y (5.16)
Equating Equations 5.14 and 5.16, we see that:
g y y
196 Mathematics for mechanical engineers
To find the function g(y), we integrate this equation with respect to y,
we have:
1 2
g y y K
2
where K is a constant. Substituting this into Equation 5.14, we have
1 2
f x, y xy 3 x 2 y K
2
This is the desired potential function.
b) Since f is conservative field, we can use Theorem 5.1 to evaluate
the given line integral. We just want to know the two end point C,
namely,
r 0 2,0
r 0,3
In the expression for f x, y in part a), the constant K is arbitrary.
So, let’s choose K 0 . Then, using Theorem 5.2 we have:
17
F dr f dr f 0,3 f 2, 0
C C 2
Obviously, when dealing with the conservative field, the Fundamental
Theorem for line integrals is very useful since it gives a much shorter
computation than the straightforward method for evaluating line
integrals that we learned in previous. The technique for finding the
potential function in 3 is much the same as for vector fields on 2 .
Law of conservation of energy: Consider a continuous force field F
that moves an object along a path C given by a vector equation:
r t , a t b , where r a A and r b B are the initial and
Line integrals and circulation of a field 197
terminal points of C. By Newton’s Second Law of Motion, we have
equation
F r t m r t .
Where F r t is the force acting on the object at a point on C, m is
the mass of object, and a t r t is the object's acceleration. So,
the work done by the force in on the object is:
W F dr
C
F r t r t dt
b
a
b
mr t r t dt
a
m bd
r t r t dt
2 a dt
m bd m 2 b
r t dt r t
2
2 a dt 2 a
m
2
r ' b r ' a
2 2
Therefore,
2 2
W 12 m v ( b ) 12 m v ( a ) (5.17)
2
The quantity 1
2 m v (t ) , which is a half the mass times the square of
the speed, is called the kinetic energy of the object. So, Equation 5.17
can be expressed as:
W K B – K A (5.18)
This says that the work done by the force field along C is equal to the
change in kinetic energy at the endpoints of C.
198 Mathematics for mechanical engineers
Now we suppose that F is a conservative force field, i.e. there exists a
potential function f such that F f . In physics, the potential function
is called potential energy function and is generally denoted as
P x, y, z f x, y, z . So, we write F P . Then, by Theorem
5.1, we have:
W F dr P dr
C C
P r b P r a P A P B
Comparing that equation with Equation 5.18, we see that:
P A K A P B K B
We can explain this formula as follows. If an object moves from one
point A to another point B under the influence of a conservative force
field, then the sum of its potential energy and its kinetic energy
remains constant. This is called the Law of Conservation of Energy. It
is the reason the vector field is called conservative.
Line integrals and circulation of a field 199
EXERCISES
I. Evaluate the line integral, where C is the given curve:
1.
C
yds , C is described by x 2t 2 , y t , 0 t 1 .
y
2.
C x
ds , C is described by x t , y t 2 , 0 t 1 .
3.
C
x 3 yds , C is the half of the circle x 2 y 2 4, y 0 .
4. xy 3ds , C is the line segment connecting 0,1 to 5, 6 .
C
5. ln xds , C is the arc of the parabola y 2 x 2 from 1, 2 to 2,8 .
C
6. xe y ds , C is the arc of the curve x e y from 1, 0 to e,1 .
C
II. Evaluate the line integral C
Fdr , where C is given by the vector
function r t :
1. F x, y x 2 y 3i yx 3/ 2 j , r t t 2 i 2t 3 j , 0 t 1
2. F x, y xyi yx 2 j+xzk , r t t 2 i 2t 2 j t 3k , 0 t 2
3. F x, y sin xi cos yj zk , r t t 2 i 2tj t 3k , 0 t 1
4. F x, y sin xi yj cos zk , r t t 2 i tj t 2k , 0 t 1
5. F x, y xyi yxj+zk , r t ti cos tj sin tk , 0 t
III. Find the work done by the force field F that move an object on C.
1. F x, y x 2 y 3i yx 3/ 2 j , C is s circle x 2 y 2 9 , oriented in the
counterclockwise direction.
200 Mathematics for mechanical engineers
2. F x, y xy 2 i yx 2 j , C consists of segments from 0,1 to 5, 6
and from 5, 6 to 8, 0 .
3. F x, y x 2 y 2 i xyj , C is the triangle with vertices 1,1 , 3, 4
and 6,1 , oriented in the counterclockwise direction.
4. F x, y y 3i x 3 j , C is the part of the astroid x cos3 t , y sin 3 t ,
in the first quadrant.
IV. Show that the line integral is independent of path and evaluate the
integral:
1. tan ydx x sec 2 ydy , C is any path from 0, 0 to 2, / 4 .
C
1 ye dx e dy , C is any path from 1, 0 to 1,1 .
x x
2.
C
ANSWERS
I.
2
y 1t 1
2.
C x
ds
0 t
4t 2 1dt
12
5 5 1
5 49 2
xy 3 ds 2 x x 1 dx
3
4.
C 0 20
6.
C
1
xe y ds e 2 y 1 e 2 y dy
0
1
3
1 e2 2 2 .
3/2
II.
352
Fdr 5t 7 4t 5 dt
2
2.
C 0 3
1 3
4.
C
Fdr 3tdt
0 2
.
Line integrals and circulation of a field 201
III.
2. The work done is:
F dr 2t 3 3t 2 t dt 2t 3 48t 2 256t dt
5 8
C 0 5
1071 1971
450
2 2
4. The work done is:
/2
C
F dr 3
0
sin10 t cos 2 t sin 2 t cos10 t dt
63
1024
IV.
2.
y
1 ye x e x , e x e x
x
1 ye dx e
x x
therefore, dy is independent of path.
C
F x, y 1 ye x i e x j f x, y x ye x is a potential
function for F.
1
1 ye dx e dy F.dr f 1,1 f 0, 0 1 .
x x
C C e
202 Mathematics for mechanical engineers
Flow of a field through a surface 203
CHAPTER 6. FLOW OF A FIELD THROUGH A SURFACE
We learn from Chapter 5 that the work done by a force field F in
moving an object can be described by the line integral of the vector
field F over the path. So what can be used to represent the amount of
fluid flowing through the surface (per unit time)? The amount of the
fluid flowing through the surface per unit time is called the flux of
fluid through the surface and it can be represented by the surface
integral. For this reason, we often call the surface integral of a vector
field a flux integral.
Imagine that water is flowing through a net in the rivers. If the net
surface is perpendicular to the water flow, a lot of water will flow
through the net surface and the flux will be large. However, if the net
is parallel to the flow, water will not flow through the net surface, and
the flux will be zero. If we know the velocity field F of water, we can
calculate the total amount of water flowing through the net surface by
adding up the component of the vector F that is perpendicular to the
surface. This establishes the flow of a field through a surface.
In general, to evaluate the surface integral we need to use convenient
coordinates which leads to the use of parametric equations to describe
the given surface. The vector functions are useful to describe more
general surfaces, called parametric surfaces. In this Chapter, we first
study parametric surfaces, how to parametrize surfaces to give the
convenient ways of computing surface integrals. Then, we will
investigate the integration of different types of surfaces. Finally, we
will take the general surface area formula and see how it applies to
special surfaces.
6.1 Parametric Surfaces
Similarly to describe a curve by a vector function r t of a single
variable t, we can describe a surface by a vector function r u , v of
two parameters u and v. Assume that:
r u, v x u, v i y u, v j z u, v k (6.1)
204 Mathematics for mechanical engineers
where r u , v is a vector function defined on a region D in the uv-
plane as shown in Figure 6.1. The set of all points x, y , z in 3 such
that:
x x u, v , y y u, v , z z u, v (6.2)
z
u
r S
r(u, v)
D (u, v)
y
v x
Figure 6.1. Vector function r is defined on a region D in the uv-plane
and u , v is defined in D, is called a parametric surface S. Equations
6.2 are called parametric equations of S. Each couple u and v maps a
point on S. By making all couples, we get all of S. In other words,
when u , v moves throughout the region D, the tip of the position
vector r u, v plots the surface S.
Example 6.1: Identify and plot the surface with vector equation:
r u , v vi 3cos uj 3sin uk
Solution: It is not easy to think of the required surface at the first
glance on this equation. However, the parametric equations for this
surface:
x v, y 3 cos u , z 3sin u
would give us an advice that y and z form a function of a circle
because:
Flow of a field through a surface 205
y 2 z 2 9 cos 2 u 9 sin 2 u
This means that, with arbitrary value of y
the projections of the given surface on the
xz-plane are all circles with radius 3.
Since x v and no restriction is placed
on v, the surface is an entire circular
cylinder with radius 3 whose axis is the x-
axis as described in Figure 6.2. But if, for
instance, u and v are restricted in a given Figure 6.2. Entire cylinder
domain as:
0 u / 2, 0 v 5
then
y 0, z 0, 0 x 5
Figure 6.3. Quarter-cylinder
So, we obtained only the quarter-cylinder
with length 5 as shown in Figure 6.3.
Grid Curve
v z
S
r
D
(𝑢 , 𝑣 )
𝐶1 y
x 𝐶
u
Figure 6.4. Parametric surface S is given by a vector function r
If a parametric surface S is given by a vector function r u , v , we can
create grid curves that lie on S by keeping u and v constant sequently.
206 Mathematics for mechanical engineers
These correspond to vertical and horizontal lines in the uv-plane.
Keeping u constant by putting u u0 , we obtain a vector function
r u0 , v of the single parameter v and create a curve C1 lying on S.
Similarly, keeping v constant by putting v v0 , we create a curve C2
given by r u , v0 that lies on S. These curves are called grid curves.
Example 6.2: Use a computer algebra system to graph the surface
r u , v 3 sin v cos u , u cos v, 3 sin v sin u
Which grid curves have u constant? Which have v constant?
Solution: We graph the portion of the surface with parameter domain
0 u 6
0 v 2
It has the appearance of a
spiral tube as shown in
Figure 6.5. To identify
the grid curves, we write
the corresponding
parametric equations:
x 3 sin v cos u
y u cos v
z 3 sin v sin u Figure 6.3. Grid curve of a spiral tube
If v is constant, then sin v and cos v are constant. So, the grid curves
with v constant are the spiral curves. If u is constant the curves must
look like circles. The reason is that, if u is kept constant, for example
u 0 , then the equations
x 3 sin v
y cos v
z0
Flow of a field through a surface 207
shows that the curve is a circle which lies in the xy-plane.
In Examples 6.1 and 6.2 we plotted parametric surfaces as the vector
equations were given. Now we want to find a vector function
representing a surface which satisfies some specific conditions.
Example 6.3: Determine a vector function representing the plane that
consists of two nonparallel vectors a and b and passes through the
point P0 with position vector r0.
Solution: This problem requires us to find the vector equation in the
form:
r u , v xi yj zk
where
x x u, v , y y u, v , z z u, v
If we can find a vector equation for an arbitrary point in the required
plane, we can establish the above equations. So let P be any point in
the plane, we can build up a vector P0 P from the given vectors a and
b by moving a certain distance in the direction of a and another
distance in the direction of b. So, there are scalars u and v such that:
P0 P ua vb
P
We can see this from Figure 6.6, vector ua
P0 P is built up from vectors a and b by
using the Parallelogram Law, for the case a
where u and v are positive. Let r be the vb b 𝑃
position vector of P, then
Figure 6.6. Parallelogram Law
r OP0 + P0 P r0 ua vb
So, the vector equation of the plane can be written as:
208 Mathematics for mechanical engineers
r u , v r0 ua vb
where u and v are real numbers.
Now, if we write the equation in term of vector's elements:
r x, y , z
r0 x0 , y0 , z0
a a1 , a2 , a3
b b1 , b2 , b3
then, the parametric equations of the plane through the point
x0 , y0 , z0 can be expressed as:
x x0 ua1 vb1
y y0 ua2 vb2
z z0 ua3 vb3
These are the required equations.
Example 6.4: Find a parametric that represents the sphere
x2 y 2 z 2 a2 .
Solution: We have learned that the sphere
can be expressed by parameters , and θ.
Here, a , so we use parameter and θ
to describe the sphere. We can use the
equations to convert from spherical to
rectangular coordinates
x a sin cos
y a sin sin
z a cos
Figure 6.7. The sphere
as the parametric equations of the sphere.
Flow of a field through a surface 209
Thus, the corresponding vector equation is:
r , a sin cos i a sin sin j a cos k
Here, we have 0 and 0 2 . So, the parameter domain is
the rectangle D 0, 0, 2 .
Example 6.5: Find a parametric representation for the cylinder
y 2 z 2 4, 0 x 2 as shown in Figure 6.8.
Solution: The cylinder has a simple
representation r2 in cylindrical
coordinates. So, we choose as parameters θ
and z in cylindrical coordinates. Then the
parametric equations of the cylinder are:
y 2 cos , z 2 sin , x x
where:
0 2 , 0 x 2 Figure 6.8. The cylinder
Example 6.6: Find a vector function that represents the elliptic
paraboloid x 2 y 2 z 2 .
Solution: If we regard x and y as parameters, then the parametric
equations are simply
x 2 y 2 z 2 , y y, z z
and the vector equation is
r x , y 2 y 2 z 2 i yj z k
In general, a surface given as an equation of the form x f y , z , we
can always consider it as a parametric surface by regarding y and z as
parameters and writing the parametric equations as:
210 Mathematics for mechanical engineers
x f y, z , y y, z z
Note that parametric representations (also called parametrizations) of
surfaces are not unique because we can use any convenient parameters
to describe the given surface. We can see this in the next examples.
Example 6.7: Find a parametric representation for the surface
z x 2 y 2 , that is, the top half of the cone z 2 x 2 y 2 .
Solution: We can see that if we use the rectangular to describe the
surface, one way to parametrize the cone is to consider x and y as
parameters:
x x, y y
Then, we can form the vector equation as:
r ( x, y ) x i y j x 2 y 2 k
Another way to describe the cone will be obtained when we recognize
that the cone can be described conveniently in polar coordinates.
Thus, another representation results from choosing r and θ as
parameters in the polar coordinates. A point (x, y, z) on the cone
shown in Figure 6.9 is expressed by the equations:
x r cos
y r sin
z x2 y2 r
Thus, a vector equation for the cone is
formed:
r r , r cos i r sin j rk
where, the parameter domain is: Figure 6.9. The cone
r 0, 0 2
Flow of a field through a surface 211
6.1.1 Tangent Planes
Let us consider a parametric surface described by a vector equation
r u , v . We can create two grid curves by keeping u constant and then
keeping v constant as shown in Figure 6.10. Let put u u0 then
r u0 , v becomes a vector function of the single parameter v and
defines a grid curve C1 lying on S. The tangent vector to C1 at P0 is
obtained by taking the partial derivative of r with respect to v:
x y z
rv (u0 , v0 ) i (u0 , v0 ) j (u0 , v0 ) k (6.3)
v v v
Similarly, keeping v constant by putting v = v0, we create a grid curve
C2 presented by r u , v0 that lies on S. Its tangent vector at P0 is:
v z
r P0
D
(𝑢 , 𝑣 )
x C1 C2 y
u
Figure 6.10. Tangent vectors
x y z
ru (u0 , v0 ) i (u0 , v0 ) j (u0 , v0 ) k (6.4)
u u u
6.1.2 Smooth Surface
If the surface has no “corners”, then it is called smooth. For a smooth
surface, the tangent plane is the plane that contains the tangent vectors
ru and rv , and the vector ru rv is a normal vector to the tangent plane.
212 Mathematics for mechanical engineers
Example 6.8: Find the tangent plane to the surface with parametric
equations x x, y y, z x 2 y 2 at the point (1, 1, 2).
Solution: Recall that if P0 x0 , y0 , z0 and P x, y , z are two points
lying on a plane, n a, b, c is an unit normal vector of the plane,
then the standard equation of a tangent plane at P0 x0 , y0 , z0 is
written in the form:
a x x0 b y y0 c z z0 0
To find a normal vector of the surface, we first form the vector
equation for the given surface
r x i y j x2 y2 k
then compute the tangent vectors:
x y z
ru i j k i 2 xk
u u u
x y z
rv i j k j 2yk
v v v
Thus, a normal vector to the tangent plane is:
i j k
ru rv 1 0 2 x 2 x i 2 y j k
0 1 2y
Notice that the point P0 1,1, 2 corresponds to
the parameter values x 1 and y 1. So, the
normal vector there is:
Figure 6.11. Tangent plane
2i 2j k
Flow of a field through a surface 213
Therefore, an equation of the tangent plane at 1,1, 2 is:
2 x 1 2 y 1 z 2 0
or
2x 2 y z 2 0
Figure 6.11 describes the given surface that is the self-intersecting
surface, and its tangent plane at 1,1, 2 .
6.1.3 Surface Area
Now, we define the surface area of a general parametric surface which
is represented by Equation 6.1. Let's start by considering a surface
whose parameter domain D is a rectangle as presented in Figure 6.12.
We divide D into subrectangles Rij . We choose ui* , v*j to be the
lower left corner of Rij . The corresponding part Sij of the surface S is
called a patch and the position vector r ui* , v*j defines the point Pij as
one of its corners.
Rij D
Sij
S
z
r
Pij
Figure 6.12. D is divided into subrectangles
Let ru* ru ui* , v*j and rv* rv ui* , v*j be the tangent vectors at Pij as
given by Equations 6.3 and 6.4. As presented in Section 3.8 of
214 Mathematics for mechanical engineers
Chapter 3, the two edges of the patch that start from Pij can be
approximated by two vectors as shown. These vectors, in turn, can be
approximated by the vectors uru* and vrv* . So, Sij is approximated
by the parallelogram determined by the vectors uru* and vrv* as
illustrated in Figure 6.13. Thus, the area of this parallelogram is:
| (u ru* ) (v rv* ) | | ru* rv* | u v
Figure 6.13. Sij is approximated by a parallelogram
So, the area of S is approximated by the following Riemann sum:
m n
| r
i 1 j 1
*
u rv* | u v
We can see intuitively that this approximation will get closer to the
area of S as we increase the number of subrectangles. Therefore,
taking the limit of this Riemann sum will lead to the double integral:
| r r | du dv
D
u v
Definition 6.1: If S is a smooth parametric surface given by
r u, v x u, v i y u, v j z u, v k , u, v D
then, the surface area of S is:
Flow of a field through a surface 215
A( S ) | ru rv | dA
D
where:
x y z
ru i j k
u u u
x y z
rv i j k
v v v Figure 6.14. A sphere
Example 6.9: Find the surface area of a sphere of radius a.
Solution: Since the surface is a sphere, we use the parameters in the
spherical coordinates as shown in Figure 6.14. The converting
equations from the spherical coordinates to the rectangular coordinates
are:
x a sin cos , y a sin sin
z a cos
where the parameter domain is:
D , 0 , 0 2
We need to compute the cross product of the tangent vectors:
i j k
x y z
r r
x y z
i j k
a cos cos a cos sin a sin
a sin sin a sin cos 0
a 2 sin 2 cos i a 2 sin 2 sin j a 2 sin cos k
216 Mathematics for mechanical engineers
Then,
| r r | a 4 sin 4 cos 2 a 4 sin 4 sin 2 a 4 sin 2 cos 2
a 4 sin 4 a 4 sin 2 cos 2
a 2 sin 2 a 2 sin
since sin 0 for 0 .
Hence, the area of the sphere is:
2 2
A | r r | dA a 2 sin d d a 2 d sin d
0 0 0 0
D
a 2 (2 )2 4 a 2
If a surface S is described by equation z f x, y where x, y lies
in D and f has continuous partial derivatives, we regard x and y as
parameters. The parametric equations are:
x x , y y , z f x, y
Thus,
f f
rx i k , ry j k
x y
and
ki j
f f f
rx ry 1 0 i jk (6.5)
x x y
f
0 1
y
Then, we have:
Flow of a field through a surface 217
2 2 2 2
f f z z
| rx ry | 1 1 (6.6)
x y x y
Finally, we derive the formula for the surface area as follows:
2 2
z z
A( S ) 1 dA (6.7)
D x y
Example 6.10: Calculate the area of the part of the paraboloid
z 4 – x 2 – y 2 that lies above the plane z 0 .
