132 CHAPTER 3 VECTOR ANALYSIS

directions of the three mutually perpendicular unit vectors x̂,

ŷ, and ẑ, which are also called base vectors. The vector A in C A
A C
Fig. 3-2(b) may be decomposed as

A = x̂Ax + ŷAy + ẑAz , (3.3)

B B
where Ax , Ay , and Az are A’s scalar components along the (a) Parallelogram rule (b) Head-to-tail rule
x-, y-, and z axes, respectively. The component Az is equal to
the perpendicular projection of A onto the z axis, and similar Figure 3-3 Vector addition by (a) the parallelogram rule and
definitions apply to Ax and Ay . Application of the Pythagorean (b) the head-to-tail rule.
theorem—first to the right triangle in the x–y plane to express
the hypotenuse Ar in terms of Ax and Ay and then again to the
vertical right triangle with sides Ar and Az and hypotenuse A—
yields the following expression for the magnitude of A: given by
q C = A + B = (x̂Ax + ŷAy + ẑAz ) + (x̂Bx + ŷBy + ẑBz )
A = |A| = + A2x + A2y + A2z . (3.4)
= x̂(Ax + Bx ) + ŷ(Ay + By ) + ẑ(Az + Bz)
Since A is a nonnegative scalar, only the positive root applies. = x̂Cx + ŷCy + ẑCz , (3.7)
From Eq. (3.2), the unit vector â is
with Cx = Ax + Bx , etc.
A x̂Ax + ŷAy + ẑAz
â = = q . (3.5)
A +
A2x + A2y + A2z ◮ Vector addition is commutative:

Occasionally, we use the shorthand notation A = (Ax , Ay , Az ) to C = A + B = B + A. (3.8)

denote a vector with components Ax , Ay , and Az in a Cartesian
coordinate system. Graphically, vector addition can be accomplished by either the
parallelogram or the head-to-tail rule (Fig. 3-3). Vector C is
3-1.1 Equality of Two Vectors the diagonal of the parallelogram with sides A and B. With the
head-to-tail rule, we may either add A to B or B to A. When
Two vectors A and B are equal if they have equal magnitudes A is added to B, it is repositioned so that its tail starts at the
and identical unit vectors. Thus, if tip of B while keeping its length and direction unchanged. The
sum vector C starts at the tail of B and ends at the tip of A.
A = âA = x̂Ax + ŷAy + ẑAz , (3.6a) Subtraction of vector B from vector A is equivalent to the
B = b̂B = x̂Bx + ŷBy + ẑBz , (3.6b) addition of A to negative B. Thus,
D = A − B = A + (−B)
then A = B if and only if A = B and â = b̂, which requires that
Ax = Bx , Ay = By , and Az = Bz . = x̂(Ax − Bx ) + ŷ(Ay − By ) + ẑ(Az − Bz ). (3.9)
Graphically, the same rules used for vector addition are also
applicable to vector subtraction; the only difference is that the
◮ Equality of two vectors does not necessarily imply that
they are identical; in Cartesian coordinates, two displaced arrowhead of (−B) is drawn on the opposite end of the line
parallel vectors of equal magnitude and pointing in the segment representing the vector B (i.e., the tail and head are
same direction are equal, but they are identical only if they interchanged).
lie on top of one another. ◭
3-1.3 Position and Distance Vectors
The position vector of a point P in space is the vector from the
3-1.2 Vector Addition and Subtraction origin to P. Assuming points P1 and P2 are at (x1 , y1 , z1 ) and
The sum of two vectors A and B is a vector (x2 , y2 , z2 ) in Fig. 3-4, their position vectors are
−→
C = x̂Cx + ŷCy + ẑCz , R1 = OP1 = x̂x1 + ŷy1 + ẑz1 , (3.10a)
3-1 BASIC LAWS OF VECTOR ALGEBRA 133