Figure 6.15. A part of paraboloid and its projection on xy-plane
Solution: Look at Figure 6.15, if we project the given paraboloid onto
the xy-plane we get a disk R with the equation x 2 y 2 4 . Therefore,
all points x, y of the given surface lies on the disk R with center the
origin and radius 2. Using Formula 6.7, we have:
2 2
z z
A 1 dA
D x y
1 2 x 2 y dA
2 2
1 4( x 2 y 2 ) dA
D
218 Mathematics for mechanical engineers
Since the xy-domain R is the disk, so we convert it to polar
coordinates, then:
2 2
A 1 4r 2 r dr d
0 0
2 2
d r 1 4r 2 dr
0 0
2 18 23 (1 4r 2 )3/ 2
2
(17 17 1)
6
6.2 Surface Integral
Having presented general knowledge of parametric surfaces, we now
define the surface integral. Let f be a function of three variables
defined in a domain that includes a surface S. First, we will define the
surface integral of f over S. We use the parametric equations to
describe S by using two parameters u and v. Then, S is a graph of a
function of two variables. Suppose a surface S has a vector equation:
r u, v x u, v i y u, v j z u, v k u, v D
Rij D
Sij
S
z
Figure 6.16. Parameter domain D and its image in xy-domain
Flow of a field through a surface 219
Let assume that the parameter domain D of u and v is a rectangle as
shown in Figure 6.16. We first divide D into subrectangles Rij with
dimensions ∆u and ∆v. Then, the surface S is divided into
corresponding patches Sij. In each patch, we take the value of f at a
*
point Pij , then multiply by the area ∆Sij of the patch, and form the
Riemann sum
m n
f ( P ) S
i 1 j 1
*
ij ij
Next, we take the limit of the Riemann sum as the number of patches
increases and finally we define the surface integral of f over the
surface S as:
m n
f ( x, y, z ) dS lim
m , n
f ( P ) S
i 1 j 1
*
ij ij (6.8)
S
We can see the similarity of the definition of the surface integral with
the definition of a line integral and the definition of a double integral.
Now we will derive the formula to evaluate the surface integral in
Equation 6.8. Similar to the case for variable change in multiple
integrals, we approximate the area ∆Sij by the area of a parallelogram
in the tangent plane. We have known that this area can be
approximated as:
∆Sij ≈ |ru x rv| ∆u ∆v
where ru and rv are the tangent vectors at a corner of Sij. If the
components of vector r are continuous and ru and rv are nonzero and
nonparallel in D, from Equation 6.8, we have:
f ( x, y, z) dS f (r(u, v)) | r
S D
u rv | dA (6.9)
This equation is also true when D is not a rectangle. Formula 6.9 gives
us the way to compute a surface integral by converting it into a double
220 Mathematics for mechanical engineers
integral over the parameter domain D. When using this formula,
remember that f r u , v is computed by parametrizing:
x x u, v , y y u, v , z z u, v
in the formula for f x, y, z . In the special case where f x, y, z 1 ,
the surface integral is equal to the surface area of S:
1 dS | r r
S D
u v | dA A( S )
y
2
Example 6.11: Compute the surface integral dS , where S is the
S
unit sphere x2 y 2 z 2 1 .
Solution: As in Example 6.4, we use the parametric equations:
x a sin cos
y a sin sin
z a cos
0 , 0 2
Then, the vector equation of the sphere is:
r , sin cos i sin sin j cos k
In Example 6.9, we have:
| r r | a 2 sin sin
Thus, applying Formula 6.9, we get:
y dS (sin sin ) | r r | dA
2 2
S D
2
(sin 2 sin 2 sin d d
0 0
Flow of a field through a surface 221
2
sin 2 d sin 3 d
0 0
2 1
(1 cos 2 ) d (sin sin cos 2 ) d
0 2 0
2
1 1 1 4
sin 2 cos cos3
2 2 0 3 0 3
If a surface S is given by equation z g x, y , we use directly
variable x, y as parameters, then we have the parametric equations:
x x , y y , z g x, y
So, we get:
g g
rx i k , ry j k
x y
Thus,
g g
rx ry i jk (6.10)
x x
and
2 2
z z
| rx ry | 1 (6.11)
x y
Therefore, the surface integral can be expressed by Formula 6.9 as:
f ( x, y, z ) dS
S
2 2 (6.12)
z z
f ( x, y, g ( x, y )) 1 dA
D x y
222 Mathematics for mechanical engineers
Similar formulas can be used if it is more convenient to project S onto
the yz-plane or xy-plane. For instance, if we are given S a surface with
equation y h x, z and D is its projection on the xz-plane, then
2 2
y y
f ( x, y, z ) dS
S D
f ( x, h( x, z ), z ) 1 dA
x z
Example 6.12: Evaluate x dS
S
where S is the surface z x 2 2 y ,
0 x 1 , 0 y 1 as shown in Figure 6.17.
Solution: We regard x and y as parameters, then:
z z
2 x, 2
x y
So, using Formula 6.12, we have:
2 2
z z
S x dS D x 1 x y dA
1 1
x 1 4 x 2 4 dx dy
0 0
1 1
dy x 5 4 x 2 dx
0 0
1
1 2
. (5 4 x 2 )3/ 2
8 3 0
Figure 6.17. Surface S
1
12
27 5 5
If S is a piecewise-smooth surface, which consists of a finite sum of
smooth surfaces S1, S2,..., Sn that intersect only along their boundaries,
then the surface integral of f over S is evaluated as:
f ( x, y, z ) dS f ( x, y, z ) dS f ( x, y, z ) dS
S S1 Sn
Flow of a field through a surface 223
Example 6.13: Evaluate z dS where S is the surface whose:
S
S1 are the sides that are given by the cylinder x 2 y 2 1.
S2 is the bottom disk x 2 y 2 1 in the plane z 0 .
S3 is the top that is a part of the plane z 1 x that lies
above S2.
Solution: S is depicted in Figure 6.18. We will compute the surface
integral over three surfaces S1, S2 and S3, then sum them up to get the
requirement.
For S1, since this surface is cylinder so we use the cylinder parameters
r, θ and z to parametrize the surface. In this case, r=1, then, as
presented in Example 6.5 we write its parametric equations as:
x cos
y sin
zz
0 2
0 z 1 x 1 cos
Thus,
i j k
r rz sin cos 0
0 0 1
cos i sin j
Figure 6.18. Graph of S
and Jacobian is:
| r rz | cos 2 sin 2 1
Thus, the surface integral over S1 is:
224 Mathematics for mechanical engineers
2 2
z z
S z dS D (1 x) 1 x y dA
3
2 1
(1 r cos ) 1 1 0 r dr d
0 0
2 1
2 (r r
2
cos ) dr d
0 0
2
2
0
12 13 cos d
2
sin
2
2 3 0
2
Therefore,
z dS z dS z dS z dS
S S1 S2 S3
3
2
0 2 3
2 2
Oriented Surfaces
To define surface integrals of vector fields, we need to distinguish the
orientable surfaces and nonorientable surfaces. Consider the Möbius
strip as shown in Figure 6.19. If we draw a line continuously along the
Möbius trip starting from point P on one side of the trip, then it would
end up at point P, but on the other side of the trip. Then, if we
continue to draw in the same direction, it would end up at the same
point P without ever lifting having crossed an edge. This says that a
Möbius strip really has only one side, or in the other words, the
Möbius strip is nonorientable surface.
The nonorientable surfaces are outside of the scope of this text book.
We consider only orientable (two-sided) surfaces. We choose a
surface S that has a tangent plane at every point x, y, z on S (except
at any boundary point) as shown in Figure 6.20 to begin with.
Obviously, in this case the surface has two unit normal vectors n1 and
Flow of a field through a surface 225
n2 n1 at x, y, z . A surface S is called an oriented surface if it is
possible to choose a unit normal vector n at every such point x, y, z
so that n varies continuously over S. The given choice of n provides S
with an orientation. Thus, there are two possible orientations for any
orientable surface.
Figure 6.19. Möbius strip Figure 6.20. Oriented surface
We now want to find the formula for the normal vector of a surface. If
a surface z g x, y is given as the graph of g, we know that the
normal vector of surface z at any point can be determined by the cross
product of two unit tangent vector n rx ry . Thus, we use Equation
6.10 to build up the formula for the unit normal vector:
g g
i jk
x y
n rx ry (6.13)
2 2
g g
1
x y
Observe that k-component is positive. In rectangular coordinates, the
k-component is positive gives the upward orientation of the surface. If
S is a smooth orientable surface described in parametric coordinates
by a vector function r u, v , then it is automatically oriented by unit
normal vector:
226 Mathematics for mechanical engineers
ru rv
n (6.14)
| ru rv |
The opposite orientation is given by –n. For instance, as presented in
Example 6.4, we know the parametric representation of the sphere
x2 y 2 z 2 a 2 by the vector equation:
r , a sin cos i a sin sin j a cos k
In Example 6.9, we found that:
r r a 2 sin 2 cos i a 2 sin 2 sin j a 2 sin cos k
and
| r r | a 2 sin
So, the orientation supplied by r , is determined by the unit
normal vector:
r r 1
n sin cos i sin sin j cos k r ( , )
| r r | a
Clearly, the unit normal vector n
points in the same direction as the
position vector, that is, outward from
the sphere as demonstrated in Figure
6.21. The opposite orientation would
have been determined by reversing
the order of the parameters in the
cross product:
n r r
since r r r r . From this Figure 6.21. Normal
vector of the sphere
results, we can give the rule for
Flow of a field through a surface 227
determining the orientation of a closed surface: For a closed surface,
which is the boundary of a solid region E, the positive orientation is
determined by normal vectors that point outward from E. The negative
orientation is defined by inward-pointing normal vectors.
6.3 Flow of a Field Through a Surface
In this section we will set up the surface integral for the flow of a field
through a surface. Suppose that S is an oriented surface supplied unit
normal vector n. Consider a fluid with density x, y , z and velocity
field v x, y, z flowing through S as shown in Figure 6.22. We regard
S as an imaginary surface that doesn’t
influences the fluid flow, for example z 𝐅
it doesn't slow down the velocity of
the fluid. Then, we define the rate of
flow (mass per unit time) per unit area n
as v . If we divide S into small
patches Sij that can be approximated
by a planar. Then, the mass of fluid
flowing Sij in the direction of the y
normal n per unit time can be x
approximated by the term Figure 6.22. Flow of
a field through a
v n A Sij surface
where ρ, v, and n are evaluated at some point on Sij. Look at this term,
we know that v n is the component of the vector ρv in the direction
of the unit vector n. Adding up these terms and taking the limit, we
derive the surface integral of the function ρv · n over S using
Definition 6.1:
v n dS ( x, y, z ) v( x, y, z ) n( x, y, z) dS
S S
(6.15)
This equation says that the rate of flow of a fluid through a surface S
is the surface integral of the function ρv · n over S. If we denote
F v , then F is also a vector field on 3 . Then, the integral in
Equation 6.15 is written in the form of abbreviation:
228 Mathematics for mechanical engineers
F n dS
S
A surface integral of this form occurs frequently in physics, even
when F is not v , for example the heat field, the electric field, and so
on. It is called in general the surface integral (or flux integral) of F
across S and is defined as follows.
Definition 6.2: If F is a continuous vector field defined on an oriented
surface S with unit normal vector n, then the surface integral of F
across S is:
F dS F n dS
S S
This integral is also called the flow (or flux) of F over S. We can see
that the term F n is the component of F in the normal direction of S.
Therefore, this equation interprets that the surface integral of a vector
field over S is the surface integral of its normal component over S.
If S is a parametric surface which is described by a vector function
r r
r u, v , then we have the normal vector n u v as given in
ru rv
Equation 6.14. Thus, from Definition 6.2 and Equation 6.9, we have:
ru rv r r
F dS F r r
S S u v
dS F(r(u, v)) u v ru rv dA
D
ru rv
where D is the parameter domain. Finally, we obtain:
F dS F (r r ) dA
S D
u v (6.16)
Example 6.13: Find the flow of the vector field
F x, y , z xi yj zk across the unit sphere x2 y 2 z 2 1 .
Solution: Since the surface is the unit sphere, we use the parametric
equation as:
Flow of a field through a surface 229
r , sin cos i sin sin j cos k
0 , 0 2
then the vector field is presented as:
F r , sin cos i sin sin j cos k
From Example 6.9 we have:
r r sin 2 cos i sin 2 sin j sin cos k
Therefore,
F r , r r
sin 3 cos 2 sin 3 sin 2 sin cos 2
sin 3 cos 2 sin cos 2
Using Formula 6.16, the flow is:
F dS
S
F (r r ) dA
D
sin cos 2 sin cos 2 d d
2
3
0 0
2 2
sin 3 d cos 2 d sin cos 2 d d
0 0 0 0
2 4
0 sin cos 2 d d
0 0 3
The vector field F in this example on the unit sphere is depicted in
Figure 6.23. We can give an example of physical meaning for this
integral. If, for instance, the sphere is a hot object and the vector field
describes the heat flow with density 1, then the answer, 4π/3,
represents the rate of heat flow through the unit sphere. in units of
mass per unit time.
230 Mathematics for mechanical engineers
0.5
-0.5
-1
1 1
0 0
-1 -1
y x
Figure 6.23. Vector field F on the unit sphere
In the case of a surface S given by an equation z = g(x, y), we can use
x and y as parameters and apply Equation 6.5 to obtain:
g g
F (rx ry ) ( P i Q j R k ) i jk
x y
Putting this equation in Formula 6.16, yields:
g g
F dS P x Q y R dA
S D
(6.17)
This formula is written in the rectangular coordinates so the positive
orientation of S is upward. The negative orientation is downward and
is determined by multiplying the upward orientation by –1. Similar
formulas can be applied if S is given by y h x, z or x k y , z .
Example 6.14: Compute F dS , if F x, y, z yi xj zk , S is the
S
boundary of the solid region E enclosed by the paraboloid z x 2 y 2
and the plane z 4 .
Flow of a field through a surface 231
Solution: The graph of S shows that it contains a parabolic bottom
surface S1 and a circular top surface S2.
Because S is a closed surface, so the positive orientation is outward.
This means that S1 is oriented downward and S2 is oriented upward.
For S1, we can see that the xy-domain D is the disk x 2 y 2 4 which
is the projection of S1 on the xy-plane as shown in Figure 6.24. Denote
that:
P x , y , z y , Q x, y , z x, R x , y , z z x 2 y 2
then,
g g
2 x, 2y
x y
Substituting these equations into Equation 6.17 we obtain:
F dS
S1
g g
P Q R dA
D
x y
[ y (2 x) x (2 y ) x 2 y 2 ] dA
D
( x 2 4 xy y 2 ) dA
D
2 1
(r 4r cos sin ) r dr d
2 2
0 0
2 1
r (1 2sin 2 ) dr d
3
0 0
2 1 Figure 6.24. Graph of S
0
1 2sin 2 d 0 r 3dr
1
(2 )
4 2
For the disk S2 at z 4 , its orientation is upward. So, its unit normal
vector is n k and we have:
232 Mathematics for mechanical engineers
F dS F (k ) dS z dA 16
S2 S2 D
Finally, by definition, we sum surface integrals of F over the pieces S1
and S2:
33
F dS F dS F dS 2 16
S S1 S2
2
Having presented some examples of surface integral of vector fields,
we know techniques to evaluate the surface integral of a vector field.
Now, we will look at some real physical situations and see how they
are described by the surface integral. For instance, if E is an electric
field, the surface integral E dS is called the electric flux of E
S
through the surface S.
Gauss’s Law: One of the important laws of electrostatics is Gauss’s
Law. This law says that the net charge enclosed by a closed surface S
is:
Q 0 E dS (6.17)
S
where ε0 is a constant called the permittivity of free space that depends
on the units used. In the SI system, ε0 ≈ 8.8542 x 10–12 C2/N · m2.
Thus, if the vector field F in Example 6.14 represents an electric field,
we can conclude that the charge enclosed by S is: Q = 4πε0/3.
Heat flow: Suppose the temperature at a point x, y, z in a body is
u x, y, z . Then, the heat flow is defined as the vector field:
F K u
where K is a constant determined experimentally and called the
conductivity of the substance. Then, the rate of heat flow across the
surface S in the body is given by the surface integral:
Flow of a field through a surface 233
F dS K u dS
S S
Example 6.15: The temperature u in a metal ball is proportional to the
square of the distance from the center of the ball. Find the rate of heat
flow across a sphere S of radius a with center at the center of the ball.
Solution: Taking the center of the ball to be at the origin, we have:
u x, y , z C x 2 y 2 z 2
where C is the proportionality constant. Then, the heat flow is:
F x, y , z K u KC 2 xi 2 yj 2 zk
where K is the conductivity of the metal. In this example, the heat
flow field is given specially so that instead of using the usual
parametrization of the sphere as usual, we can use the rectangular
coordinates. Observe that the outward unit normal to the sphere
x 2 y 2 z 2 a 2 at the point x, y, z is:
1
n xi yj zk
a
Thus,
2 KC 2
F n (x y2 z2 )
a
However, on S, we have:
x2 y 2 z 2 a2
Thus,
F n 2aKC
Thus, the rate of heat flow across S is:
234 Mathematics for mechanical engineers
F dS F n dS
S S
2aKC dS
S
2aKCA( S )
2aKC (4 a 2 )
8KC a3
Flow of a field through a surface 235
EXERCISES
I. Evaluate the surface integral.
1. S
x 2 y 2 zdS , S is a part of the plane z 1 2x 3 y that lies above
the rectangle 0,1 0,1 .
2. xyzdS , S is a triangular region with vertices 1, 0, 0 , 0, 3, 0
S
and 0, 0,3 .
3. S
xydS , S is a part of the plane x y z 2 that lies in the first
octant.
4.
S
xy 2 dS , S is the surface z x 2 y 2 , 0 x 1 , 0 y 1 .
5.
S
y 2 z 2 dS , S is the surface z x 3 y 3 , 0 x 1 , 0 y 1 .
6.
S
yz 2 dS , S is the surface x y 2 z , 0 y 1 , 0 z 1 .
II. Find the flux of vector field F across the oriented surface S
(Evaluate the surface integral F dS for the given vector field F
S
and the oriented surface S). For closed surfaces, use the positive
orientation.
1. F x, y , z xyi yzj zxk , S is a part of the paraboloid
z 5 x 2 y 2 that lies above the square 0 x 1 , 0 y 1 and has
upward orientation.
2. F x, y , z xyzi y 3 zj zx 2k , S is the surface z xe2 y , 0 x 1 ,
0 y 1 with upward orientation.
236 Mathematics for mechanical engineers
3. F x, y , z xi yj zk , S is the hemisphere x 2 y 2 z 2 9 , y 0
oriented in the direction of the positive y-axis.
4. F x, y , z xi yj zk , S consists of the paraboloid x y 2 z 2 ,
0 x 1 and the disk y 2 z 2 1, x 1 .
5. F x, y , z xi yj xzk , S the part of the sphere x 2 y 2 z 2 4 ,
in the first octant, with orientation toward the origin.
ANSWERS
I.
2. The equation of plane is z 3 3x y , thus:
1 3 3 x 9 11
xyzdS xy 3 3x y 11dydx
S 0 0 40
1 1 2
4. S
xy 2 dS
0 0
xy 2 2dxdy
6
1 1 3 6 2
6. S
yz 2 dS
0 0
yz 2 2 4 y 2 dydz
18
.
II.
F dS x ye 4 y 2 x 2 y 3e 4 y x 3e 2 y dxdy
1 1
2
2.
S 0 0
1
192
31 24e2 29e4
4. F dS
S S1
F dS F dS
S2
Since S is a closed surface, we use the outward orientation.
Flow of a field through a surface 237
On 𝑆 :
F dS y 2 z 2 2 y 2 2 z 2 dA
S1
y 2 z 2 1
3
3 y z 2 dA
2
y 2 z 2 1
2
On 𝑆 : S2
F dS 1 dA
y z 2 1
2
Hence F dS 2
S
238 Mathematics for mechanical engineers
Green, Ostrogradski and Stoke theorems 239
CHAPTER 7. GREEN, STOKE AND
OSTROGRADSKI THEOREMS
7.1 Green Theorem
Green's theorem transforms the line integral on a simple closed path C
which has no holes into a double integral over the plane region D
bounded by C. Here, region D contains all points inside C and the
curve C as shown in Figure 7.1.
D D
Figure 7.1. The plane region D bounded by C
In Green’s Theorem, we define the positive orientation of a simple
closed curve C as a single counterclockwise traversal of C. Thus, we
can observe that if we move on a curve C in the positive orientation,
then the region D is always on the left as the point r t traverses C.