Scalar or Dot Product

z
The scalar (or dot) product of two co-anchored vectors A
and B, denoted A · B and pronounced “A dot B,” is defined
z2 geometrically as the product of the magnitude of A and the
P1 = (x1, y1, z1)
scalar component of B along A, or vice versa. Thus,
z1 R12 P2 = (x2, y2, z2)
A · B = AB cos θAB , (3.14)

R1 R2 where θAB is the angle between A and B (Fig. 3-5) measured

from the tail of A to the tail of B. Angle θAB is assumed to be
y1 y2
O y in the range 0 ≤ θAB ≤ 180◦. The scalar product of A and B
yields a scalar whose magnitude is less than or equal to the
x1 products of their magnitudes (equality holds when θAB = 0)
x2 and whose sign is positive if 0 < θAB < 90◦ and negative if
90◦ < θAB < 180◦. When θAB = 90◦ , A and B are orthogonal,
x and their dot product is zero. The quantity A cos θAB is the
−−→ scalar component of A along B. Similarly, B cos θBA is the
Figure 3-4 Distance vector R12 = P1 P2 = R2 − R1 , where R1 scalar component of B along A.
and R2 are the position vectors of points P1 and P2 , respectively.

The dot product obeys both the commutative and

−→ distributive properties of multiplication:
R2 = OP2 = x̂x2 + ŷy2 + ẑz2 , (3.10b)
where point O is the origin. A · B = B · A, (3.15a)
The distance vector from P1 to P2 is defined as (commutative property)
−−→
R12 = P1 P2 = R2 − R1 A ·(B + C) = A · B + A · C. (3.15b)
= x̂(x2 − x1 ) + ŷ(y2 − y1 ) + ẑ(z2 − z1 ), (3.11)
(distributive property)
and the distance d between P1 and P2 equals the magnitude
of R12 :
The commutative property follows from Eq. (3.14) and the fact
d = |R12 | = [(x2 − x1 )2 + (y2 − y1 )2 + (z2 − z1 )2 ]1/2 . (3.12) that θAB = θBA . The distributive property expresses the fact that
Note that the first and second subscripts of R12 denote the the scalar component of the sum of two vectors along a third
locations of its tail and head, respectively (Fig. 3-4). one equals the sum of their respective scalar components.
The dot product of a vector with itself gives

3-1.4 Vector Multiplication A · A = |A|2 = A2 , (3.16)

There exist three types of products in vector calculus: the

simple product, the scalar (or dot) product, and the vector (or
A B
cross) product.
θBA θAB
Simple Product θBA
θAB B A
The multiplication of a vector by a scalar is called a simple
(a) (b)
product. The product of the vector A = âA by a scalar k results
in a vector B with magnitude B = kA and direction the same
as A. That is, b̂ = â. In Cartesian coordinates, Figure 3-5 The angle θAB is the angle between A and B,
measured from A to B between vector tails. The dot product
B = kA = âkA = x̂(kAx ) + ŷ(kAy ) + ẑ(kAz ) is positive if 0 ≤ θAB < 90◦ , as in (a), and it is negative if
90◦ < θAB ≤ 180◦ , as in (b).
= x̂ Bx + ŷ By + ẑBz . (3.13)
134 CHAPTER 3 VECTOR ANALYSIS

which implies that

z
√
+
A = |A| = A · A . (3.17) A × B = nˆ AB sin θAB

Also, θAB can be determined from

nˆ B
A·B
 
θAB = cos−1 √ +
√ . (3.18) θAB
A·A + B·B y
Since the base vectors x̂, ŷ, and ẑ are each orthogonal to the
other two, it follows that x A
(a) Cross product
x̂ · x̂ = ŷ · ŷ = ẑ · ẑ = 1, (3.19a)
A×B
x̂ · ŷ = ŷ · ẑ = ẑ · x̂ = 0. (3.19b)

If A = (Ax , Ay , Az ) and B = (Bx , By , Bz ), then B

A · B = (x̂Ax + ŷAy + ẑAz ) ·(x̂Bx + ŷBy + ẑBz ). (3.20)

Use of Eqs. (3.19a) and (3.19b) in Eq. (3.20) leads to A

A · B = Ax Bx + Ay By + Az Bz . (3.21) (b) Right-hand rule

Figure 3-6 Cross product A × B points in the direction n̂,

Vector or Cross Product which is perpendicular to the plane containing A and B and
defined by the right-hand rule.
The vector (or cross) product of two vectors A and B, denoted
× B and pronounced “A cross B,” yields a vector defined as
A×

by A and (B + C) equals the sum of those formed by (A and B)

× B = n̂ AB sin θAB ,
A× (3.22) and (A and C):

where n̂ is a unit vector normal to the plane containing A × (B + C) = A×

A× × B + A×
× C, (3.23b)
and B (Fig. 3-6(a)). The magnitude of the cross product,
(distributive)
AB| sin θAB |, equals the area of the parallelogram defined by the
two vectors. The direction of n̂ is governed by the right-hand
rule (Fig. 3-6(b)): n̂ points in the direction of the right thumb The cross product of a vector with itself vanishes. That is,
when the fingers rotate from A to B through the angle θAB .
× A = 0.
A× (3.24)
Note that, since n̂ is perpendicular to the plane containing A
and B, A× × B is perpendicular to both vectors A and B. From the definition of the cross product given by Eq. (3.22),
it is easy to verify that the base vectors x̂, ŷ, and ẑ of
the Cartesian coordinate system obey the right-hand cyclic
The cross product is anticommutative and distribu- relations:
tive:

× B = −B×
A× ×A (anticommutative). (3.23a) × ŷ = ẑ,
x̂× × ẑ = x̂,
ŷ× × x̂ = ŷ.
ẑ× (3.25)

Note the cyclic order (xyzxyz . . .). Also,

The anticommutative property follows from the application of
the right-hand rule to determine n̂. The distributive property
× ẑ = 0.
x̂ × x̂ = ŷ × ŷ = ẑ× (3.26)
follows from the fact that the area of the parallelogram formed
3-1 BASIC LAWS OF VECTOR ALGEBRA 135

p √
If A = (Ax , Ay , Az ) and B = (Bx , By , Bz ), then use of Eqs. (3.25) A = |A| = 22 + 32 + 32 = 22 ,
and (3.26) leads to A √
â = = (x̂2 + ŷ3 + ẑ3)/ 22 .
× B = (x̂Ax + ŷAy + ẑAz )×
A× × (x̂Bx + ŷBy + ẑBz ) A
= x̂(Ay Bz − AzBy ) + ŷ(Az Bx − Ax Bz ) (b) The angle β between A and the y axis is obtained from
+ ẑ(Ax By − Ay Bx ). (3.27) A · ŷ = |A||ŷ| cos β = A cos β ,
The cyclical form of the result given by Eq. (3.27) allows us to or
A · ŷ
   
express the cross product in the form of a determinant: −1 −1 3
β = cos = cos √ = 50.2◦.
A 22
x̂ ŷ ẑ (c)
×B =
A× Ax Ay Az . (3.28)
Bx By Bz B = x̂(1 − 2) + ŷ(−2 − 3) + ẑ(2 − 3) = −x̂ − ŷ5 − ẑ.