Theorem 7.1 (Green’s Theorem): Let C be a positively oriented,
piecewise-smooth, simple closed curve in the plane and let D be the
region bounded by C. If P and Q are functions of x, y defined on
open region that contains D and having continuous partial derivatives,
then
Q P
C
P dx Q dy
D
dA
x y
We usually use the notation for the positively oriented boundary curve
of D by ∂D. Then, the equation in Green’s Theorem can be written as:
240 Mathematics for mechanical engineers
Q P
x y dA
D
D
P dx Q dy (7.1)
In physics, Green's Theorem finds many applications. For example, it
can be used to solve two-dimensional flow integrals, which is the sum
of fluid outflowing from a volume that is equal to the total outflow
summed about an enclosing area. In plane geometry, and in particular,
area surveying, Green's Theorem can be used to determine the area
and centroid of plane figures solely by integrating over the perimeter.
Green’s Theorem is considered as the counterpart of the Fundamental
Theorem of Calculus (FTC) for double integrals.
It should be noted that any simply connected region D (which has no
holes) can be divided into small regions such that they are all type III
regions (defined as simple regions which are both type I and type II)
as shown in Figure 7.2.
Figure 7.2. A simply region can be divided into type III regions
To prove Green’s Theorem for a general region, we first prove it for a
simple region. The following is a proof of half of the theorem for the
simplified area D, a type I region where C1 and C3 are curves
connected by vertical lines (possibly of zero length). A similar proof
exists for the other half of the theorem when D is a type II region
where C2 and C4 are curves connected by horizontal lines (again,
possibly of zero length). Putting these two parts together, the theorem
is thus proven for regions of type III. The general case can then be
deduced from this special case by decomposing D into a set of type III
regions.
Green, Ostrogradski and Stoke theorems 241
Proof: Notice that, if it can be shown that if:
P
C
P dx
D
y
dA (7.2)
and
Q
C
Q dy
D
x
dA (7.3)
are true, then Green’s Theorem follows immediately for the region D.
We prove Equation 7.2 by expressing D as a type I region as shown in
Figure 7.3:
D x, y | a x b, g1 x y g 2 x
where g1 and g2 are continuous functions. We calculate the double
integral on the right side of Equation 7.2 as follows:
P b g 2 ( x ) P
D y dA a g1 ( x) y ( x, y) dy dx
(7.4)
b
[ P ( x, g 2 ( x)) P( x, g1 ( x))] dx
a
where the last step follows from the FTC. Now we compute the line
integral in Equation 7.2. C can be rewritten as the union of four curves
C1, C2, C3, and C4.
𝑦 = 𝑔 (𝑥)
On C1 we regard x as the parameter and
get the parametric equations as: C3
x x, y g1 x , a x b D
C2
C4
So,
C1
b 𝑦 = 𝑔 (𝑥)
C1 P ( x , y ) dx a P ( x , g1 ( x )) dx
a b
Figure 7.3. Type I region
242 Mathematics for mechanical engineers
On C3, we use the parametric equations:
x x, y g 2 x , a x b
Then,
b
C3
P ( x, y ) dx
C3
P ( x, y ) dx P ( x, g 2 ( x )) dx
a
On C2 and C4, x remains constant, so dx = 0. Thus,
C2
P ( x, y ) dx 0 P ( x, y ) dx
C4
Therefore,
C
P ( x, y ) dx
P( x, y ) dx P( x, y ) dx P( x, y ) dx P ( x, y ) dx
C1 C2 C3 C4
b b
P( x, g1 ( x)) dx P( x, g 2 ( x)) dx
a a
Comparing this expression with the one in Equation 7.4, we see that:
P
C
P ( x, y ) dx
D
y
dA
A similar treatment yields Equation 7.3 for regions of type II. Then,
by adding Equations 7.2 and 7.3, we obtain Green’s Theorem. Since
Green’s Theorem gives us two choices to evaluate an integral, either a
line integral or a double integral, so we can choose the convenient
integral to work with. The following examples show that while line
integrals are complicated to evaluate, they are easier to be evaluate by
the double integrals using Green’s Theorem.
Example 7.1: Evaluate C
x 2 dx 2 xy dy , where C is the triangular
curve consisting of the line segments:
Green, Ostrogradski and Stoke theorems 243
from 0, 0 to 2, 0
from 2, 0 to 1,1
from 1,1 to 0, 0
Solution: If we evaluate the given line integral as usual, we would
involve setting up three separate integrals along the three sides of the
triangle. While, if we use Green’s Theorem then the region D
enclosed by C is simple as demonstrated in Figure 7.4. Thus we use
Green’s Theorem instead. Here C has positive orientation. Let
P x, y x 2 and Q x, y 2 xy , then
C
x 2 dx 2 xy dy
Q P
dA
D
x y
1 2 y
(2 y 0) dx dy
0 y
1
[2 xy ]xx 2y y dy
0
1
4 y y 2 dy
0 Figure 7.4. The region D
1
y 2 y3 2
4
2 3 0 3
Example 7.2: Evaluate
C
(3 y x 2 3) dx 7 x y cos y dy , where
C is the circle x 2 y 2 4 .
Solution: Look at the function under the integral, it would be
complicated if we evaluate the line integral as conventional methods.
However, observe that the region D bounded by C is the disk
x 2 y 2 4 . If we apply Green’s Theorem, the function under double
integral become simpler. So, let’s apply Green’s Theorem and then
change to polar coordinates, we have:
244 Mathematics for mechanical engineers
C
(3 y x 2 3) dx 7 x y cos y dy
7 x y cos y (3 y x 2 3) dA
D
x y
2 2
(7 3) r dr d
0 0
2 2
4 d r dr 16
0 0
In Examples 7.1 and 7.2, we found that the double integral was easier
to evaluate than the line integral. Sometimes, though, it’s easier to
evaluate the line integral, and Green’s Theorem is used in the reverse
direction. For example, if P x, y Q x, y 0 on the curve C, then
the double integral over D bounded by C is
Q P
x y dA
D
C
P dx Q dy 0
no matter what values P and Q are given in D.
7.1.1 Reverse Direction
We can apply the reverse direction of the theorem in computing areas.
As we have known, the area of D is 1 dA . There are several
D
possibilities of P and Q:
P x , y 0 , Q x, y x
P x, y y , Q x, y 0
y x
P x, y , Q x, y
2 2
so that:
Q P
1
x y
Green, Ostrogradski and Stoke theorems 245
Then, from Green’s Theorem we can see that the following formulas
can be applied to find the area of D:
A
x dy y dx 1
2 x dy y dx (7.5)
C C c
x2 y2
Example 7.3: Find the area enclosed by the ellipse 2 2 1 .
a b
Solution: The ellipse has parametric equations
x a cos t , y b sin t , 0 t 2
Using the first formula in Equation 7.5, we have:
A 1
2 x dy y dx
C
2
1
2 (a cos t )(b cos t ) dt (b sin t )(a sin t ) dt
0
ab 2
2 0
dt ab
If D is not a simple region as shown in Figure 7.5, we decompose D
into set of simple regions. Then we can apply Green's Theorem for
each simple area bounded by a closed curve. For example, if D is the
region shown here, we can write: D D1 D2 where D1 and D2 are
both simple.
C1
D1 C
C3
– C3
D2
C2
Figure 7.5. D is decomposed into simple regions
246 Mathematics for mechanical engineers
The boundary of D1 is C1 C3 . The boundary of D2 is C2 C3 .
So, applying Green’s Theorem to D1 and D2 separately, we get:
Q P
C1 C2
P dx Q dy
D1
dA
x y
Q P
C2 ( C3 )
P dx Q dy
D2
dA
x y
Summing up these two equations, the line integrals along C3 and –C3
cancel. Thus, we have:
Q P
C1 C2
P dx Q dy
D
dA
x y
This is Green’s Theorem for D D1 D2 with boundary
C C1 C2 . This argument can extend to any finite union of
nonoverlapping simple regions.
Example 7.4: Evaluate
C
e x dx 3 xy 2 dy , where C is the boundary of
the semiannular region D in the upper half-plane between the circles
x 2 y 2 1 and x 2 y 2 4 as shown in Figure 7.6.
Solution: Notice that, though D is not
simple, the y-axis divides it into two simple
regions. We apply Green’s Theorem for each
of these two regions and then add up them
together. Therefore, the given line integral is
equal to the double integral over the entire
region D. Using polar coordinates, we have:
Figure 7.6. Semiannular
region
D r , |1 r 2, 0
So, Green’s Theorem gives:
Green, Ostrogradski and Stoke theorems 247
C
e x dx 3 xy 2 dy (3xy 2 ) (e x ) dA
D
x y
3 y 2 dA
D
2
3 ( r 2 sin 2 ) r dr d
0 1
3 2
1 cos 2 d r 3dr
2 0 1
2
3 1 1 4 45
[ sin 2 ]0 r
4
2 2 1 8
7.1.2 Regions with Holes
Green’s Theorem can be extended to the regions that are not simply
connected regions, that is, regions with holes as presented in Figure
7.7. Observe that the boundary C of the region D here consists of two
simple closed curves C1 and C2. We can see that the positive direction
is counterclockwise for C1 but clockwise for C2 so that the region D is
always on the left as the curves C1 and C2 are traversed.
Figure 7.7. Regions with holes
To apply Green's Theorem, we decompose this region into simply
connected regions. Suppose that D is divided into two regions D and
D by two lines as shown. Then, applying Green’s Theorem to each of
D and D , yields:
248 Mathematics for mechanical engineers
Q P Q P Q P
x y dA x y dA x y dA
D D' D"
P dx Q dy P dx Q dy
D ' D
Observe that in
D '
P dx Q dy and D
P dx Q dy the line integrals
along the common boundary lines are in opposite directions, they
vanish. Then, we obtain:
Q P
x y dA
D
D '
P dx Q dy
D
P dx Q dy
P dx Q dy P dx Q dy
C1 C2
P dx Q dy
C
This is Green’s Theorem for the region D which has a hole.
yi xj
Example 7.5: Suppose F x, y
, find F dr where
x2 y2 C
C is a positively oriented, simple closed path that encloses the origin
as depicted in Figure 7.8.
Solution: Since C is an arbitrary closed
path that encloses the origin, we cannot
apply Green's Theorem for the given
integral directly. If we remove from the
region bounded by C a disk bounded by
a small circle C with center the origin
and radius a, that lies in C then we can
apply the theorem to the remaining
region D. Suppose C is Figure 7.8. C
counterclockwise-oriented. Then, the encloses the origin
positively oriented boundary of D is
C C . Applying Green’s Theorem for the region D, we have:
Green, Ostrogradski and Stoke theorems 249
C
P dx Q dy
C '
P dx Q dy
Q P y 2 x2 y 2 x2
dA ( x 2 y 2 ) 2 ( x 2 y 2 ) 2 dA 0
D x y D
Therefore,
C
P dx Q dy P dx Q dy
C
That is,
C
F dr F dr
C
We now easily compute this last integral using the parametrization
given by:
r t a cos ti b sin tj, 0 t 2
Thus,
C
F dr F dr
C
2
F(r (t )) r(t ) dt
0
2 ( a sin t )( a sin t ) (a cos t )(a cos t )
dt
0 a 2 cos 2 t a 2 sin 2 t
2
dt 2
0
7.2 Stokes Theorem
Stokes’ Theorem can be considered as a 3D version of Green’s
Theorem. Green’s Theorem transform a double integral on a plane
region D to a line integral along its plane boundary curve. Stokes’
Theorem transform a surface integral over a surface S to a line integral
along the space boundary curve of S.
250 Mathematics for mechanical engineers
Let consider an oriented surface S with unit normal vector n as shown
in Figure 7.9. We define the orientation of the boundary curve C from
unit normal vector n as follows: If you walk in the positive direction
around C with your head pointing in the direction of n, the integral
surface will always be on your left.
Figure 7.9. Unit normal vector of the oriented surface S
Theorem 7.2 (Stokes’ Theorem):
Let S be an oriented piecewise-smooth surface bounded by
a simple, closed, piecewise-smooth boundary curve C with
positive orientation.
If a vector field F is defined and has continuous partial
derivatives on an open region in 3 that contains S.
Then,
C
F dr curl F dS
S
The Stoke’s Theorem states that “the surface integral of the curl of a
function over a surface bounded by a closed surface is equal to the
line integral of the particular vector function around that surface.”
The theorem is named after the Irish mathematical physicist Sir
George Stokes (1819–1903). Stokes' Theorem also known as Kelvin–
Stokes Theorem because Stokes’ Theorem was actually discovered by
the Scottish physicist Sir William Thomson (1824–1907, known as
Lord Kelvin). Stokes learned of it in a letter from Thomson in 1850.
Green, Ostrogradski and Stoke theorems 251
The theorem is sometimes called the Fundamental Theorem for
Curls or simply the Curl Theorem. Noting that:
F dr F T ds
C C
curl F dS curl F n dS
S S
then, we can express the theorem as
C
F T ds curl F n dS
S
Therefore, Stokes’ Theorem tell us that the line integral of the
tangential component of F along the boundary C of a surface S is
equal to the surface integral of the normal component of the curl F .
The positively oriented boundary of the oriented surface S is often
denoted as ∂S. We can rewrite Stokes’ Theorem in from:
curl F dS
S
S
F dr (7.6)
Recall that curl F is a sort of derivative of F, we can see the analogy
among Green’s Theorem and Stokes’ Theorem as seen in Equation 7.1
and 7.5. That is, the left sides of these equations involve derivatives,
while the right sides of them involve the values of F only on the
boundary of S.
Consider the special case where the surface S is flat and lies in the xy-
plane with upward orientation. The unit normal is n 0i 0 j k . If
F Pi Qj Rk is a vector field, then Stokes’ Theorem becomes:
Q P
C
F dr curl F dS
S S x
y
k dA
It is very difficult to prove Stokes’ Theorem in its full generality.
Here, we will give a proof when S is a graph, F, S, and C are "well
behaved" as follows.
252 Mathematics for mechanical engineers
Proof: Suppose S is a surface with space boundary curve C as shown
in Figure 7.10. Assume that S is
described by the equation:
z g x, y , x, y D
where: g has continuous second-
order partial derivatives. D is a
simple plane region with boundary
C1 which is the projection of C. If
we choose the orientation of S to
be upward, the positive orientation
of C corresponds to the positive Figure 7.10. Surface S
orientation of C1.
Suppose that F Pi Qj Rk , where the partial derivatives of P, Q,
and R are continuous functions of x, y , g x, y . Then, we have:
curlF dS
S
(7.7)
R Q z P R z Q P
dA
D y z x z x y x y
If we parametrize C1 by equations:
x x t , y y t , a t b
Then, a parametric representation of C is:
x x t , y y t , z g x t , y t , a t b
Therefore, we can compute the line integral of F along C as follows:
Green, Ostrogradski and Stoke theorems 253
b dx dy dz
C F d r a P dt Q dt R dt dt
b dx dy z dx z dy
P Q R dt
a
dt dt x dt y dt
b z dx z dy
P R Q R dt
a
x dt y dt
z z
P R dx Q R dy
C1
x y
z z
Q R P R dA
D
x y y x
Notice that step 3 is derived by using Chain Rule and the last step is
obtained by using Green’s Theorem. Since P, Q, and R are functions
of x, y, and z where z is itself a function of x and y. Thus, using Chain
Rule and rearrange, we obtain:
z z
C
F dr Q R P R dA
D
x y y x
Q Q z R z R z z 2z
R
D
x z x x y z x y xy
P P z R z R z z 2 z
R dA
y z y y x z y x yx
R Q g P R g Q P
dA
D
y z x z x y x y
Observe that the last integral coincides the double integral in Equation
6.17 in Chapter 6:
g g
F dS P x Q y R dA
S D
254 Mathematics for mechanical engineers
R Q P R Q P
We also observe that , , are three
y z z x x y
components of vector curlF . Hence, the last integral becomes:
R Q g P R g Q P
y z x z x y x y dA
D
curl FdS
S
Thus,
C
F dr curl F dS
S
This establishes the theorem.
Example 7.6: Evaluate C
F dr , where:
F x, y, z y 2i xj z 2k
C is the curve of intersection of a half of the sphere
x 2 y 2 z 2 4, z 0 and the cylinder x 2 y 2 2 y as
shown in Figure 7.11. (Orient C to be counterclockwise
when viewed from above.)
Solution: To find the boundary curve C,
we solve the equations:
x2 y 2 2 y
x2 y 2 z 2 4
Subtracting, we get z 4 2 y .
Thus C is the curve given by the equations
Figure 7.11. Curve C
z 4 2 y , x2 y 2 2 y .
Green, Ostrogradski and Stoke theorems 255
C
F dr could be evaluated directly as usual. However, we can see
that it can be solve easier by using Stokes’ Theorem. We have:
i j k
curl F 1 2 y k
x y z
y2 x z2
As can be seen from the figure, there are some surfaces with boundary
C, but the simplest surface is the region S in the sphere
x 2 y 2 z 2 4 that is bounded by C. Here, S is oriented upward and
the positive orientation of C is counterclockwise.
The projection D of S on the xy-plane is the disk x 2 y 2 2 y . So, let
apply Equation 6.17 in Chapter 6 with z g x, y 4 2 y , then
using the polar coordinates, we have:
F dr curl F dS 1 2 y dA
C
S D
2 1
1 2 1 r sin r dr d
0 0
1
2 3r 2 r3
2 sin d
0
2 3 0
2 3 2
sin d 3
0
2 3
Example 7.7: Use Stokes’ Theorem to evaluate curl F dS where:
S
F x, y, z xzi yzj xyk
S is the part of the sphere x 2 y 2 z 2 4 that lies inside
the cylinder x 2 y 2 2 and above the xy-plane as depicted
in Figure 7.12.
Solution: We first find the boundary curve C by solving equation:
256 Mathematics for mechanical engineers
x2 y 2 z 2 4
x2 y 2 2
Subtracting, we get z2 = 2. Since z 0 , so C is the circle x 2 y 2 2
at z 2 . A vector equation of C in polar coordinates is:
r t 2 cos t i 2 sin t j 2 k , 0 t 2
Thus,
r t 2 sin t i 2 cos t j
On the other hand, we have:
F r t 2cos t i 2 sin t j sin 2t k
Figure 7.12. A part of
So, Stokes’ Theorem gives the sphere
curl F dS
S
C
F dr
2
F (r (t )) r '(t ) dt
0
2
2
2 cos t sin t 2 2 sin t cos t dt
0
2
0 dt 0
0
Note that, in this example, we computed a surface integral from values
of F only on the boundary curve C. So any another oriented surface
with the same boundary curve C, we get exactly the same value for the
surface integral. Therefore, in general, if S1 and S2 are oriented
surfaces with the same oriented boundary curve C and both satisfy the
hypotheses of Stokes’ Theorem, then
curl F dS
S1
C
F dr curl F dS
S2
Green, Ostrogradski and Stoke theorems 257
This fact allows us useful to choose an easy surface to integrate
instead of other difficult surfaces.
7.3 Ostrogradski Theorem
Theorem 7.3:
Let E be a simple solid region and let S be the boundary
surface of E, given with positive (outward) orientation.
Suppose F is a vector field whose component functions
have continuous partial derivatives on an open region that
contains E.
Then,
F dS div FdV
S E
This theorem says that the surface integral of a vector field F over a
closed surface S, which is called the flux through the surface, is equal
to the triple integral of the divergence over the region E bounded by S.
The theorem is named after the Russian mathematician Mikhail
Ostrogradsky (1801–1862). He published this result in 1826. This
theorem is also called Divergence Theorem and is sometimes called
Gauss’s Theorem after the great German mathematician Karl
Friedrich Gauss (1777–1855). He discovered this theorem during his
investigation of electrostatics.
Proof: Let F Pi Qj Rk , then,
P Q R
div F
x y z
Hence,
P Q R
div F dV x dV y dV z dV
E E E E
258 Mathematics for mechanical engineers
If n is the unit outward normal of S, then the surface integral on the
left side of the Divergence Theorem is:
F dS F n dS
S S
P i Q j R k n dS
S
P i n dS Q j n dS R k n dS
S S S
So, to prove the theorem, we prove the following equations:
P
P i n dS x dV
S E
(7.8)
Q
Q j n dS y dV
S E
(7.9)
R
R k n dS z dV
S E
(7.10)
Assume that E is a type I region as shown in Figure 7.13:
E x, y, z x, y D, u x, y z u
1 2 x, y
where D is the projection of E onto the xy-plane. Then we prove
Equations 7.10 first. We have:
R u2 x , y R
dV x, y, z dz dA
E
z D
u1 x , y z
So, the Fundamental Theorem of Calculus gives
R
z dV R x, y, u x, y R x, y, u x, y dA
E D
2 1 (7.11)
Green, Ostrogradski and Stoke theorems 259
The boundary surface S consists of
bottom surface S1, top surface S2,
possibly a vertical surface S3.