(d)
A·B
   
−1 −1 (−2 − 15 − 3)
Example 3-1: Vectors and Angles θAB = cos = cos √ √ = 145.1◦.
|A||B| 22 27
(e) The perpendicular distance between the origin and vector B
In Cartesian coordinates, vector A points from the origin to −→
point P1 = (2, 3, 3), and vector B is directed from P1 to point is the distance | OP3 | shown in Fig. 3-7. From right triangle
P2 = (1, −2, 2). Find: OP1 P3 ,
(a) vector A, its magnitude A, and unit vector â, −→
| OP3 | = |A| sin(180◦ − θAB )
(b) the angle between A and the y axis, √
(c) vector B, = 22 sin(180◦ − 145.1◦) = 2.68.
(d) the angle θAB between A and B, and
(e) perpendicular distance from the origin to vector B.
Solution: (a) Vector A is given by the position vector of Example 3-2: Cross Product
P1 = (2, 3, 3) (Fig. 3-7). Thus,
A = x̂2 + ŷ3 + ẑ3, Given vectors A = x̂2 − ŷ + ẑ3 and B = ŷ2 − ẑ3, compute
× B) · A.
× B, (b) ŷ × B, and (c) (ŷ×
(a) A×
Solution: (a) Application of Eq. (3.28) gives
z
θAB x̂ ŷ ẑ
3 ×B =
A× 2 −1 3
P1 = (2, 3, 3) 0 2 −3
P2 = (1, –2, 2) B 2 = x̂((−1) × (−3) − 3 × 2) − ŷ(2 × (−3) − 3 × 0)
P3
A + ẑ(2 × 2 − (−1 × 0))
1
β = −x̂3 + ŷ6 + ẑ4.
–2 O 3
y × B = ŷ×
(b) ŷ× × (ŷ2 − ẑ3) = −x̂3.
1 (c) (ŷ × B) · A = −x̂3 ·(x̂2 − ŷ + ẑ3) = −6.

2
Exercise 3-1: Find the distance vector between
x P1 = (1, 2, 3) and P2 = (−1, −2, 3) in Cartesian coor-
dinates.
Figure 3-7 Geometry of Example 3-1. −−→
Answer: P1 P2 = −x̂2 − ŷ4. (See EM .)
136 CHAPTER 3 VECTOR ANALYSIS

result is a scalar. A scalar triple product obeys the cyclic order:

Exercise 3-2: Find the angle θAB between vectors A and B
of Example 3-1 from the cross product between them.
Answer: θAB = 145.1◦. (See EM
.) A ·(B × C) = B ·(C×
× A) = C ·(A×
× B). (3.29)

Exercise 3-3: Find the angle between vector B of Exam- The equalities hold as long as the cyclic order (ABCABC . . .) is
ple 3-1 and the z axis. preserved. The scalar triple product of vectors A = (Ax , Ay , Az ),
B = (Bx , By , Bz ), and C = (Cx ,Cy ,Cz ) can be expressed in the
Answer: 101.1◦. (See EM
.) form of a 3 × 3 determinant:

Exercise 3-4: Vectors A and B lie in the y–z plane and Ax Ay Az

both have the same magnitude of 2 (Fig. E3.4). Determine A ·(B×
× C) = Bx By Bz . (3.30)
(a) A · B and (b) A×
× B. Cx Cy Cz

z The validity of Eqs. (3.29) and (3.30) can be verified by

B expanding A, B, and C in component form and carrying out
the multiplications.

2 30◦ Vector Triple Product

A The vector triple product involves the cross product of a vector
y with the cross product of two others, such as
2
× (B × C).
A× (3.31)
x
Since each cross product yields a vector, the result of a vector
Figure E3.4
triple product is also a vector. The vector triple product does
not obey the associative law. That is,
Answer: (a) A · B = −2; (b) A×
× B = x̂ 3.46. (See EM
.)
× (B × C) 6= (A×
A× × B)×
× C, (3.32)
Exercise 3-5: If A · B = A · C, does it follow that B = C?
which means that it is important to specify which cross multi-
Answer: No. (See EM
.) plication is to be performed first. By expanding the vectors A,
B, and C in component form, it can be shown that

3-1.5 Scalar and Vector Triple Products × C) = B(A · C) − C(A · B),

× (B×
A× (3.33)
When three vectors are multiplied, not all combinations of dot
and cross products are meaningful. For example, the product
which is known as the “bac-cab” rule.
× (B · C)
A×

does not make sense because B · C is a scalar, and the cross Example 3-3: Vector Triple Product
product of the vector A with a scalar is not defined under the
rules of vector algebra. Other than the product of the form
A(B · C), the only two meaningful products of three vectors Given A = x̂ − ŷ + ẑ2, B = ŷ + ẑ, and C = −x̂2 + ẑ3, find
are the scalar triple product and the vector triple product. (A×× B)×
× C and compare it with A× × (B × C).
Solution:
Scalar Triple Product
x̂ ŷ ẑ
The dot product of a vector with the cross product of two other ×B =
A× 1 −1 2 = −x̂3 − ŷ + ẑ
vectors is called a scalar triple product, so named because the 0 1 1
3-2 ORTHOGONAL COORDINATE SYSTEMS 137

and Why do we need more than one coordinate system? Whereas

a point in space has the same location and an object has the
x̂ ŷ ẑ same shape regardless of which coordinate system is used
× B)×
(A× ×C = −3 −1 1 = −x̂3 + ŷ7 − ẑ2. to describe them, the solution of a practical problem can be
−2 0 3 greatly facilitated by the choice of a coordinate system that
best fits the geometry under consideration. The following sub-
A similar procedure gives A × (B × C) = x̂2 + ŷ4 + ẑ. The sections examine the properties of each of the aforementioned
fact that the results of two vector triple products are different orthogonal systems, and Section 3-3 describes how a point or
demonstrates the inequality stated in Eq. (3.32). vector may be transformed from one system to another.

Concept Question 3-1: When are two vectors equal

and when are they identical?
3-2.1 Cartesian Coordinates
Concept Question 3-2: When is the position vector of a The Cartesian coordinate system was introduced in Section 3-1
point identical to the distance vector between two points? to illustrate the laws of vector algebra. Instead of repeating
these laws for the Cartesian system, we summarize them in
Table 3-1. Differential calculus involves the use of differen-
Concept Question 3-3: If A · B = 0, what is θAB ?
tial lengths, areas, and volumes. In Cartesian coordinates, a
differential length vector (Fig. 3-8) is expressed as
Concept Question 3-4: × B = 0, what is θAB ?
If A×
dl = x̂ dlx + ŷ dly + ẑ dlz = x̂ dx + ŷ dy + ẑ dz, (3.34)
Concept Question 3-5: Is A(B · C) a vector triple prod-
uct?
where dlx = dx is a differential length along x̂, and similar
interpretations apply to dly = dy and dlz = dz.
Concept Question 3-6: If A · B = A · C, does it follow A differential area vector ds is a vector with magnitude ds
that B = C? equal to the product of two differential lengths (such as dly
and dlz ) and direction specified by a unit vector along the third

3-2 Orthogonal Coordinate Systems

A three-dimensional coordinate system allows us to uniquely
specify locations of points in space and the magnitudes and z dsz = zˆ dx dy
directions of vectors. Coordinate systems may be orthogonal
or nonorthogonal. dy
dx

dsy = yˆ dx dz
◮ An orthogonal coordinate system is one in which
dz
coordinates are measured along locally mutually perpen-
dz
dicular axes. ◭
dl dv = dx dy dz
dsx = xˆ dy dz
Nonorthogonal systems are very specialized and seldom used
in solving practical problems. Many orthogonal coordinate dy
y
systems have been devised, but the most commonly used are
dx
• the Cartesian (also called rectangular), x

• the cylindrical, and Figure 3-8 Differential length, area, and volume in Cartesian
coordinates.
• the spherical coordinate system.
138 CHAPTER 3 VECTOR ANALYSIS

Table 3-1 Summary of vector relations.