However, S3 may doesn’t exist, for
instance, E is a sphere.
On S3, we have k ∙ n = 0, because k
is vertical and n is horizontal. So,
R k n dS 0 dS 0
S3 S3
Figure 13. E is type I
Then, we have: region
R k n dS R k n dS R k n dS
S S1 S2
(7.12)
On S2, the representative equation is z u2 x, y , x, y D , and the
outward normal n points upward. So, using Equation 6.17 in Chapter
6 (with F replaced by Rk ), yields:
R k n dS R x, y, u x, y dA
S2 D
2
The equation of S1 is z u1 x, y . However, on this surface the
normal unit vector n points downward. So, we multiply by –1:
R k n dS R x, y, u x, y dA
S1 D
1
Thus, from Equation 7.12 we have:
R k n dS R x, y, u x, y R x, y, u x, y dA
S D
2 1
Comparison with Equation 7.11 we obtain:
260 Mathematics for mechanical engineers
R
R k n dS z dV
S E
Similar proofs can be obtained for Equations 7.8 and 7.9 when E is
regarded as a type 2 or type 3 region, respectively. We can notice the
analogue of the methods for proving the Divergence Theorem and
Green’s Theorem.
Example 7.8: Find the flux of the vector field
F x, y, z yzi 3 yj x k over the unit sphere x y z 1
2 2 2 2
Solution: To apply Ostrogradski's Theorem, we first compute the
divergence of F:
div F yz 3 y x 2 3
x y z
The unit sphere S is the boundary of the unit ball B given by:
x2 y2 z 2 1
Applying Ostrogradski's Theorem, the flux is obtained:
F dS div F dV 3 dV
S B B
4
3V B 3 1 4
3
Example 7.9: Evaluate F dS , where S is the surface of the region
S
E bounded by the parabolic cylinder z 2 y 2 and the planes
z 0, x 0, x 2 as shown in Figure 7.14, and
F x, y , z cos yz x 2 i 3 xyj e xy k .
Solution: It would be very complicated to evaluate the given surface
integral directly. We would have to evaluate four surface integrals
Green, Ostrogradski and Stoke theorems 261
corresponding to the four pieces of S. However, the div F is much less
complicated than F itself:
div F
x
cos yz x 2 3xy e xy
y z
2 x 3x x
So, we use Ostrogradski's Theorem to evaluate the triple integral
instead of the surface integral. Observe that the simplest way to
evaluate the triple integral is to consider the region E as a type III
region:
E x, y, z 0 x 2, 2 y 2, 0 z 2 y 2
Thus, from Ostrogradski's Theorem we have:
F dS div F dV
S E
x dV
E
2 2 2 y 2
x dz dy dx
0 2 0
2 2
xdx 2 y 2 dy
0 2
8 2 16 2
2 Figure 7.14. E is
3 3 type III region
The Ostrogradski's Theorem is also true for regions that are finite
unions of simple solid regions. The proof is similar to the one we have
done to extend Green’s Theorem.
For example, Figure 7.15 describes the solid region E bounded by the
closed surface S2. E consists of the solid region E1 that is bounded by
the closed surface S1 and the solid region E2 that is bounded by S1 and
S2, where S1 lies inside S2. Let n1 and n2 be outward normal vectors of
S1 and S2. Then, the boundary surface of E2 is: S S1 S2 . The
262 Mathematics for mechanical engineers
normal n of S is n n1 on S1 and n n 2 on S2. Applying
Ostrogradski's Theorem to E1 and E2, we obtain:
div F dV F dS F n
E1 S1 S1
1 dS (7.12)
div F dV F dS
E2 S
F n dS F n dS
S1 S2
(13)
F n1 dS F n 2 dS
S1 S2
F n1 dS F n 2 dS
S1 S2
Summing up Equations 7.13 and 7.14, we have:
div F dV div F dV div F dV
E1 E2 E
F n1 dS F n1 dS F n 2 dS
S1 S1 S2
F n 2 dS
S2
This is Ostrogradski's Theorem for the solid region Figure 7.15.
E E1 E2 with boundary S2.
Example 7.10: Applications of Ostrogradski's Theorem
Let’s consider the electric field. Suppose that we want to find the flux
Q
of an electric field E x 3 x over S2. Then, we choose S1 as a
x
small sphere with radius a and center the origin. Observe that:
Green, Ostrogradski and Stoke theorems 263
Qx Qy Qz
E x i j k
x y z
2 3/2
x y z
2 3/ 2
x y2 z2
2 2 2 2 2 3/2
Then,
div E
Qx Qy Qz
x x 2 y 2 z 2 3/ 2
y x 2 y 2 z 2 3/2
z x 2 y 2 z 2 3/ 2
x y2 z2 3x 2 x 2 y 2 z 2
2 3/ 2 1/2
Q
x y z
3
2 2 2
x z 3y x y z2
3/2 1/ 2
2
y2 2 2 2 2
Q
x y z 2 2 2 3
x z 3z x y
2 3/2
z2
1/ 2
2
y2 2 2 2
Q 0
x y z
3
2 2 2
Thus, from Equation 7.13, we have:
E dS E dS div E dV E dS E n dS
S2 S1 E2 S1 S1
The result says that the integral of the field E over surface S2 is equal
to the surface integral over S1. Because S1 is a sphere, the normal
x
vector at x is . Therefore,
x
Q x Q Q Q
En x 4 x x 2 2
3
x x x x a
since the equation of S1 is x x 2 y 2 z 2 a . Thus, we have:
264 Mathematics for mechanical engineers
Q Q
E dS E n dS 2 dS A S1
S2 S1
a S1
a2
Q
2
4 a 2 4 Q
a
This result says that the electric flux of E through any closed surface
S2 that contains the origin is 4πεQ.
Another application of Ostrogradski's Theorem is in fluid flow. Let's
consider the continuous velocity field v x, y, z of a fluid with
constant density ρ. Then, F v is the rate of flow per unit area. Let
P0 x0 , y0 , z0 be a point in the fluid. Ea is a very tiny sphere with
center P0 and radius a. Then, div F P div F P0 for all points in
Ea. Thus, the flux over the boundary sphere Sa can be approximated as
the surface integral F dS . Applying Ostrogradski's Theorem, we
Sa
have:
F dS div F dV div F P dV div F P V B
Sa Ba Ba
0 0 a
This approximation becomes better as a → 0, then we have:
1
div F P0 lim
a 0 V B
F dS (7.15)
a Sa
Equation 7.15 states that div F P0 is the net rate of outward flux per
unit volume at P0. If div F P 0 , P is called a source. In this case
the net flow is outward near P. If div F P 0 , P is called a sink.
The net flow is inward near P.
Figure 7.16 describes a velocity field F 2 x 2 i y 3 j . For this vector
field, we can see that the vectors that come to P1 are longer than the
vectors that go out from P1. Thus, the net flow is inward near P1. So,
div F P1 0 and P1 is a sink. Near P2, we can observe that the
Green, Ostrogradski and Stoke theorems 265
incoming vector are shorter than the outgoing arrows. Thus, the net
flow is inward. So, div F P2 0 and P2 is a source.
Figure 7.16. Velocity field
We can use the formula for F to confirm this impression:
Since F 2 x 2 i y 3 j , we have div F 4 x 3 y 2 . At P1:
x>0 and y=0, then div F 4 x 0 . So, P2 is a sink.
At P2: x=0 and y>0, then div F 3 y 2 0 . So, P2 is a
source.
266 Mathematics for mechanical engineers
EXERCISES
I. Evaluate the line integral by using Green’s Theorem.
1.
C
ydx xdy , C is the triangle with vertices (0, 0), (3, 0), (3, 3), and
(0, 3).
2. C
xy 2 dx x 2dy , C is the rectangle with vertices (0, 0), (3, 0) and
(3, 3).
3.
C
xydx x 2dy , C is the circle with the center the origin and radius
1.
4. C
ydx xdy , C consists of segments (-3, 4) to (1, 4) and parabol
y x 2 2 x 1 from (-3, 4) to (1, 4).
II. Use the Green’s Theorem to find the work done on an object:
1. By the force F x x y i x 2 yj in moving the object from the
origin along the y-axis to (0, 1) then along the line segment to (-1, 0),
and then back to the origin along the x-axis.
2. By the force F xi x3 3xy 2 j in moving the object from (-2, 0)
1/2
along the x-axis to (2, 0) then along the semicircle y 4 x 2 to
the starting point.
III. Use Stokes’ Theorem to evaluate curlF dS
S
1. F x, y, z yzi xzj xyk , S is the part of paraboloid
z 10 x 2 y 2 that lies above the plane z 4 , oriented upward.
2. F x, y, z xe yz i ye xz j ze xy k , S is the hemisphere
x 2 y 2 z 2 4, z 0 , oriented upward.
Green, Ostrogradski and Stoke theorems 267
3. F x, y, z x 2 y 2 z 2i cos xyz j xyzk , S is the part of cone
y 2 x 2 z 2 that lies between the planes y 0 and y 3 oriented in
the direction of the positive y-axis.
IV. Verify that Stokes’ Theorem is true for the given vector field F
and surface S.
1. F x, y, z y 2i xzj z 2k , S is the part of paraboloid z x 2 y 2
that lies below the plane z 2 , oriented upward.
2. F x, y, z xyzi yzj zk , S is the part of the plane x y z 1
that lies in the first octant, oriented upward.
3. F x, y, z x 2 y 2i y 2 z 2 j xyzk , S is the part of cone
x 2 y 2 z 2 1 , y 0 that lies between the planes y 0 and y 2
oriented in the direction of the positive y-axis.
V. Verify that the Frean- Ostrogradski Theorem is true for the vector
field F on the region E.
1. F x, y, z 2 xi 3xyj 3xzk , E is the cube bounded by the planes
x 0 , x 1 , y 0 , and z 1 .
2. F x, y, z xy 2 i 2 x 2 yj zk , E is the solid bounded by the
paraboloid z 1 x 2 y 2 and xy-plane.
3. F x, y, z y 2i 3xyj zxk , E is the solid cylinder x 2 y 2 4 ,
0 z 4.
4. F x, y , z xi yj zk , E is the ball x 2 y 2 z 2 4 .
ANSWERS
I.
268 Mathematics for mechanical engineers
9
3 x
2.
C
xy 2 dx x 2dy
0 2 x 2 xy dydx 4
0
64
1 4
4. ydx xdy 2 dydx .
C 2
3 x 2 x 1 3
II.
2. W xdx x3 3 xy 2 dy 3 x 2 3 y 2 0 dA , where D is the
C D
semicircular region bounded by C.
Converting to polar coordinates, we have:
2
W 3 r 2 .rdrd 12
0 0
III.
2. The boundary curve C is the circle x 2 y 2 4 in the xy – plane.
curlF.dS F.dr curlF.dS where 𝑆
S1 C S2
is the original hemisphere
and 𝑆 is the disk x y 2 4 , z 0 .
2
curlF xze xy xye xz i xye yz yze xy j yze xz xze yz k
For 𝑆 choose n k 0, 0,1
Then, curlF.n yze xz xze yz on 𝑆 , where z 0 .
Thus:
curlF.dS yze xze yz dA 0dA 0
xz
S2 x2 y 2 4 x2 y2 4
IV.
Green, Ostrogradski and Stoke theorems 269
2. The plane intersects the coordinate axes at x y z 1, so the
boundary curve C consists of the three line segments:
C1 : r1 t 1 t i tj, 0 t 1
F r1 0, 0, 0 , r1 1,1, 0
C2 : r2 t 1 t j tk , 0 t 1
F r2 0, t 1 t , t , r2 0, 1,1
C3 : r3 t ti 1 t k , 0 t 1
F r3 0, 0,1 t , r3 1, 0, 1
Thus:
1 1 1 1
F.dr 0 0 dt 0
t 1 t t
dt 0 t 1 dt 6
C
On the other hand,
curl F yi xyj xzk , z g x, y 1 x y
g g
1, 1
x y
g g
curl F.dS P x Q y R dA
S D
y xy x 1 x y dA
D
1
x 2 xy y x dydx
1 1 x
2
0 0 6
V.
2. We should show that FdS div FdV , in which the surface S
S E
consists of the paraboloid S1 : z 1 x 2 y 2 , 0 z 1 and the disk
270 Mathematics for mechanical engineers
S 2 : x 2 y 2 1, z 0 . The solid E is bounded by the paraboloid
z 1 x 2 y 2 0 z 1 and the disk S 2 : x 2 y 2 1, z 0 .
FdS FdS FdS
S S1 S2
Since S is a closed surface, we use the outward orientation.
On 𝑆 , the paraboloid z 1 x 2 y 2 n 2 x, 2 y,1
FdS xy , 2 x y, z . 2 x, 2 y,1 dA
2 2
S1 D
6 x 2 y 2 1 x 2 y 2 dA
D
On 𝑆 , the disk is in Oxy – plane x 2 y 2 1 n 0, 0, 1
FdS xy , 2 x y, z . 0, 0, 1 dA x 2 y 2 1 dA
2 2
S2 D D
On the other hand, div F 2 x 2 y 2 1
1 x 2 y 2
div FdV 2 x y 2 1 dzdA , where D is the disk in
2
E D 0
Oxy – plane x y 1 .
2 2
4. We should show that FdS div FdV , in which the surface S
S E
is the sphere x y z 4 2 2 2
and the solid E is the ball
x2 y2 z 2 4 .
Fourier series 271
CHAPTER 8. FOURIER SERIES
Fourier introduced the series for the purpose of solving the heat
equation in a metal plate. He found that that an arbitrary function can
be represented by a trigonometric series. The first publish of this great
discovery was made by Fourier in 1807. It turns out that expressing a
periodic function as a Fourier series is advantageous than expanding it
as a power series in many fields, for example astronomical
phenomena, heartbeats, tides, vibration v.v.
8.1 Fourier Series Definition
Definition 8.1: An following infinite series of sine and cosine
functions
f x a0 an cos nx bn sin nx
n 1
a0 a1 cos x a2 cos 2 x a3 cos 3 x (8.1)
b1 sin x b2 sin 2 x b3 sin 3 x
is called trigonometric series or Fourier series.
The Fourier series is named after the French mathematician Joseph
Fourier (1768–1830), who made important contributions to the study
of trigonometric series, after preliminary investigations by Leonhard
Euler, Jean le Rond d'Alembert, and Daniel Bernoulli.
8.2 Lemma
To study the convergence of Fourier series we need to present the
following lemma:
Lemma 8.1:
sin mx dx 0
cos mx dx 0,
m0
272 Mathematics for mechanical engineers
cos mx sin nx 0
0 mn
cos mx cos nx dx
mn
0 mn
sin mx sin nx dx mn
Lemma 8.2: If f x is an even function then f x cos nx is also even
function, f x sin nx is odd function, and:
2
am
0
f ( x) cos mxdx, bm 0, m N
Lemma 8.3: If f x is an odd function then f x cos nx is also odd
function, f x sin nx is even function, and:
2
f ( x)sin mx dx,
bm am 0, m N
0
8.3 Fourier Convergence Theorem
Notice that Definition 1 associates an infinite trigonometric series to a
function f x . But the questions raised from the Definition are that:
how the coefficients an , bn are calculated? Under what conditions the
series converge? And, to what function the series converge?
Let's start by supposing that the trigonometric series converges to a
continuous function f x on the interval , , that is:
f x a0 an cos nx bn sin nx , x (8.2)
n 1
We want to find formulas for the coefficients an and bn in terms of f.
To give a comparison with the power series, we recall that for a power
Fourier series 273
series f x cn x a , the coefficients is computed in terms of
n
n 1
f
n
a
derivatives cn . For Fourier series, we use integrals instead
n!
of differentials. Let's integrate both sides of Equation 8.2 and assume
that it’s permissible to integrate the series term-by-term, we get:
f ( x)dx a0 dx
a
n 1
n cos nx bn sin nx dx
2 a0 an cos nx dx bn sin nx dx
n 1 n 1
From Lemma 8.1 we have:
cos nx dx 0
and:
sin nx dx 0
So,
f ( x)dx 2 a
0
Solving this equation for a0 we have:
1
a0
2 f ( x)dx
(8.3)
To compute an for n 1we multiply both sides of Equation 8.2 by
cos mx (where m is an integer and m 1 ) and integrate term-by-
term from to :
274 Mathematics for mechanical engineers
f ( x) cos mx dx
a0 an cos nx bn sin nx dx
n 1
(8.4)
a0 cos mx dx an cos nx cos mx dx
n 1
bn sin nx cos mx dx
n 1
It’s easy to show that:
sin nx cos mx dx 0
m, n
0 nm
sin nx cos mx dx
nm
So,
f x cos mx dx am cos mx cos mx dx am
Solving this equation for am and replacing m n we have:
1
f x cos nx dx
an n 1, 2, 3,... (8.5)
To find bn , we multiply both sides of Equation 8.2 by sin mx and
integrate from to , then we obtain:
1
bn
f x sin nx dx
n 1, 2, 3,... (8.6)
Thus, if a Fourier series converge to a function f x , then the
coefficients of the Fourier series are determined by Equation 8.3, 8.5
and 8.6. These results are true for a continuous function f x . Now
we can extend these to a piecewise continuous function f x on
Fourier series 275
a, b , where f x is continuous except perhaps for finite number of
removable and jump discontinuities, but not for infinitive
discontinuities.
Definition 8.2: Suppose f is a piecewise continuous function on
, then the Fourier series of f is the series
a0 an cos nx bn sin nx
n 1
where the coefficients a0 , an and bn are defined by:
1
a0
2 f x dx
1
f x cos nx dx
an (8.7)
1
f x sin nx dx
bn
These are called the Fourier coefficients of f. It should be noted that,
in Definition 8.2 we are not saying f is equal to its Fourier series. Later
we will discuss conditions under which the Fourier series of f is
actually equal to f. So far, we are just saying that if f is a piecewise
continuous function on , there is a certain series corresponding
to f called a Fourier series of f.
Example 8.1: Find the Fourier coefficients and Fourier series of the
square-wave function f defined by:
0 if x
f x
1 if 0 x
f x 2 f x
Solution: We know that f is periodic with period 2. Its graph is
illustrated in Figure 8.1.
276 Mathematics for mechanical engineers
Figure 8.1. Square-wave function
Using Equation 8.7 the formulas for the Fourier coefficients, are
obtained as:
1 1 0 1 1
a0 f x dx 0dx 1dx
2 2 2 0 2
1
an
f x cos nx dx
1 0 1 0
0dx
cos nx dx
1 sin nx
0
n 0
1
sin n sin 0 0
1
bn
f x sin nx dx
1 0 1
0dx cos nx dx
0
1 cos nx
n 0
1
cos n cos 0
n
0 n is even
2
n n is odd
Thus, the Fourier series of f is
Fourier series 277
a0 a1 cos x a2 cos 2 x a3 cos 3 x
b1 sin x b2 sin 2 x b3 sin 3x
1
0 0 0
2
2 2 2
sin x 0 sin 2 x sin 3x 0 sin 4 x sin 5 x
3 5
1 2 2 2
sin x sin 3 x sin 5 x
2 3 5
Since n is an odd integer, we present n 2k 1 , where k is an integer.
The Fourier series can be written as:
1 2
sin 2k 1 x
2 k 1 2k 1
The graphs in Figure 8.2 present some of the partial sums of this
Fourier series. Notice that, as n increases the graph of S n x becomes
closer to the graph of square-wave function. Observe that the graph of
S n x is approaching the graph of f x , except where x 0 or x is
equal to m, where m is an integer. The graphs suggest that f x is
equal to the sum of its Fourier series except at the points where is
discontinuous.
y
y
278 Mathematics for mechanical engineers
y
y
y
Figure 8.2. Some of the partial sums of this Fourier series
If a piecewise continuous function f x has a finite discontinuities
on an interval , , then at a jump discontinuity a, the one-sided
limits of f x exist:
f a lim x a f x
f a lim x a f x
With observations above, we have the following theorem for the
Fourier series of piecewise continuous functions.
Theorem 8.1: Let f be a function with period 2 and f and its first
derivative f are piecewise continuous on , that has only finite
jump discontinuities, then:
Fourier series 279
The series in Equation 8.7 is convergent.
The sum of the Fourier series is equal to f x at all
continuous points.