Cartesian Cylindrical Spherical
Coordinates Coordinates Coordinates
Coordinate variables x, y, z r, φ , z R, θ , φ
Vector representation A = x̂Ax + ŷAy + ẑAz r̂Ar + φ̂φAφ + ẑAz R̂AR + θ̂θAθ + φ̂φAφ
q q q
Magnitude of A |A| = +
A2x + A2y + A2z +
A2r + Aφ2 + A2z +
A2R + Aθ2 + Aφ2

Position vector
−→
OP = x̂x1 + ŷy1 + ẑz1 , r̂r1 + ẑz1 , R̂R1 ,
1
for P(x1 , y1 , z1 ) for P(r1 , φ1 , z1 ) for P(R1 , θ1 , φ1 )

Base vector properties x̂ · x̂ = ŷ · ŷ = ẑ · ẑ = 1 r̂ · r̂ = φ̂φ · φ̂

φ = ẑ · ẑ = 1 R̂ · R̂ = θ̂ θ · θ̂
θ = φ̂φ · φ̂φ = 1
x̂ · ŷ = ŷ · ẑ = ẑ · x̂ = 0 r̂ · φ̂φ = φ̂φ · ẑ = ẑ · r̂ = 0 R̂ · θ̂θ = θ̂
θ · φ̂
φ = φ̂φ · R̂ = 0
x̂ × ŷ = ẑ r̂ × φ̂φ = ẑ R̂× × θ̂θ = φ̂φ
ŷ × ẑ = x̂ φ × ẑ = r̂
φ̂ θ × φ̂
θ̂ φ = R̂
ẑ × x̂ = ŷ ẑ × r̂ = φ̂ φ φ × R̂ = θ̂θ
φ̂
Dot product A·B = Ax Bx + Ay By + Az Bz Ar Br + Aφ Bφ + Az Bz AR BR + Aθ Bθ + Aφ Bφ
x̂ ŷ ẑ r̂ φ̂φ ẑ R̂ θ̂θ φ̂φ
Cross product ×B =
A× Ax Ay Az Ar Aφ Az AR Aθ Aφ
Bx By Bz Br Bφ Bz BR Bθ Bφ

Differential length dl = x̂ dx + ŷ dy + ẑ dz r̂ dr + φ̂φr d φ + ẑ dz R̂ dR + θ̂θR d θ + φ̂φR sin θ d φ

Differential surface areas dsx = x̂ dy dz dsr = r̂r d φ dz dsR = R̂R2 sin θ d θ d φ
dsy = ŷ dx dz φ dr dz
dsφ = φ̂ dsθ = θ̂θR sin θ dR d φ
dsz = ẑ dx dy dsz = ẑr dr d φ dsφ = φ̂φR dR d θ
Differential volume d υ = dx dy dz r dr d φ dz R2 sin θ dR d θ d φ

direction (such as x̂). Thus, for a differential area vector in the φ , and z (Fig. 3-9). The coordinate r is the radial distance in
y–z plane, the x–y the azimuth angle measured from the positive x axis,
and z is as previously defined in the Cartesian coordinate
dsx = x̂ dly dlz = x̂ dy dz (y–z plane), (3.35a)
system. Their ranges are 0 ≤ r < ∞, 0 ≤ φ < 2π , and
with the subscript on ds denoting its direction. Similarly, −∞ < z < ∞. Point P(r1 , φ1 , z1 ) in Fig. 3-9 is located at the
intersection of three surfaces. These are the cylindrical surface
dsy = ŷ dx dz (x–z plane), (3.35b)
defined by r = r1 , the vertical half-plane defined by φ = φ1
dsz = ẑ dx dy (x–y plane). (3.35c) (which extends outwardly from the z axis), and the horizontal
plane defined by z = z1 .
A differential volume equals the product of all three differen-
tial lengths:
d υ = dx dy dz. (3.36)
◮ The mutually perpendicular base vectors are r̂, φ̂
φ , and ẑ
with r̂ pointing away from the origin along r, φ̂φ pointing
3-2.2 Cylindrical Coordinates in a direction tangential to the cylindrical surface, and ẑ
The cylindrical coordinate system is useful for solving pointing along the vertical. Unlike the Cartesian system,
problems involving structures with cylindrical symmetry, such where base vectors x̂, ŷ, and ẑ are independent of the
as calculating the capacitance per unit length of a coaxial location of P, both r̂ and φ̂φ are functions of φ in the
transmission line. In the cylindrical coordinate system, the cylindrical system. ◭
location of a point in space is defined by three variables: r,
3-2 ORTHOGONAL COORDINATE SYSTEMS 139