At discontinuities, the sum of the Fourier series is the
average of the right and left limits (Dirichlet conditions):
1
f x f x
2
This is called Fourier Convergence Theorem. Applying this theorem
to the square-wave function in Example 8.1, we derive what we
expected from the graphs. We observe that:
f 0 lim x 0 f x 1
f 0 lim x 0 f x 0
and similarly for the other discontinuities. The average of these left
and right limits is 1/2, so for any integer n the Fourier Convergence
Theorem gives:
1 2 f x n n
sin 2k 1 x
2 k 1 2k 1 0 n n
8.4 Functions with Period 2L
If f is a function with period 2L, that is f x 2 L f x x . Let
x Lt
t and g t f x f , then g t is a function with
L
period 2. We can check this by letting x L , then t . This
means that f L g . Thus, Fourier series of function g t
is:
a0 an cos nt bn sin nt (8.8)
n 1
280 Mathematics for mechanical engineers
where:
1
a0
2 g t dt
1
an
g t cos nt dt
(8.9)
1
bn
g t sin nt dt
x
Now, we substitute t into Equation 8.9 then dt
dx .
L L
Therefore, if f is a piecewise continuous function on L, L , its
Fourier series is:
n x n x
a0 an cos bn sin
n 1 L L
Here
1 L
f x dx
2 L L
a0
1 L n x
an f x cos dx
L L L
1 L n x
bn f x sin dx
L L L
Example 8.2: Find the Fourier series of triangular wave function as
shown in Figure 8.3 which is defined by:
f x x , 1 x 1
f x 2 f x , x
For what value of x function f x converges to its Fourier series?
Solution: Substituting L 1 into the Equation 8.8 and 8.9, we get:
Fourier series 281
1 1
2 1
a0 x dx
1 0 1 1
x dx xdx
2 1 2 0
1 0 1 1 1
x2 x2
4 1 4 0 2
Figure 8.3. Triangular wave function
For n 1 , we compute
1 1
an x cos n x dx 2 x cos n x dx
1 0
Notice that x cos n x is even function so an 0 . Let's integrate by
1
parts with u x, dv cos n x dx , then du dx, v sin n x .
n
Thus,
1
x 2 1
an 2 sin n x sin n x dx
n 0 n 0
cos n x
1
2 2
0 2 2 cos n 1
n n 0 n
In addition, x sin n x is odd function, so:
1
bn x sin n x dx 0
1
282 Mathematics for mechanical engineers
Hence, the Fourier series of f is:
1 2 cos n 1
cos n x
2 n 1 n 2 2
Notice that cos x 1 when n is odd and cos x 1 when n is
even, so we have:
0 n is even
2
an 2 2 cos n 1 4
n n 2 2 n is odd
Finally, the Fourier series of f is:
1 4 4 4
2 cos x 2 cos 3 x cos 5 x
2 9 25 2
1 4
cos 2k 1 x
2 n 1 2k 12 2
Since the triangle wave function is continuous x then applying the
Fourier Convergence Theory we have:
1 4
f x cos 2k 1 x
2 n 1 2k 12 2
Example 8.3: If f is a function with period of 2 on , which is
defined as:
1 x 0
f x
1 0 x
a) Find the Fourier coefficients of f
b) Find the Fourier series of f. For what value of x the function f x
converges to its Fourier series?
Fourier series 283
c) Plot the graphs of sum of Fouries series S2 , S4 , S6 .
Solution:
a) Since the rectangular wave function is continuous x except
number 0, applying the Fourier Convergence Theory, the Fourier
coefficients of f are computed as:
a0
1
2 f x dx 2 dx
1 0
0
dx 0
1 1
f x cos nx dx cos nx dx cos nx dx 0
0
an
0
1
f x sin nx dx
bn
1 0
sin nx dx sin nx dx
0
2
sin nx dx
0 n is even
2
1 cos nx 4
n n is odd
b) The Fourier series of f is
4
f x sin 2k 1 x
k 0 2k 1
and the function f x converges to its Fourier series when
x 0; 0 x .
c) The graphs of sum of Fourier series S2 , S4 , S6 are presented in
Figure 8.4.
284 Mathematics for mechanical engineers
S4 S6
S2
Figure 8.4. The graphs of sum of Fourier series
8.5 Complex form of Fourier series
If f: is a function with period of 2 satisfying Dirichlet
conditions, then:
a0
f ( x) an cos nx bn sin nx
2 n 1
From Euler formula we have:
einx cos nx i sin nx
e inx cos nx i sin nx
Then,
einx e inx
cos nx
2
e e inx
inx
sin nx
2i
Therefore,
Fourier series 285
einx e inx einx e inx
an cos nx bn sin nx an bn
2 2i
a ibn inx an bn inx
n e e
2 2
and:
a0 an ibn inx an bn inx
f x e e
2 n 1 2 n 1 2
Now, let:
a0
c0
2
a ibn
cn n n 1
2
a b
c n n n n 1
2
then
f x ce
n
n
inx
where c n cn is the conjugate of cn . This is the complex form of
Fourier series. If n is an integer and n 0 , we have:
an ibn
cn
2
1
f ( x ) cos nx dx i f ( x) sin nx dx
2
1
2
f ( x)e inx dx
If n is a negative integer, we have:
286 Mathematics for mechanical engineers
cn c n
1
f ( x) cos nx dx i f ( x) sin nx dx
2
1
2
f ( x)e inx dx
Therefore, for x Z , we have :
1
f ( x )e
inx
cn dx
2
For functions with period 2L we use the same procedure presented in
Section 8.4 to derive the Fourier series expansion:
nx
ce
i
f x n
L
n
nx
1 L i
2 L L
cn f x e L
dx
8.6 Fourier Series and Vibration of a String
The classical equation of the vibrating string is well-suited to Fourier
series analysis. Consider a string of a guitar that is stretched tight
between two points separated by a distance L, and displaced away
from its rest position, as shown in Figure 8.5. You may imagine that
x 0 corresponds to the bridge of a guitar (on the body of the
instrument), and x L corresponds to the nut (at the top of the neck,
just below the tuning pegs). The string is about to be plucked at a
distance rL from the bridge, where 1 r 0 . When you pluck the
string, the string is pulled out from its neutral position and then
released. The string is assumed to be released with zero initial
velocity. The vertical displacement of the string obeys a type of wave
equation,
2 y x,t 2 y x,t
(8.10)
x 2 F t 2
Fourier series 287
f(x)
rL L
Figure 8.5. A string is stretched tight between two points
Where is the mass per unit length of the string and F is the tension
in the string. The wave equation for the vibrating string is solved by
separation of variables. The solution y x, t of the wave equation is
assumed in the form:
y x,t X x T t
Substituting this into Equation 8.10 gives:
X x T t X x T t (8.11)
F
Here: X is the second derivative of X with respect to space variable
and T is the second derivative of T with respect to time variable.
Letting:
1
F 2
and dividing both sizes of Equation 8.11 by XT and separate the
variable we have:
X x T t
2
X x T t
288 Mathematics for mechanical engineers
It should be noted that the left size of this equation depends only on
variable x while the right size depends only on variable t. Thus, both
sizes of this function should be a constant:
X x T t
2
X x T t
This equation can be decomposed into two ordinary differential
equations:
X x X x 0 (8.12)
T t T t 0 (8.13)
Let's solve Equation 8.12 first. This is the second order differential
equation. The characteristic equation of Equation 8.12 is:
k2 0
The solution of this equation is:
k
There are three possible forms for the eigenfunctions X x :
x
C1e C2 e x 0
X x C1 x C2 0
C1 sin x C2 cos x 0
It should be noted that the vertical displacements of the string at the
ends of the string are always equal to zero:
y 0 ,t 0
y L,t 0
Fourier series 289
These equations describe the boundary conditions of the string. Since
the string is stretched at the beginning, then the displacement of the
string is different zero, but the velocity and acceleration are zero.
Therefore, the following equations are called initial conditions:
y x,0 f x, 0
y x,0
y x,0 0
Applying the boundary conditions for the case 0 we have:
X L C1e L
C2 e L
0
We see that the X(L) function is zero only if C1 C2 0 which is
trivial.
For the case 0 we have:
X 0 0 C1.0 C2 0 C2 0
X L 0 C1.L 0 C1 0
Thus, solution C1 C2 0 is trivial.
For the case 0 we have:
X 0 0 C1.0 C2 cos 0 C2 0
X L 0 C1 sin x 0
The latter equation says that C1 is non-zero only if sin
x 0.
Thus,
L n , n 1, 2,...
Then, we have
290 Mathematics for mechanical engineers
n 2 2
n
L2
Therefore, the solution of X is:
n x
X x C1 sin (8.14)
L
Equation 8.13 can be solved by the same way. The characteristic
equation and its solution is:
k 2 2 0
k
Thus,
t
C1e C2e t 0
T t C1t C2 0
C1 sin t C2 cos
t 0
However, we just know that 0 , so we have:
T t C1 sin
t C2 cos t
Applying the initial condition, gives:
T 0 C1 0 C1 0
Thus
n t
Tn t C2 cos (8.15)
L
From Equation 8.14 and 8.15 we have:
Fourier series 291
n x n t n x n t
X n x Tn t C1C2 sin cos bn sin cos
L L L L
This describes the vibration of nth mode of the string. The general
solution of the wave equation y x, t is a linear superposition of the
modes,
n x n t
y x, t bn X n x Tn t bn sin cos (8.16)
n 1 n 1 L L
F
Now, if we substitute into Equation 8.16 then we have:
n x F n t
y x, t bn sin cos (8.17)
n 1 L L
The precise mixture of modes depends on the coefficients bn , that
are determined by the initial condition:
n x
y x, 0 bn sin
n 1 L
Obviously, this is a Fourier sine series. The initial condition describes
n x
the initial shape of the string. If we denote it as f x bn sin ,
n 1 L
the Fourier coefficients bn are computed from the initial shape of the
string as:
2 L n x
bn f x sin dx (8.18)
L 0 L
The solution of the wave equation is a sinusoidal function of two
variables space and time. The spatial profile of the string at each mode
2L
is sinusoidal, with period (wavelength) equal to as presented in
n
Figure 8.6. The amplitude of each mode varies sinusoidally in time
292 Mathematics for mechanical engineers
n F
with angular frequency n . When a guitar string is plucked,
L
F
the fundamental frequency, 1 will dominates and it is
L
associated with the pitch of the sound made by the string.
X2(x)
X1(x)
X3(x)
X4(x)
Figure 8.6. Vibration mode shapes of the string
The formula of fundamental frequency suggests that, if the length L is
shorter, the pitch increases. Tightening the string (increasing the
tension F) also increases the pitch. This gives a convenient way to
tune the string to a particular pitch. Observe that the lower strings on
guitars are wound with wire to increases the mass per unit length ,
then lowers their pitch. The Fourier coefficients bn give the relative
strengths of the modes. The basic pitch is unaffected by this
distribution as long as b1 0 . However, the quality of the sound,
Fourier series 293
called its timbre, is very much dependent on bn that makes the
distribution of harmonic overtones.
We can use the Fourier series to investigate the harmonic overtones.
Let's consider an example where the initial string profile is presented
by two linear segments:
x
0 x rL
f x rL
1 x / L rL x L
1 r
Using Equation 8.18, the Fourier coefficients are given by the sum of
two integrals:
2 rL x n x 2 L 1 x / L n x
bn
L 0 rL
sin
L
dx
L rL 1 r
sin
L
dx
(8.19)
2sin r n
2 2 , n0
n r 1 r
Substituting Equation 8.19 into Equation 8.17, we have:
2sin r n F n t n x
y x, t cos sin (8.20)
n 1 n r 1 r L
2 2
L
bn
294 Mathematics for mechanical engineers
bn
bn
bn
Figure 8.7. Spectrum of Fourier coefficients for different values of r
As can be seen from Formula 8.20, the parameter r affects the Fourier
coefficients. Figure 8.7 shows the spectrum of Fourier coefficients for
different values of r. As observed from this figure, the harmonics are
affected significantly by r. The harmonics are weakest and the sound
is thinnest when the string is plucked in the middle, r 12 . As the
string is plucked closer to the bridge, r 110 , r 210 , the harmonics
strengthen and the sound becomes richer.
Fourier series 295
EXERCISES
Find the Fourier coefficients and Fourier series of a function f if it is
defined on the interval , and it is periodic with period 2 :
2 if x 0
1. f x
2 if 0 x
1 if x 0
2. f x
x if 0 x
3. f x 2 x
4. f x x 2 1
sin x if x 0
5. f x
0 if 0 x
1 if x 0
6. f x
1 if 0 x
1 if x 1
7. f x f x 4 f x
1 if 1 x 2
0 if 2 x 1
8. f x 1 if 1 x 1 f x 4 f x
0 if 1 x 2
1 if 2 x 1
9. f x 0 if 1 x 1 f x 4 f x
1 if 1 x 2
10. f x sin x x 1
296 Mathematics for mechanical engineers
ANSWERS
1.
1 1 0 1
a0
2 f x dx 2 2dx 2
0
2dx 0
1 1 0 1
f x cos nx dx 2 cos nx dx 2 cos nx dx
an
0
2sin nx 2sin nx
0
n n
1 1 0 1
bn f x sin nx dx 2sin nx dx 2sin nx dx
0
4 cos nx 4
n
3.
1 1
a0
2 f x dx 2 2 xdx 0
1 1
f x cos nx dx 2 x cos nx dx 0
an
1 1
f x sin nx dx 2 x sin nx dx
bn
4sin nx 4n cos n
n2
5.
Fourier series 297
1 1 0 1 1
a0
2 f x dx 2 sin xdx 2
0
0dx
1 1 0 1
f x cos nx dx sin x cos nx dx
an 0dx
0
cos nx 1
n 2 1
1 1 0 1
f x sin nx dx sin x sin nx dx
bn 0dx
0
sin nx
n 2 1
7.
1 2 1 1 1 1 1 2
a0 f x dx 1dx 1dx 1dx 0
4 2 4 2 4 1 4 1
1 2
f x cos n x dx
2 2
an
1 1 1 1 1 2
1cos n x dx 1cos n x dx 1cos n x dx
2 2 2 1` 2 1
sin 2 x
n
1 2
f x sin n x dx
2 2
bn
1 1 1 1 1 2
1sin n x dx 1sin n x dx 1sin n x dx 0
2 2 2 1` 2 1
9.
298 Mathematics for mechanical engineers
1 2 1 1 1 1 1 2 1
a0 f x dx 1dx 0 dx 1dx
4 2 4 2 4 1 4 1 2
1 2
an f x cos n x dx
2 2
1 1 1 1 1 2
1cos n x dx 1cos n x dx 1cos n x dx
2 2 2 1` 2 1
sin 2 x
n
1 2
bn f x sin n x dx
2 2
1 1 1 1 1 2
1sin n x dx 0sin n x dx 1sin n x dx 0
2 2 2 1` 2 1
Fourier transform 299
CHAPTER 9. FOURIER TRANSFORM
In Chapter 8 we learned that a function can be expressed as a Fourier
series which is an infinitive sum of sine and cosine functions (or
exponential functions) with different periods (or different frequencies). If
we take the limit of the Fourier series with respect to frequency, we will
get an integral and it is called Fourier integral. The Fourier transform can
be considered as a limiting case of the Fourier series. Similar to the
Fourier series, several questions raise for the Fourier transform such as:
what kinds of functions have Fourier transforms, what is the relationship
between the give function and its Fourier integral, and whether they can
be inverted to return to the original functions, how we compute the
Fourier transform, and what are the applications of the Fourier transform.
In this chapter we will consider the relationship between the Fourier
series and Fourier transform, investigate properties of the Fourier
transform, then explore some applications of it.
9.1 From Fourier Series to Fourier Transform
From Section 8.6 of Chapter 8, a Fourier series expansion in complex
L L
form of a function on the interval , can be expressed as:
2 2
2 nx
i
f x cn e L
n
where
2 nx
1 L /2 i
cn f x e L dx
L L/ 2
We can see that the function f is the sum of components with angular
2 n n
frequency n , or with the ordinary frequency n n . Here,
L 2 L
1
the spacing of the Fourier components is n . Substituting these
L
into the above expressions, we have:
300 Mathematics for mechanical engineers
f e i 2 n d ei 2 n x
L/2
f x
n
L/2
Now, let's see what will be happened if the period of f x increases to
the limiting case, L . In this case, the function f x is aperiodic
(not periodic) and we can consider that f x has only one period. The
line spacing becomes an infinitesimal d , the frequency n
becomes a continuous frequency variable , and the summation on the
right side of this equation becomes a Riemann integral:
f x f e i 2 d ei 2 x d
The Fourier coefficients cn now becomes a function F . We call
the inner integral the forward Fourier transform, and call the outer
integral the inverse Fourier transform.
f x F ei 2 x d (9.1)
F f x e i 2 x dx (9.2)
This pair of integrals always occurs together to describe the Fourier
transform. Equation 9.1 is called the forward Fourier transform and
Equation 9.2 is called the inverse Fourier transform.
In many engineering problems f is a function of time such as vibration
signals, wind velocity fields, fluid velocity fields. Replacing 2 and
t x in the above Equations 9.1 and 9.2, the Fourier transform is
presented as:
1
f t F eit d
2
F f t e it dt
Fourier transform 301
Where, is called angular frequency whose units are radians per second.
9.2 Theorems and Properties of Fourier Transform
The Fourier transform is an integral, and we know that integrals are
linear operations:
af a f
f g f g
Thus, we have the following theorem.
Theorem 9.1 (Linearity): Let f and F, g and G are two pairs of Fourier
integrals: f F and g G , and let a, b be constants. Then,
af bg aF bG
Theorem 9.2 (Symmetry): Let f F . Then,
If f is real, then F is Hermitian: F F * .
If f is even (odd), then F is even (odd).
If f is real and even (real and odd), then F is real and even
(imaginary and odd).
Here: F * is the complex conjugate of F. The converses of this theorem
hold, as well. For example, if F is Hermitian then f is real. The Fourier
transform obeys the same symmetries as the other transforms. The
following example will illustrate the Fourier transform symmetries.
Example 9.1: The one-sided exponential function e xU x which can be
decomposed into the sum of an even, continuous function, and an odd,
discontinuous function:
1 x 1 x
e xU x e e sgn x
2 2
302 Mathematics for mechanical engineers
where U is Heaviside function, sign is the signum function:
1 x0
1
U x x0
2
0 x0
1 x0
sgn x 0 x0
1 x0
1 x 1 x
Find the Fourier transforms of components e and e sgn x .
2 2
Solution: Using the definition, Fourier transform of function e xU x
can be calculated as follows:
F e xU x e x e i 2 x dx
0
1
e
1 i 2 x
1 i 2 0
1
1 i 2
1 i 2
1 2 1 2
2 2
Separating the even and odd parts of f and the real and imaginary parts of
F, gives:
1 x 1 x 1 i 2
e e sgn x
1 2 1 2
2 2
2 2
By the Fourier transform symmetries, we associate the even part of f with
the real, even part of F, and the odd part of f with the imaginary, odd part
of F. This yields two additional transform pairs:
Fourier transform 303
2
e x
1 2
2
i 4
sgn x
x
e
1 2
2
Theorem 9.3 (Reversal): Let f x F , then
f x F
f * x F *
f * x F *
Proof: Here we will give the proof for the first relationship. By making
the
change of variable x , we have:
f x limt f x e i 2 x dx
t
t
limt f ei 2 d
t
t
limt f ei 2 d
t
t
f e i 2 d
F
Theorem 9.4 (Repeated Transform): Let f x F , then
F x f v
or
1 f x f x
Proof: The forward Fourier transform of a function F(x) is:
304 Mathematics for mechanical engineers
F x e i 2 x dx F x e
i 2 x
dx
F e
i 2 x
d
f x
Notice that this formula has the form of an inverse transform, with x and
exchanged, and also with a minus sign on , that is, f .
Theorem 9.5 (Shift): Suppose f x F and f x a is a
translation of f along the x-axis, and F b is a translation along the
-axis, then
f x a e i 2 a F
ei 2 bx f x F b
Proof: We give the proof for the first equation. The forward transform
integral is:
f x a
f x a ei 2 x dx
We change the variable x a then we have:
f x a e i 2 x dx f ei 2 a d
e 2 a f ei 2 d
e 2 a F
The second equation can be also proved by using the forward transform:
ei 2 bx f x e i 2 x dx f x e i 2 b x dx F b
The Shift Theorem is very important in engineering problems. For
example, for audio engineering, if we need to amplify an input audio we
use the audio amplifier. The amplifier will transform the input audio
Fourier transform 305
f t into the frequency domain F , multiply with some gain factor
H , and finally converse into the time domain as an output audio. The
frequency response of an ideal audio amplifier is described by a
complex-valued function:
H H ei
The magnitude H is called the magnitude response and the
argument is called the phase response. These quantities are very
important for the amplifier to modify the amplitude and phase of the
Fourier components of the input signal before producing the output.
We can describe the procedure for an amplifier as follows. First it apply
the forward Fourier transform for f t , then multiply with H , and
finally converse to the output audio:
f t F H ei v F f t
here f t is the output which has been amplified by the amplifier. Now
consider the special case of an amplifier with linear phase, by which we
mean the phase response is linearly proportional to frequency:
2 a
Then,
H H 0 e i 2 a
Applying the amplifying procedure for a pure tone input, cos 2 0t , the
output is:
f t cos 2 0t F H 0 e 2 a F
H 0 f t a H 0 cos 2 0 t a
306 Mathematics for mechanical engineers
Thus, a cosine input cos 2 0t will be passed to the output as:
H 0 cos 2 0 t a
The cosine signal is amplified by the frequency-dependent magnitude
response, but is delayed by a time a, which is independent of frequency.
This means that all frequencies in the input signal will be delayed by the
same amount of time in passing through the amplifier and will be
preserved in the output. Consequently, there is no phase distortion due to
the time delay of frequencies relative to one another. Therefore, linear
phase is highly prized for high-fidelity sound reproduction.
The frequency-domain version of the Shift Theorem (the second equation
of Theorem 9.5) is often expressed by the following theorem.
Theorem 9.6 (Modulation Theorem): If f x has the Fourier transform
F , then f x cos x has the Fourier transform
1 1
f x cos 2 0 x F 0 F 0
2 2
Proof: The forward Fourier transform gives:
ei 2 0 x e i 2 0 x i 2 x
f x cos 2 0 xe i 2 x
dx f x e dx
2
1 1
f x e i 2 0 x i 2 x
e dx f x e i 2 0 x e i 2 x dx
2 2
1 1
f x e 0 dx f x e 0 dx
i 2 x i 2 x
2 2
1 1
F 0 F 0
2 2
Theorem 9.7 (Dilation): Let f x F and a is a nonzero constant,
then
Fourier transform 307
1
f ax F
a a
Proof: Using the forward Fourier transform, with making the change of
variable ax , we have:
d
f ax e i 2 x dx f e i 2 / a
a
1 1
f e i 2 / a d F
a a a
The Dilation Theorem says that if we squeeze a function in the time
domain, you stretch its transform in the frequency domain, and vice
versa.
Theorem 9.8 (Derivative): Suppose f , f are integrable (e.g., L1 or L2)
and f F , then
f x i 2 F
Proof: Writing down the integral, and using directly the definition of the
derivative, we obtain:
f f x e i 2 x dx
f x x f x i 2 x
lim x 0 e dx
x
f x x i 2 x f x
lim x 0 e dx e i 2 x dx
x x
Shift ei 2x F F
lim x0 x x
ei 2x 1 L ' hopital
F lim x 0 F lim x 0 i 2 ei 2x
x
i 2 F
308 Mathematics for mechanical engineers
Theorem 9.9 (Integral): Suppose f is integrable with f x dx 0 and
f F , then:
x F
f x dx
i 2
F
If is integrable, then the inverse relationship also holds.
i 2
Proof: We know that:
d x
f x dx f x
dx
By Derivative Theorem of Fourier transform,
d x
dx
f x dx f x dx i 2
x
f x dx F
d x
f x dx and f x , this relationship becomes:
dx
Removing the terms
i 2
x
f x dx F
Therefore,
x F
f x dx
i 2
Theorem 9.10 (Area): If f, F are continuous at the origin and f F ,
then:
f x dx F 0
f 0 F dx
Fourier transform 309
This theorem says that the value of f at 0 is equal to the area bounded by
F and the frequency-axis, and the value of F at 0 is equal to the area
bounded by f and the time-axis.
Proof: The Definition of the Fourier transform gives:
f x e i 2 x dx F
If we set 0 then,
f x dx F 0
Similarly, if we set x 0 in the inverse Fourier transform then,
f 0 F ei 2 0 dx
F dx
Definition 9.1 (Convolution): The convolution of two functions f and h,
denoted f ∗ h, is defined by the integral:
f h f h x d
if the integrals exist.
The convolution is commutative f h h f . Indeed, from the
definition of the infinitive integral, we have:
t
f h x d lim t f h x d
t
Set x and bear in mind that x is a finite number, then:
t x t
limt f h x d limt f x h d
t x t
x t
limt f x h d
x t
310 Mathematics for mechanical engineers
f x h d
f x h d
In the special case that f 0 for all x outside a bounded interval a, b
(the function is then said to have bounded support), the convolution is a
finite integral:
b
f h f h x d
a
Definition 9.2 (Correlation): The correlation which is close relative of
convolution is defined as:
f g f * t g t dt f * t g t dt
The correlation can be considered as a special case of the convolution.
The relationship between the correlation the convolution the that is
expressed as:
f g f * t g t dt
f * t g t dt f * g
Theorem 9.11 (Convolution Theorem): If f , h L1 then f h L1 , and,
f h FH
Proof:
f h f h x d e i 2 x dx
f h x e i 2 x d dx
f h x e i 2 x dx d
Fubini's Theorem
Fourier transform 311
f h x e i 2 x dx d
f e i 2 x H d Shift Theorem
f e i 2 x d H
F H
Theorem 9.12 (Symmetric Convolution): If both F and H are integrable,
then we also have F ∗ H ∈ L1 and,
fh F H
Theorem 9.13 (Correlation): Let f F and g G , then the cross-
correlation and the autocorrelation have their Fourier transforms:
Correlation: f g F *G
Autocorelation: f f F *G F
2
for functions f and g such that the correlations exist.
Proof: Applying the Convolution Theorem we have:
f f f * g
Then using the Reversal Theorem of Fourier transform f * F * ,
then:
f g f * g F *G
The Autocorrelation Theorem follows by setting g f and G F .
f f f * f F *F F
2
312 Mathematics for mechanical engineers
Orthogonality property
From the definition of the Fourier transform, we have:
F f t e i 2 t dt
F ei 2 t d e i 2 t dt
F ei 2 2 t d dt
F ei 2 2 t dt d
F ei 2 2 t dt d
Now we introduce the Diract delta function which is defined by:
0 t0
t t dt 1
0 t 0
and it has a property:
h t t dt h 0
Then, comparing the last expression of F with the property of Diract
delta function:
F d F
we have:
ei 2 dt
2 t
This is the orthogonality result which underlies our Fourier transform.
9.3 The Sampling Theorem
Fourier transform 313
Before presenting the Sampling Theorem we introduce the definitions of
a periodic generalized function and the comb function.
Definition 9.3 (Generalized Function): Suppose f is a function with
period L, a generalized function is periodic if:
f x x dx f x x L dx
Definition 9.4 (Comb Function): The comb function is defined as a
infinite unit impulse train:
III x x n
n
Some periodic generalized functions can be obtained by differentiating
periodic functions. For example, if we differentiate a square wave, or
take the second derivative of a triangle wave we will derive the periodic
generalized functions.
Definition 9.5 (Sampled Function): A sampled function f s x is:
1 x
f s x =f x III
x x
1 x
f x x x n f x x nx
n n
The sampled function is expressed by the product of f x with the
dilated comb function. The convolution of a function and a comb
produces a periodic replication:
1 x
f x III f x nx
x x n
Figure 9.1 shows a function f x sampled at points x k x can be
1 x
modeled as a weighted impulse train f s x =f x III . This
x x
314 Mathematics for mechanical engineers
sampled function is also known as the impulse-sampled function. From
Theorems 9.5 and 9.10, the Fourier transform of the impulse-sampled
function is:
1 x
f s x f x III F III x
x x
and the Fourier transform of a periodic replication is:
1 x
f x nx f x x III x F III x
n
s
Figure 9.1. Impulse-sampled function
The first relationship says that the Fourier transform of a sampled
function is a periodic replication of the Fourier transform of the
unsampled function. The latter relationship says that the Fourier
transform of a periodically replicated function is an impulse sampling of
the Fourier transform of the base function.
Fourier transform 315
Definition 9.6 (Bandlimited Functions): A function f x is bandlimited
if its Fourier transform F is zero for frequencies higher than a
finite value B.
Theorem 9.14 (Sampling): Suppose f is a bandlimited function,
F 0 for B 0 , and have finite energy, f L2 . If
1
2B , then f can be reconstructed from samples f nx as:
x
x n x
f x f nx sinc x
n
1
When the continuous analog signal is sampled at a frequency s , the
x
resulting discrete signal has more frequency components than did the
analog signal. To be precise, the frequency components of the analog
signal are repeated at the sample rate. How many samples are necessary
to ensure we are preserving the information contained in the signal? If
the signal contains high frequency components, we will need to sample at
a higher rate to avoid losing information that is in the signal. In general,
to preserve the full information in the signal, it is necessary to sample at
twice the maximum frequency of the signal. This is known as the
Nyquist rate. The Sampling Theorem states that a signal can be exactly
reproduced if it is sampled at a frequency s where s is greater than
twice the maximum frequency in the signal.
9.4 Discrete Fourier Transform
In general, the Fourier transform of a function is not always found
analytically or looked up in a table of transform pairs. Thus, numerical
computation of the Fourier transform is appropriate. To compute the
Fourier transform numerically we need to discretize the Fourier integrals.
General methods for computational integration are described in
numerical analysis texts. Many algorithms are available in mathematical
software packages (e.g., the fft function in Matlab). While the details
vary, all numerical integration methods are based on approximating an
integral:
316 Mathematics for mechanical engineers
N 1
f x dx f xn xn 1 xn
n 0
Figure 9.2. Approximation of an integral
Figure 9.2 shows an integral approximated by the sum of N rectangular
areas. The height of the nth rectangle is f xn , and its width is xn 1 xn .
A rectangular approximation to the Fourier transform is:
F f x e i 2 x dx
X /2
f x e i 2 x dx
X /2
N
f xn e i 2 xn x
n0
X X
where the interval , is chosen such that f is acceptably close to
2 2
X X
zero for x , and x is the sampling interval.
N N
Fourier transform 317
9.4.1 Forward Transform
Let's begin with a rectangular approximation to the forward Fourier
transform, using a uniform sampling grid, x nx :
F
f x e i 2 x dx f nx e
n
i 2 nx
x
This is called the Discrete Fourier Transform (DFT). The sampling
interval Δx should be chosen sufficiently small to avoid aliasing. That is,
1
F is effectively bandlimited to . The complex exponential
2x
1
ei 2 nx is periodic in with period s . Thus, the approximation is
x
also periodic with period s and it is sufficient to evaluate it over a single
period, s , s .
2 2
Assume that a uniform sampling grid with M sample points is
implemented on the frequency range:
M M
m , m ,...,0,..., 1
2 2
where
1 1
2x 2x 1
s
M M x M
is called the resolution in the frequency domain and this quantity
should be chosen small enough (by choosing M sufficiently large) to
reveal the fine features of F. From these representations, the
approximation is obtained:
318 Mathematics for mechanical engineers
N
2
F m f nx e
N
i 2 m nx
x
n
2
N
2 mn
2 i M M
N
f nx e M
x, m
2
,..., 0,..., 1
2
n
2
Notice that, the very important step in making the sum in a DFT is setting
M N so the vectors ( f nx ) and ( F m ) have the same number
N
of samples. For convenience, we can shift the vector f nx 2
N into
n
2
vector f n n 0 as follows:
N 1
N
f n N x , n
2
,..., 1
f c n
f nx , N
n 0 ,..., 1
2
Thus,
N 1 i 2 mn
F m f c n e N
x
n 0
9.4.2 Inverse Transform
The derivation of a DFT approximation for the inverse Fourier transform
follows the same steps, leading up to:
N
1 2 mn
2
N N
F m e
i
f nx N
, m ,...,0,..., 1
m
N 2 2
2
N
into vector Fc m m0 as follows:
N 1
We can shift the vector F m 2
N
m
2
Fourier transform 319
N
F m N , m
2
,..., 1
Fc m
F m , N
m 0 ,..., 1
2
Therefore,
N 2 mn
f nx Fc m e
i
N
m 0
9.5 Spectrum Analysis
In the time domain the amplitude of electrical signals is plotted versus
time, a display mode that is customary with oscilloscopes. The
relationship between the amplitude of a simple sinusoidal signal
f t A sin 0 t and the time is shown in Figure 9.3.
Figure 9.3. Simple sinusoidal signal
However, for complicated signals that consist of different sinusoidal
components, their presentations give only little information. For example,
Figure 9.4 presents a record of a sound wave. We don't know what
320 Mathematics for mechanical engineers
frequencies are contained in the wave, and we don't know how the
loudness of each frequency component is.
To reveal component properties of a signal, the frequency spectrum is
emerged as a very important tool for signal processing. The frequency
spectrum of a periodic signal is the distribution of the amplitudes and
phases of each frequency component against frequency. The frequency
spectrum is computed by mean of DFT. The frequency spectrum is then
presented as the graph of amplitude of frequency component versus
frequency.
Figure 9.4. A complicated signal
Recall that, any signal that is periodic in the time domain can be derived
from the sum of sine and cosine signals of different frequency and
amplitude. Such a sum is referred to as a Fourier series. In complex form
the signal with period of T can be expanded as:
2 nt
i
f t cn e T
n
Applying the Fourier transform for a signal f t in the interval [0, T]
using Equation 9.2 in Section 9.1, we have:
Fourier transform 321
T
F f t e i 2 t dt f t e i 2 t dt
0
2 nt
cne ce
T i T
T
e i 2 t dt n
i 2 n
e i 2 t dt
0 0
n n
ce
T i 2 n t
n dt
0
n
n
here: n
T
Applying the orthogonal property of the Fourier transform, we have:
0 n
ce
T
F
i 2 n
dt T
0 cn dt cnT n
0 n
n
This says that the Fourier transform of a periodic signal gives coefficients
cn of the Fourier series of the signal at corresponding frequency 2T n .
Thus, the coefficients cn can be calculated from the Fourier transform of
the signal as:
1
cn F
T
Now, we apply the DFT to compute the coefficients of Fourier series of
the signal f t in the interval 0,T :
2 mn
1 1 1 N 1 i
cn F
T N t
F m fc n e N t
N t n 0
N 1 2 mn N 1 2 mn
1 i 1 i
N
f n e
n0
c
N
N
f nt e
n 0
N
Example 9.2: Let f t is a signal that includes two harmonic
components:
322 Mathematics for mechanical engineers
f sin 2 * 40t 5sin 2 *80t , t 0, 2
Use function fft in Matlab to calculate the spectrum of f .
Solution: Performing the discrete Fourier transform of f to get F
using the above formula then plot F with respect to , we have the
frequency spectrum. The discrete Fourier transform is performed by
using the fft function in Matlab. The Matlab codes are presented here,
where N=1024.
Figure 9.5. Frequency spectrum
As can be seen from the spectrum presented in Figure 9.5, the harmonic
with frequency of 40 Hz has amplitude of 1, while the harmonic with
frequency of 80 Hz has amplitude of 5 as described in the given equation.
Since the spectrum is symmetric over the y-axis, we just present only one
side of the spectrum. The amplitude of one side is therefore doubled.
N=1024;
dt=1/N; t=-1:dt:1;
omega1=40*2*pi;
omega2=80*2*pi;
f=sin(omega1*t)+5*sin(omega2*t);
Fourier transform 323
NFFT=2^nextpow2(N);
F=abs(fft(f,NFFT)/N);
F=2*F;
plot(F(1:round(NFFT/2));
xlabel('Frequency','color','black','fontsize',10);
ylabel('Amplitude','color','black','fontsize',10);
324 Mathematics for mechanical engineers
EXERCISES
I. Calculate following Fourier coefficients:
1 1
1. fˆ xe ix dx
2 1
1
1 x e
1
2. fˆ ix
dx
2 1
1 /2
3. fˆ cos xe ix dx
2 /2
1 /2
4. fˆ sin xe ix dx
2 /2
II. Calculate the transform of the following function by using Fourier
theorems:
1. f x t
2. f x 1
3. f x e at H t
4. f x e at H t
5) f x e
a t
6. f x te at H t
7. f x rect x 1 ei x
8. f x e x
2
Fourier transform 325
ANSWERS
I.
1.
1 1 1 1
fˆ xe i x dx x cos x i sin x dx
2 1
2 1
sin x x cos x
1
i 1 2i 1
2 1 x sin x dx 2 1 2 dx
0
i 2 sin cos
2
2.
1 1 1 1
fˆ 1 x e i x dx 1 x cos x dx
1
2 2 1
2 1
2 1 x cos x dx
0
sin x
1
2 1 2 1
0 sin x dx
2 0
1 x dx
2 0
2 1 cos
2
3.
1 /2 2 /2
fˆ cos xe i x dx cos x cos x dx
2 /2
2 0
2 /2
cos x 1 x cos x 1 x dx
2 0
/2
2 sin x 1 x sin x 1 x
2 x 1 x 1 0
326 Mathematics for mechanical engineers
cos
2 2
1
12
Treat the case 1 separately and we will have
2 /2 2
fˆ 1 cos x cos xdx
2 0 4
II.
1. Applying properties of the Dirac delta function we have:
x x e i x dx e 0 1
2. Applying properties of the Dirac delta function we have:
1 1 0 1
1 ei x d
e
2 2 2
1
1 1 2
2
4. Applying properties of the Heaviside function, we have:
e at H t e at H t e it dt e at e it dt
0
u
1 1 1
lim u e i a t lim u e i a u
i a 0 a i i a
2 2
a
1 1 u 1
lim u e a i
a i i a a i
Laplace transform 327
CHAPTER 10. LAPLACE TRANSFROM
We have earlier learned how to define and compute the Fourier
transform for functions which are absolutely integrable. The Laplace
transform extends the capability of the Fourier transform to certain
important functions of rapid growth which are not Fourier
transformable. The Laplace transform, named after its France inventor
Pierre-Simon Laplace (1749-1827), is an integral transform that
converts a function of a real variable t (often time) to a function of a
complex variable s (complex frequency). The transform has many
applications in science and engineering because it is a tool for solving
differential equations. In particular, it transforms linear differential
equations into algebraic equations and convolution into multiplication.
This chapter presents basic definitions and properties of the Laplace
transform with illustration examples. The inverse Laplace is then
defined and its applications for solving ordinary differential equations
are provided to show the power of the Laplace transform and its
inverse transform. Finally, some engineering problems are presented
and solved by using Laplace transforms.
10.1 Definition, Basic Properties
Definition 10.1: Let F : 0 , . If F t e t L1 for some real 𝜎,
then the Laplace transform of F is:
f s F t e st dt
0
where s i i 2 . The operator notation for the Laplace
transform is F t .
Definition 10.2: A function F t is a said to be exponential order c if
there exists constants c, M 0, T 0 such that F t Me ct for all
t T .
328 Mathematics for mechanical engineers
This says that the function is of exponential order if F t is slower
growing in comparison with ect , that is, there exists constant c such
F t
that limt ct 0 . Thus, we can see from Definition 1 that not
e
every function has a Laplace transform, but only functions that are of
exponential order.
Theorem 10.1 (Linearity): If F1 t and F2 t have their Laplace
transforms, then:
aF1 t bF2 t a F1 t b F2 t
Proof: We have,
aF1 bF2 aF1 bF2 e st dt
0
aF e bF2 e st dt
1
st
0
a F1e st dt b F2 e st dt
0 0
a F1 b F2
This establishes the theorem.
Theorem 10.2 (First Shift): If there exists constants M and α such that
F t Me at , that is, F t is of exponential order, then:
e bt F t f s b
provided b a . (In practice if F t is of exponential order then the
constant a can be chosen such that this inequality holds.)
Proof: Using directly the Definition 10.1, we have:
Laplace transform 329
e bt F limT e bt Fe st dt
T
0
e bt Fe st dt
0
since the limit exists
s b t
Fe dt
0
f s b
This establishes the theorem. This theorem is very useful once we
have established a few elementary Laplace transforms as follows.
Example 10.1: Find the Laplace transform of the function F t 1 .
Solution: The Definition 1 gives:
1 1
1 e st dt e st
0 s 0 s
Example 10.2: Find the Laplace transform of the function F t t .
Solution: Using directly the definition of Laplace transform, we have:
T
t limT te st dt
0
Then, integrating by parts we get:
T
T t T t
te dt e st e st dt
st
0 s 0
0 s
T
T sT t st
e 2e
s s 0
T sT 1 sT 1
e 2e 2
s s s
We need to evaluate the limit of the last expression:
330 Mathematics for mechanical engineers
T 1 1
limT e sT 2 e sT 2
s s s
T 1 1
limT e sT limT 2 e sT limT 2
s s s
T 1 1
limT sT limT 2 sT limT 2
se se s
1
00 2
s
Thus, we have:
1
t
s2
This result can be generalized as the following corollary.
n!
Corollary 10.1: t n , where n is a positive integer.
s n 1
Proof: Using the Definition 1, we have:
tn nt n 1 st n
t e t n 1
n
t e dt e st
n st
0 s 0
0 s s
We will prove this by induction as follows. If we put n 2 in this
recurrence relation we obtain:
2 2 2!
t 2 t 3 21
s s s
Assume that:
n!
t n
s n 1
then, we need to prove:
Laplace transform 331
t n 1
n 1 !
s n2
Put m = n + 1 then:
m ! n 1 ! n 1 !
t m n2
s m 1 s n 1 1 s
This proves the Corollary by induction.
Example 10.3: Find the Laplace transform of functions te at and t n eat ,
where a is a real constant and n a positive integer.
Solution: Using the First Shift Theorem with b a gives:
F t eat f s a
1
So, if F t t then f s .
s2
Thus,
1
F t e at
s a
2
n!
Now, if F(t) = tn then f s . Therefore:
s n 1
n!
t n e at
s a
n 1
Example 10.4: Find the Laplace transform sin t and cos t .
Solution: We know that:
eit cos t i sin t
332 Mathematics for mechanical engineers
Let's calculate the Laplace transform of this expression:
cos t i sin t eit
eit e st dt
0
e
i s t
dt
0
e
is t
is 0
e
is t
1
lim t
is is
We evaluate the limit:
1 s 2 t
e i s t e i s t e i s t e
limt limt limt 0
i s i s e i s t
i s e i s t
Thus,
1
eit
is
s 1
2 i 2
s 1 s 1
cos t i sin t
cos t i sin t
Equating real and imaginary parts, the two results are obtained:
s
cos t
s 1 2
1
sin t 2
s 1
Laplace transforms of these trigonometric function are very useful. So
remember these results and we will use them in the next sections.
Laplace transform 333
Theorem 10.3: If F t f s then,
d
tF t f s
ds
and in general:
dn
t n F t 1 f s .
n
ds n
Proof: Let's write down the definition of Laplace transform:
F t e st F t dt
and differentiate this with respect to s to give:
d d
f s e st F t dt te st F t dt tF t
ds ds 0 0
Assume that two integrals in this equation are absolute convergence,
so we can interchange the differentiation and the improper integration:
d
tF t f s
ds
We now prove the theorem in general case by induction. Assume the
result holds for n, that is:
dn
t n F t 1 f s
n
ds n
Differentiating both sides with respect to s (assuming all appropriate
convergence properties), yields:
dn
t n 1e st F t dt 1 f s
n
0 ds n
334 Mathematics for mechanical engineers
or
d n 1
t n 1e st F t dt 1 f s
n 1
0 ds n 1
Thus,
d n 1
t n 1 F t 1 f s
n 1
ds n 1
This establishes the theorem for the general case.
Example 10.5: Find the Laplace transform t sin t .
1
Solution: If F t sin t then f s . Using Theorem 10.3 we
s 1
2
have:
d d 1
t sin f s
ds ds s 2 1
2s
s 1
2 2
This is the required result.
10.2 Further Laplace Transform Theorems
Theorem 10.4 (Dilation): If F t has Laplace transform f s , then:
1 s
F at f
a a
Proof: We use the definition of Laplace transform,
F at e st F at dt
0
Laplace transform 335
du s
Changing the variable u at then dt , st u and replacing
a a
these into the integral, we have:
1 as u 1 s
F at e F u du f
a 0 a a
This is the proof of theorem.
Example 10.6: Evaluate Laplace transform sin t .
Solution: We have already known that:
1
sin t
s 12
Applying the Dilation Theorem, we have:
1 1
sin t
s 2
s 2
2
1
Theorem 10.5 (Derivative): If F t is a differentiable function and
has Laplace transform f s , then:
F t e st F t dt F 0 sf s
Proof: Integrating by parts once, we obtain:
F t e st F t se st F t dt
0 0
F 0 sf s
This is an important result that can be applied in many applications,
especially in solving linear differential equations. The interesting
property is that the Laplace transform of a derivative F t does not
336 Mathematics for mechanical engineers
itself involve a derivative, only F 0 sf s which is an algebraic
expression involving f s . The only thing is that the value F 0 is
required to complete the transform. In solving differential equation,
the integration constant F 0 known as the initial condition is in
general provided. We will investigate this problem through solving
differential equations.
Theorem 10.6: If F t is a twice differentiable function of t then
F t s 2 f s sF 0 F 0
Proof: Integrating by parts twice, gives:
F t e st F dt
0
e st F se st F dt
0 0
F 0 se st F s 2 e st Fdt
0 0
F 0 sF 0 s f s
2
This establish the result. The general result which can be proved by
induction, is
F
n
t s n f s s n1 F 0 s n 2 F 0 ... F n 1 0
where n is a positive integer. Observe that there are n constants on the
right hand side. This because we integrate by parts n times. This result
has wide applications.
Example 10.7: Evaluate cos t .
Solution: Since we have known sin t , so we can derive easily
cos t by the Derivative Theorem:
Laplace transform 337
1
cos t sin t
1
s 2 sin 0
s 2
s
2
s 2
Heaviside’s Unit Step Function
Heaviside’s unit step function H t is defined as follows:
0 t0
H t
1 t0
Thus
1
H t e st H t dt e st .1dt
0 0 s
The Laplace transform of H t t0 , t 0 , is:
H t t0 He st dt e st dt
0 t0
e st e st0
s t0 s
This result is generalized through the following theorem.
Theorem 10.7 (Second Shift): If F t is a function of exponential
order in t, then:
H t t0 F t t0 e st0 f s
Proof: Using directly the definition of Laplace transform, yields:
338 Mathematics for mechanical engineers
H t t0 F t t0 e st H t t0 F t t0 dt
0
e st F t t0 dt
t0
F u du put u t t0
s u t0
e
0
e st0 e su F u du
0
e st0
f s
This gives the theorem.
10.3 The Inverse Laplace Transform
We know that many operations have inverses, for example, addition
has subtraction, multiplication has division, differentiation has
integration, Fourier transform has inverse Fourier transform. The
Laplace transform is no exception, and we can define the Inverse
Laplace transform as follows.
Definition 10.3: Assume F t has the Laplace transform f s , that
is:
F t f s
then the inverse Laplace transform is defined by:
1 f s F t
and is unique apart from null functions. In engineering problems, the
most important property of the inverse transform is probably its
linearity. This property can be stated as the following theorem.
Theorem 10.8 (Second Linearity):
1 af1 s bf 2 s a1 f1 s b1 f 2 s
Laplace transform 339
Proof: Linearity is easily established as follows. Since the Laplace
transform is linear, we have for suitably well behaved functions F1 t
and F2 t :
aF1 t bF2 s a F1 s b F2 s
af1 s bf 2 s
Taking the inverse Laplace transform of this expression gives
aF1 t bF2 t 1 af1 s bf 2 s
From the definition of the reverse Laplace transform, we have:
F1 t 1 f1 s
F2 t 1 f 2 s
and,
aF1 t a1 f1 s
bF2 t b1 f 2 s
Thus,
aF1 t bF2 t a1 f1 s b1 f 2 s
Finally,
a1 f1 s b1 f 2 s 1 af1 s bf 2 s
This establishes linearity of 1 f s .
a
Example 10.8: Find 1 2 2
s a
340 Mathematics for mechanical engineers
Solution: Using the partial fraction decomposition:
a 1 1 1
s a
2 2
2 s a s a
Notice that,
1
L 1
s
Using the First Shift Theorem:
1
L e at .1 f s a
sa
1
L e at .1 f s a
sa
Taking the inverse Laplace transform, we finally find:
1 1 1 1 at at
L1 e e sinh at
2 s a s a 2
s 2
Example 10.9: Find 1 3
s 3
Solution: Using the partial fraction decomposition:
s2 1 6 9
s 3 s 3 s 3 s 3 3
3 2
Applying the First Shift Theorem, we have:
s 2 1 1 6 1 9
1 3
1 2
3
s 3 s 3 s 3
s 3
Laplace transform 341
9
e3t 6te 3t t 2 e 3t
2
where we have used the linearity property of the 1 operator. To
improve the computational skill, we can do the following examples.
Example 10.10: Determine the following inverse Laplace transform:
1
s 3
s s 1 s 2
Solution: Using partial fraction decomposition, we get:
s 3 3 4 1
s s 1 s 2 2 s 3 s 1 6 s 2
Applying the inverse Laplace transform operator gives:
1
s 3 1 3 1 4 1 1
s s 1 s 2 2s 3 s 1 6 s 2
3 1 4 1 1 1
1 1 1
2 s 3 s 1 6 s 2
3 4 1
e t e 2 t
2 3 6
Theorem 10.9: If the Diract delta function is defined as having the
following properties
0 t0
t
0 t 0
h t t dt h 0
t dt 1
for any function h t continuous in , , then
342 Mathematics for mechanical engineers
t 1
Proof: The definition of the delta function gives:
0
h t t dt h t t dt h t t dt h 0
0
If h t e st , we have
e st t dt e st t dt
0 0
lim 0 e st t dt
0
e st t dt
0
e0 1
Thus,
t 1
Function t in Laplace transform theory is usual used to define the
impulse function.
s2
Example 10.11: Determine 1 2
s 1
Solution: We have:
s2 1
1 2 1 1 2
s 1 s 1
1
1 1 1 2
s 1
t sin t
Laplace transform 343
Theorem 10.10 (Convolution): If f t and h t are two functions of
exponential order, and denote f t f s , h t h s as
the two Laplace transforms then:
1 f s h s f t h t
is the convolution operator introduced above. Here we define that the
convolution of two given functions f t and g t is written f g is
the integral:
t
f h f h t d
0
Proof: We have:
f t h t e st f h t d dt
t
0 0
t
e
st
f h t d dt
0 0
This is the double integral with respect to variables and t. Domain of
t , can be sketched as Figure 10.1. Changing the order of
integration, we have:
t
e st f h t d dt e st f h t dtd
0 0 0
f e st h t dtd
0
Now we change the variable u t in
the inner integral so that it becomes:
e st h t dt e su h u du
0
t
e st e su h u du
0
e h s
st
Figure 10.1. Domain of
344 Mathematics for mechanical engineers
Thus
f h e st h t dt
f e s h s d
h s f s
f s h s
This establish the theorem.
10.4 Laplace Transform for Solving Ordinary
Differential Equations
An ordinary differential equation ODE is an equation where the
unknown is in the form of a derivative. The word "ordinary" means
that there is only differentiation with respect to a single independent
variable so that the solution is the required function of this variable.
Laplace transform of a derivative function can eliminate the
derivative, replacing each differentiation with a multiple of s. Thus, it
is interesting that Laplace transforms are a handy tool for solving
certain types of differential equation. In this text book, we will learn
how to solve the first and second order differential equations using
Laplace transform.
The Laplace transform of a derivative was found easily by direct
integration by parts. Theorem 10.5 and 10.6 give two useful results
that will be used extensively here:
F t sf s F 0
and
F t s 2 f s sF 0 F 0
where the prime denotes differentiation with respect to t. Taking
Laplace transforms of a linear ODE, the derivatives themselves
disappear. The ODE is transformed into the Laplace transform of the
Laplace transform 345
function multiplied by s (for a first derivative) or s2 (for a second
derivative). Moreover, the constants also appear in the form of F 0
(for a first order ODE) and F 0 (for a second order ODE). If F 0
and F 0 (or the first only) are given, the ODE is called initial value
problems, that is, there are enough information to solve the problems.
10.4.1 First Order Differential Equations
Example 10.12: Solve the first order differential equation:
y t 3 y t 0
where y 0 1 .
Solution: Here we have changed F t to the more usual y t .
Taking Laplace transforms leads to:
y t 3 y t 0
This means that
sy s y 0 3 y s 0
where y denote Laplace transform. Since y 0 1 , solving for y s
we have
1
y s
s3
Whence,
1
y t 1 e 3t
s3
346 Mathematics for mechanical engineers
using the standard form. That this is indeed the solution is easily
checked. Let us look at the same equation, but with a different
boundary condition.
This differential equation is simple and can in fact be solved by a
variety of methods. It is useful to check the answer by solving again
using, for example, separation of variables or integrating factor
methods as these will be familiar to most students.
Example 10.13: Solve the first order differential equation:
y t 3 y t 0
where y 1 1 .
Solution: Taking the same procedure as before, but we don't know the
value of y (0) , so we have the solution:
y 0
y t 1 y 0 e 3t
s3
We now use the given boundary condition to obtain:
y 1 y 0 e 3 1
which gives
y 0 e3
Thus, the solution is
y t e3e 3t e31t
We will deal with a slightly more challenging problem in the next
example.
Example 10.14: Solve the differential equation:
Laplace transform 347
y 3 y cos 3t
given y 0 0 .
Solution: Taking the Laplace transform, we get:
s
sy s y 0 3 y s
s 9
2
Applying the initial condition and solving this for y s , yields:
s
y s
s 3 s 2 9
This solution is in the form of the product of two known Laplace
transforms. Thus we can invert either using partial fractions or the
Convolution Theorem. Here, let's use the Convolution Theorem.
Notice that:
1
1 e et
s 3
s
1 cos 3t
s 92
Using the Convolution Theorem yields:
s
y t 1
s 3 s 2 9
t
e3 t cos 3 d
0
t
e 3t e3 cos 3 d
0
1 1 1
e 3t e3t cos 3t e3t sin 3t
6 6 6
1 1 1
cos 3t sin 3t e 3t
6 6 6
348 Mathematics for mechanical engineers
10.4.2 Second Order Differential Equations
We consider some examples here to know how Laplace transforms are
used to solve second order ordinary differential equations. We will see
that the technique is similar as for solving the first order ODEs, but
finding the inverse Laplace transform is often more challenging. To
begin with, we find the solution of a homogeneous second order ODE
that is well-known in the vibration problem.
Example 10.15: Use Laplace transforms to solve the equation:
y t y t 0
with y 0 1, y 0 0 .
Solution: Taking the Laplace transform of this equation, we obtain:
s 2 y s sy 0 y 0 y s 0
Replacing the initial conditions into this equation, yields:
s
y s
s 1
2
This is a standard form that we can inverts to
s
y t 1 cos t
s 1
2
This is the correct solution of a simple harmonic motion problem that
can easily be checked. Since we have the solution for the
homogeneous differential problem, we are now ready to solve the
inhomogeneous problem that follows.
Example 10.16: Use Laplace transforms to solve the equation:
y t y t t
Laplace transform 349
with y 0 1, y 0 0 .
Solution: Apart from the trivial change of variable, we follow
Example 15 and take Laplace transforms to obtain
1
s 2 y s sy 0 y 0 y s
s2
Applying the initial conditions y 0 1, y 0 0 , gives:
s 1
y s 2 2
s 1 s s 1
2
Using inverse Laplace transform with the use of Convolution
Theorem, we have:
s 1 1
y t 1 1 2 2
s 1
2
s s 1
cos t
0
t sin d
cos t sin t t
The first two terms are the complementary functions and the third is
the particular integral. The whole is easily checked to be the correct
solution. The Laplace transform method provides the solution of the
differential equation with a general right hand side in a simple and
straightforward manner.
10.4.3 Some Applications of Laplace Transform to
Mechanical Engineering
10.4.3.1 Motion of a falling object
Consider a parachutist jumping from an airplane towards Earth with
the initial velocity v 0 v0 . Assume that forces applied to the
parachutist consist of the gravity and air resistance. We need to find
the velocity of the parachutist as a function of time. To solve this
350 Mathematics for mechanical engineers
problem, applying Newton’s second law we can express this by the
equation:
dv
m f
dt
Here: m, v and f are the mass, velocity of the object. f is the total
forces acting on the parachutist.
Figure 10.2. A parachutist
The force caused by gravitational acceleration g is the same as the
object's weight and equal to mg. The air resistance acting on the object
can be expressed by -cv where c is the resistant constant which
depends on the density of the air and the shape of the parachutist.
Thus, the motion equation of the object can be rewritten as:
dv
m mg cv
dt
or
dv c
m vg
dt m
Taking the Laplace transform we have:
Laplace transform 351
c g
s v t v 0 v t
m s
or
c c g
s v s v0 v s
m m s
This can lead to the partial fraction equation:
g v0
v s
c c
ss s
m m
A B C
s s c s c
m m
g 1 1 v
0
mc s s c s c
m m
Using the inverse Laplace transform with the use of Convolution
Theorem, yields:
g t
c c
t
v t 1 e v0 e
m m
mc
This is the velocity of the falling parachutist with the initial velocity
v 0 v0 as a function of time.
10.4.3.2 Vibration of a mass-spring-damper system
Consider the motion equation of an object under an excitation force:
my t cy t ky t f t
352 Mathematics for mechanical engineers
where m, c and k are constants. In mechanics, m is the mass, c is the
damping, k is the stiffness and y(t) itself is the displacement of the
mass, f(t) is external force. The right-hand side is called the external
force or excitation. In terms of systems engineering, f t is the
system input, and y t is the system output. Since m, c and k are all
constant the system described by the equation is linear and time
invariant.
Figure 10.3. Spring-damper-mass system
Applying the Laplace transform for this second order differential
equation, yields:
m s 2 y s sy 0 y 0 c sy s y 0 ky s f s
Solving this equation for y s , gives:
f s ms c y 0 my 0
y s
ms 2 cs k
Theoretically, y t can be found by taking inverse Laplace transform.
Let's consider the simplest case to consider is when y(0) and y 0 are
both zero. The output is then free from any difficulties that might be
caused by special initial conditions. In this special case we have:
Laplace transform 353
f s
y s
ms 2 cs k
This equation is in the standard form: “response = transfer function ×
input”. This explains why we are interested in using Laplace
transforms for engineering problems.
Example 10.17: Consider a mass-spring-damper system with m = 1, c
= 5, k = 6 and f t 2e t . The initial conditions are:
y 0 1, y 0 0 . Find the vibration of the mass-spring-damper
system by using Laplace transform.
Solution: We need to solve the second order differential equation by
using Laplace transform:
y t 5 y t 6 y t 2e t
Taking the Laplace transform of this equation we get:
2
s 2 y s sy 0 y 0 5 sy s y 0 6 y s
s 1
Here, the tilde notation implies Laplace transform. Substituting the
initial conditions into this equation to form the equation for y s ,
yields:
2
s 2
5s 6 y s
s 1
s 5
Factorizing this, gives:
2 s5
y s
s 1 s 2 s 3 s 2 s 3
Using the partial fraction method, we have:
354 Mathematics for mechanical engineers
1 1 1
y s
s 1 s 2 s 3
Finally, applying the inverse Laplace transform, the solution of the
second order differential equation is obtained:
y t e t e 2 t e 3 t
If we want to know which terms are the elementary functions and
which one is particular solution, we need to solve the homogeneous
problem first. In this case, the right side of given equation is zero, then
2
the term in Equation 1 is cancelled, so:
s 1
s5 3 2
y s
s 2 s 3 s 2 s 3
Thus, the solution of the homogeneous equation is found by taking the
inverse Laplace transform:
y t 3e 2t 2e3t
Comparing Solution 2 and 3 we see that, the first term in Equation 2 is
the particular solution and the last two terms are the complementary
functions. Obviously, each term of the solution approaches zero very
quickly. Thus, the whole solution is overdamped and therefore
nonoscillatory. Although this is the easiest case and it is also least
physical meaning as no vibrating period is done, this example serves
to demonstrate the power of the Laplace transform to solve this kind
of ordinary differential equation.
Example 10.18: Find the vibration response of a mechanical system
described by the second order differential equation using Laplace
transform:
y t 6 y t 9 y t sin t
Laplace transform 355
with the initial conditions y 0 1, y 0 0 .
Solution: If there is no force acting on the mechanical system, that is,
the right hand side is zero, so with zero initial conditions the system
would be stationary, y t 0 . However, with the sinusoidal forcing,
the solution will be quite interesting. Taking Laplace transforms, the
equation becomes:
1
s 2 y s sy 0 y 0 6 sy s 6 y 0 9 y s
s 1
2
Replacing the initial conditions and rearranging, gives:
1 1
y s
s 1 s 3 2
2
We can apply either partial fractions or convolution. Here, we apply
Convolution. Note that:
1
sin t
s 1
2
1
te 3t
s 3
2
So, taking the invers Laplace transform, we obtain:
1 1 t
1 2 2
e 3 sin t d
s 1 s 3
0
e 3t 3 2
5t 3 cos t sin t
50 50 25
The first term is the particular solution and is called the transient
response by engineers since it dies away for large times. The final two
terms are the complementary function and are called the steady state
response by engineers since it persists as long as the force remains.
After a “long time” the influence of the first term vanishes, the
356 Mathematics for mechanical engineers
response is harmonic at the same frequency as the forcing frequency.
The “long time” in practice may be quite short depending on
particular circumstance. The figure here presents the graph of the
output y t . As can be seen from the Figure 10.4, the output is
clearly sinusoid after about t=0.5. However, the amplitude and phase
of the input and output are different. Indeed, the steady state solution
can be combined to give a sinusoid:
3 2 1
cos t sin t sin t
50 25 10
Figure 10.4. The output signal
with amplitude and phase form. Where, 1/10 is the amplitude of the
vibrating response and is the phase difference between input and
output. Here, cos 4 / 5 , sin 3 / 5 . It is worth noting that the
response frequency is the same as the forcing frequency that is
important in practical applications.
10.4.3.3 Deflection of a beam
Let's consider a beam with non-uniform loading as shown in Figure 5.
Assume that the flexural rigidity k = EI, where E and I are the Young's
Laplace transform 357
modulus of the moment of inertia of the beam, respectively. f(x) is the
transverse force per unit length along the beam. The governing
bending equation of the beam can be expressed as:
d4y
k 4 f x (10.1)
dx
y
f(x)
Figure 10.5. A beam with non-uniform loading
We will use Laplace transforms to find the deflection of beam under
the non-uniform force. In this problem, the fourth derivative of y x
with respect to the space variable x 0, L is present so in order to
use Laplace transforms the domain needs to extend so that x 0,
and the Heaviside step function will be applied. First, we define the
Laplace transform as follows:
y s e xs y x dx
0
Then, taking the Laplace transform of Equation 10.1 , we have:
k s 4 y s s 3 y 0 s 2 y 0 sy 0 y 0 f s
Thus,
f s y 0 y 0 y 0 y 0
y s 4
2
3
(10.2)
ks s s s s4
However, the length of beam is finite, so we need to use the
Heaviside’s step function to describe the domain of the beam. Here we
358 Mathematics for mechanical engineers
replace y x by the combination y x 1 H x L then this
function will satisfy conditions for the existence of the Laplace
transform. This means that:
y s e xs y x 1 H x L dx
0
If the forcing is known then the solution y x will be found by
inversion. In general, the Convolution Theorem is particularly useful.
Indeed, put:
3!
g s
s4
then
g x x3
Applying the Convolution Theorem for function f x and g x , we
have the following equation:
3!
1 g s f s 1 4 f s x f d
x 3
s 0
Taking the inverse Laplace transform Equation 10.2 we have:
f s
1 y s 1 4
ks
y 0 1 y 0 1 y 0 1 y 0
1 2 3 4
s s s s
or
1 x
y x 1 H x L x f d
3
6k 0
(10.3)
1 1
y 0 xy 0 x 2 y 0 x 3 y 0
2 6
Laplace transform 359
If the beam is simply supported at both ends, then boundary
conditions of the beam are:
y 0 0, y L 0
y 0 0 y L 0
These equations describe that the displacements and moments at two
ends are zero. From these conditions, the four constants
y 0 , y 0 , y 0 , y 0 in Equation 10.3 will be found. From
y 0 0, y 0 0 we have:
1 x
y x 1 H x L x f d
3
6k 0
(10.4)
1
xy 0 x 3 y 0
6
The application of boundary conditions at x L is less easy. One
method is to put u x y x and Equation 10.1 becomes:
d 2u
k f x
dx 2
Applying Laplace transforms as before, the solution of this equation is
found in the form:
1 x
u x 1 H x L x f d u 0 xu 0
k 0
Substituting u x y x into this equation, yields:
1 x
y x 1 H x L x f d
k 0
y 0 xy 0
From this equation we now have:
1 x
y 0 y x 1 H x L y 0 x f d
k 0
360 Mathematics for mechanical engineers
Applying the boundary condition y 0 0 for this equation, we
have:
1 1 x
y 0 y x 1 H x L x f d
x kx 0
Thus, applying the boundary conditions at x = L:
y L 0
we obtain:
1 L
y 0 L f d
kL 0
(10.5)
From equation 10.4 we have:
1 1 x 1
y 0 y x 1 H x L x f d x 2 y 0
3
x 6kx 0 6
Applying the boundary conditions at x = L again for this equation,
yields:
1 L 1
y 0 L f d L2 y 0
3
6kL 0 6
Substituting Equation 10.5 into this equation we obtain:
1 L L L
y 0 L f d L f d
3
(10.6)
6kL 0 6k 0
From Equations 10.4, 10.5 and 10.6, the general solution to the
problem in terms of integrals is derived:
1 x
y x 1 H x L x f d
3
6k 0
(10.7)
1 L
x L 2 L x f d
6kL 0
2 2
It is now possible to insert any loading function into this expression
and calculate the displacement caused.
Laplace transform 361
Example 10.19: Find the deflection of a beam with length of L = 3 and
the flexural rigidity k = 1 subjected to a uniform loading of unit
magnitude f = 1.
Solution: The deflection of the beam can be easily derived by
calculating the integrals in Equation 10.8:
1 x 1 L
y x x d x L 2 2 L x 2 d
3
6 0 6 0
4 3
x x 9x
, x 0, L
24 4 8
This deflection of the beam is depicted in Figure 5.
Figure 5. Deflection of beam
362 Mathematics for mechanical engineers
TABLE OF LAPLACE TRANSFORMS
The following Table presents Laplace transforms of some of the
commonly used formulas. In this table, t is a real variable, s is a
complex variable and a, b and x are real constants.
f s F t F(t)
1 1
s
1
t
s2
1 t n 1
(n=1, 2, ...)
sn n 1!
1 1
(n=1, 2,...) t n 1e at
s a
n
n 1!
1 1
(k>0) t k 1e at
s a
k
1 1
s t
1 2
s3/2 t /
1 t x 1
, x0
sx x
1
eat
sa
s
cos at
s a2
2
a
sin at
s a2
2
s
cosh at
s a2
2
a
sinh at
s a2
2
Laplace transform 363
1 ebt e at
, ab
s a s b ba
s bebt ae at
, ab
s a s b ba
1 sin at at cos at
s 2
a 2 2
2a 3
s t sin at
s 2
a 2 2
2a
sin at at cos at
2
s
s 2
a 2 2
2a
s3 1
cos at at sin at
s 2
a 2 2
2
1 e bt e at
sa sb s b a t 3
e a / s cos 2 at
s s t
a/ s n/ 2
e
s n 1
1
a
J n 2 at
1 t
sn n t
e a/ s
H t a
s
e a / s
a
erfc
s 2 t
2
a / s
ae a / 4t
e
2 t3
1 at
erf
s sa a
364 Mathematics for mechanical engineers
1 e at erf at
s s a a
1
sa b
1
eat
t
2
beb t erfc b t
1
J 0 at
s2 a2
a sin at
tan 1
s t
t
a 0t a
1 as F t
2
tanh 2 t a t 2a
as 2 a
F t F t 2a
1 0t a
1 as F t
tanh 1 a t 2a
s 2
F t F t 2a
n/ 2
e a/ s
s n 1
t
a
J n 2 at
a as t
coth sin
a s2 2
2 a
t
sin 0ta
a F t a
a s 2 1 eas
2 2 0 a t 2a
F t F t 2a
t
1 e as F t sin 0t a
a
as 2 s 1 e as
F t F t a
1 e at / 2 3at 3at
3 sin cos e 3at / 2
s a3
3
3a 2 2 2
Laplace transform 365
s e at / 2 3at 3at
3 sin cos e 3at / 2
s a3
3
3a 2 2 2
s2 1 at 3at
e 2e cos
at / 2
s3 a3 3 2
1 e at / 2 3at / 2 3at 3at
2
e 3 sin cos
s a3
3
3a 2 2
s e at / 2 3at 3at
2
3 sin cos e3at / 2
s a3
3
3a 2 2
s2 1 at at / 2 3at
e 2e cos
s3 a3 3 2
1
1
4a 3
sin at cosh at cos at sinh at
s 4a 4
4
s sin at sinh at
s 4a 4
4
2a 2
366 Mathematics for mechanical engineers
EXERCISES
I. Find the Laplace transform:
1. t sin 2t
2. t 2 2t 1
3. t 2 cos 3t
4. t 2 e at
II. Given suitably well behaved functions f, g and h establish the
following properties of the convolution f *g where:
t
f g f g t d
0
1. f g g f
2. f g h g f h
III. Determine the following inverse Laplace transforms:
1. 1
s 1
s 2s 8
2
3s 7
2. 1
s 2s 5
2
e 7 s
3. 1
s 3
3
IV. Solve the following differential equations by using Laplace
transforms:
dx
1. 3 x e 2t , x 0 1
dt
dx
2. 3x sin t , x 1
dt
Laplace transform 367
dx 2 dx
3. 2
3 2 x sin t , x 0 x 0 0
dt dt
2
dx dx
4. 2 4 2 x 5, x 0 x 0 1
dt dt
2
dx
5. 2 2 y 5sin 3t , y 0 y 0 1
dt
ANSWERS
I.
1
1. If F t sin t then f s . Using Dilation Theorem we
s 1
2
have:
1 s 1 1 a
sin at f 2 .
a a a s 2
s a2
1
a
Using Theorem 10.3 we have:
d d 2 4s
t sin 2t f s
ds ds s 2 s 2 2 2
2
II.
1. Put t x , then:
x t d dx , and:
0 xt
t x0
Thus:
t 0
f g f g t d f t x g x dx
0 t
t t
g x g t x dx g g t d g f
0 0
2. We have:
f g h 0 0 f g x d h t x dx
t x
368 Mathematics for mechanical engineers
Domain of x, is sketched as the figure. Changing the order of
integration, we have:
f gh
t t
f g x h t x dxd
0
Let u x we have:
f gh
f g u h t u du d
t t
0 0
f g h
III. All of these problems are computed in the same manner, by
decomposing the expression into partial fractions, using Shift
Theorems, then identifying the simplified expressions with various
standard forms.
1. Using partial fraction decomposition we have:
s 1
1
5
s 2s 8
2
6 s 2 6 s 4
Therefore, taking inverse Laplace transforms gives:
1
s 1 1
1
1
5
s 2s 8
2
6 s 2 6 s 4
1 1 5 1 1 5
1 1 e 2t e 4t
6 s 2 6 s 4 6 6
3s 7
2. The denominator of the rational function does not
s 2s 5 2
factorize. In this case we use completing the square and standard
trigonometric forms as follows:
Laplace transform 369
3s 7 3 s 1 10
s 2 s 5 s 12 4
2
So
3s 7 3 s 1 10
1 1
s 2s 5 s 1 4
2 2
3 s 1 10
1 1
s 1 4 s 1 4
2 2
31
s 1 51 2
s 1 4 s 1 4
2 2
3et cos 2t 5et sin 2t
Again, the First Shift Theorem has been used.
3. The final inverse Laplace transform is different with the above
e 7 s
examples. Notice that the expression contains an exponential
s 3
3
in the numerator. Thus, we can use the Second Shift Theorem since:
e 7 s
e 7 s f s
s 3
3
1
where f s . We first find F t as the inverse Laplace
s 3
3
1 1
transform of f s : 1 t 2 e3t using the First Shift Theorem.
s 3
3
2
1
This says that if F t t 2e3t . We recall the Second Shift Theorem:
2
H t t0 F t t0 e st0 f s
370 Mathematics for mechanical engineers
1 1
Now, put t0 7 , F t t 2e3t , f s to this equation, we
s 3
3
2
1 1
obtain: H t 7 t 7 e3t 7 e 7 s
2
. So
s 3
3
2
e 7 s 1
1 H t 7 t 7 e3t 7
2
s 3
3
2
Where H is Heaviside unit step function.
373
REFERENCES
C.F. Chan Man Fong, D De Kee, P.N. Kaloni. Advanced Mathematics
for Engineering and Science. World Scientific Publishing Co. Pre.
Ltd, 2003.
Christoph Rauscher. Fundamentals of Spectrum Analysis. Rohde &
Schwarz GmbH & Co. KG, 2001.
Eric W. Hansen. Fourier Transforms Principles and Applications.
John Wiley & Sons, Inc, 2014.
Fortney, J.P. A Visual Introduction to Differential Forms and Calculus
on Manifolds. Springer International Publishing, 2019.
Georges Fiche. Gérard Hébuterne, Mathematics for Engineers. John
Wiley & Sons, Inc, 2008.
J. F. James. A Student’s Guide to Fourier Transforms with
Applications in Physics and Engineering. Cambridge University Press,
New York, 2011.
James Stewart. Essential Calculus, Second Edition. Brooks/Cole,
Cengage Learning, 2012.
Jon Rogawski. Calculus, Second Edition. W. H. Freeman and
Company, 2012.
P.P.G. Dyke. An Introduction to Laplace Transforms and Fourier
Series, Second Edition. Springer, 2014.
R. J. Beerends, H. G. ter Morsche, J. C. van den Berg and E. M. van
de Vrie. Fourier and Laplace Transforms. Cambridge University
Press, 2003.
Stephen Lee. An Introduction to Mathematics for Engineers. Taylor
& Francis Group, LLC, 2008.
You might also like
- Applications of Vector Analysis and Complex VariablesNo ratings yetApplications of Vector Analysis and Complex Variables228 pages
- A Modern Text-Book On Statics For Students of Applied Mathematics, Physics and Engineering (Eliezer, C. J) (Z-Library)No ratings yetA Modern Text-Book On Statics For Students of Applied Mathematics, Physics and Engineering (Eliezer, C. J) (Z-Library)316 pages
- Dover Books On Engineering Rutherford Aris Vectors Tensors and The Basic Equations of Fluid Mechanics Dover Publications 1989100% (1)Dover Books On Engineering Rutherford Aris Vectors Tensors and The Basic Equations of Fluid Mechanics Dover Publications 1989314 pages
- Mathematics - Advanced Mathematical Methods in Science and Engineering (Hayek)88% (8)Mathematics - Advanced Mathematical Methods in Science and Engineering (Hayek)737 pages
- (Dover Books On Mathematics) John David Jackson - Mathematics For Quantum Mechanics - An Introductory Survey of Operators, Eigenvalues, and Linear Vector Spaces - Dover Publications (2006)0% (1)(Dover Books On Mathematics) John David Jackson - Mathematics For Quantum Mechanics - An Introductory Survey of Operators, Eigenvalues, and Linear Vector Spaces - Dover Publications (2006)107 pages
- Mathematical Physical Chemistry Practical and Intuitive - Hota, S - 2018 - Springer Singapore, Singapore - Anna's ArchiveNo ratings yetMathematical Physical Chemistry Practical and Intuitive - Hota, S - 2018 - Springer Singapore, Singapore - Anna's Archive704 pages
- (Dover Books On Mathematics) Rutherford Aris, Mathematics - Vectors, Tensors and The BasicNo ratings yet(Dover Books On Mathematics) Rutherford Aris, Mathematics - Vectors, Tensors and The Basic314 pages
- The Structures of Mathematical Physics An Introduction (Steven P. Starkovich) (Z-Library)100% (2)The Structures of Mathematical Physics An Introduction (Steven P. Starkovich) (Z-Library)274 pages
- Electromagnetic Fields For Engineers and Scientists100% (1)Electromagnetic Fields For Engineers and Scientists856 pages
- Mathematical Techniques An Introduction For The Engineering, PH PDF100% (8)Mathematical Techniques An Introduction For The Engineering, PH PDF812 pages
- Application Vector Integral in EngineeringNo ratings yetApplication Vector Integral in Engineering19 pages
- Ebook Vector and Tensor Analysis With Applications - A.borisenko I.tarapov100% (1)Ebook Vector and Tensor Analysis With Applications - A.borisenko I.tarapov265 pages
- Gerd Baumann - Mathematics For Engineers III - Vector Calculus (2011, Oldenbourg Wissenschaftsverlag) (10.1524 - 9783486714470) - Libgen - Li100% (1)Gerd Baumann - Mathematics For Engineers III - Vector Calculus (2011, Oldenbourg Wissenschaftsverlag) (10.1524 - 9783486714470) - Libgen - Li434 pages
- Roald K. Wangsness - Electromagnetic Fields-Wiley (1986)No ratings yetRoald K. Wangsness - Electromagnetic Fields-Wiley (1986)598 pages
- Von Karman, The Engineer Grapples With Nonlinear Problems PDFNo ratings yetVon Karman, The Engineer Grapples With Nonlinear Problems PDF69 pages
- Bishop, Goldberg - Tensor Analysis On Manifolds (Dover 1980) (288s) PDF100% (6)Bishop, Goldberg - Tensor Analysis On Manifolds (Dover 1980) (288s) PDF290 pages
- Applied Theory of Functional Differential EquationsNo ratings yetApplied Theory of Functional Differential Equations245 pages
- EEC 2301 Transmission Lines - Waveguides - Course Book 1No ratings yetEEC 2301 Transmission Lines - Waveguides - Course Book 1424 pages
- Application of Mathematical Tools in Geophysics50% (2)Application of Mathematical Tools in Geophysics28 pages
- (Dover Books On Engineering) Rutherford Aris - Vectors, Tensors and The Basic Equations of Fluid Mechanics (Dover Books On Engineering) - D PDFNo ratings yet(Dover Books On Engineering) Rutherford Aris - Vectors, Tensors and The Basic Equations of Fluid Mechanics (Dover Books On Engineering) - D PDF314 pages
- Advanced Mathematical Methods in Science and Engineering S. I. Hayek PDFNo ratings yetAdvanced Mathematical Methods in Science and Engineering S. I. Hayek PDF737 pages
- Semester I: 15L101 Calculus and Its Applications 3 2 0 4No ratings yetSemester I: 15L101 Calculus and Its Applications 3 2 0 497 pages
- Edward K. Blum, Sergey V. Lototsky - Mathematics of Physics and Engineering - Selected Topics-World Scientific Pub Co ( (2006)No ratings yetEdward K. Blum, Sergey V. Lototsky - Mathematics of Physics and Engineering - Selected Topics-World Scientific Pub Co ( (2006)498 pages
- Mathematics of Physics and Engineering PDF67% (3)Mathematics of Physics and Engineering PDF498 pages
- Differential Calculus: Reviewer InformationNo ratings yetDifferential Calculus: Reviewer Information43 pages
- Introduction To Calculus of Vector FieldsNo ratings yetIntroduction To Calculus of Vector Fields46 pages
- 28 ESL Discussion Topics For Adults That Everyone Has Opinions OnNo ratings yet28 ESL Discussion Topics For Adults That Everyone Has Opinions On17 pages
- Importance of Linear Algebra in Engineering Design MethodologyNo ratings yetImportance of Linear Algebra in Engineering Design Methodology3 pages
- Module 2 Kinetics of A Particle-Force and AccelerationNo ratings yetModule 2 Kinetics of A Particle-Force and Acceleration72 pages
- Practical Astrodynamics Springer Aerospace Technology Alessandro de Iaco Veris Download PDF100% (3)Practical Astrodynamics Springer Aerospace Technology Alessandro de Iaco Veris Download PDF65 pages
- Noor Science Academy Multan: Director: Naveed Noor Cell # 03052440984No ratings yetNoor Science Academy Multan: Director: Naveed Noor Cell # 030524409843 pages
- 3. Gợi ý giải đáp câu hỏi của chủ đề Animals: 14. Have you ever had a pet before?No ratings yet3. Gợi ý giải đáp câu hỏi của chủ đề Animals: 14. Have you ever had a pet before?1 page
- The SUVAT Equations and Their Derivation - StudyWellNo ratings yetThe SUVAT Equations and Their Derivation - StudyWell5 pages
- Welcome To Discord, Bé Thương So Kiu!: 2 Ways To Add A FriendNo ratings yetWelcome To Discord, Bé Thương So Kiu!: 2 Ways To Add A Friend1 page