z = z1 plane

z1 P = (r1, φ1, z1)

R1

r = r1 cylinder
O y
φ1 r1 zˆ
ˆ
φ φ = φ1 plane
rˆ
x

Figure 3-9 Point P(r1 , φ1 , z1 ) in cylindrical coordinates; r1 is the radial distance from the origin in the x–y plane, φ1 is the angle in the
x–y plane measured from the x axis toward the y axis, and z1 is the vertical distance from the x–y plane.

The base unit vectors obey the following right-hand cyclic position vector of point P(r1 , φ1 , z1 ), it is necessary to specify
relations: that r̂ is at φ1 .
Figure 3-10 shows a differential volume element in cylin-
× φ̂φ = ẑ,
r̂× φ̂φ × ẑ = r̂, φ,
× r̂ = φ̂
ẑ× (3.37) drical coordinates. The differential lengths along r̂, φ̂φ, and ẑ
are

and like all unit vectors, r̂ · r̂ = φ̂φ · φ̂φ = ẑ · ẑ = 1, and dlr = dr, dlφ = r d φ , dlz = dz. (3.41)
× r̂ = φ̂φ × φ̂φ = ẑ×
r̂× × ẑ = 0.
In cylindrical coordinates, a vector is expressed as Note that the differential length along φ̂φ is r d φ , not just d φ .
The differential length dl in cylindrical coordinates is given by
A = â|A| = r̂Ar + φ̂φAφ + ẑAz , (3.38)
dl = r̂ dlr + φ̂φ dlφ + ẑ dlz = r̂ dr + φ̂φr d φ + ẑ dz. (3.42)
where Ar , Aφ , and Az are the components of A along the r̂, φ̂φ,
and ẑ directions. The magnitude of A is obtained by applying As was stated previously for the Cartesian coordinate system,
Eq. (3.17), which gives the product of any pair of differential lengths is equal to the
√
+
q magnitude of a vector differential surface area with a surface
|A| = A · A = + A2r + A2φ + A2z . (3.39) normal pointing along the direction of the third coordinate.
Thus,
−→
The position vector OP shown in Fig. 3-9 has components
along r and z only. Thus, dsr = r̂ dlφ dlz = r̂r d φ dz (φ –z cylindrical surface),
−→ (3.43a)
R1 = OP = r̂r1 + ẑz1 . (3.40)
dsφ = φ̂φ dlr dlz = φ̂
φ dr dz (r–z plane), (3.43b)
The dependence of R1 on φ1 is implicit through the depen-
dence of r̂ on φ1 . Hence, when using Eq. (3.40) to denote the dsz = ẑ dlr dlφ = ẑr dr d φ (r–φ plane). (3.43c)
140 CHAPTER 3 VECTOR ANALYSIS

z z

P1 = (0, 0, h)
dz dsz = zˆ r dr dφ
aˆ
r dφ dr
dsφ = ϕˆ dr dz
h
dz dv = r dr dφ dz

dsr = rˆ r dφ dz A
O y
φ0 r0
O y P2 = (r0, φ0, 0)
φ
r x

x Figure 3-11 Geometry of Example 3-4.

dr r dφ

Figure 3-10 Differential areas and volume in cylindrical

coordinates. coordinate system, which is not true. The ambiguity can be
resolved by specifying that A passes through a point whose
φ = φ0 .
The differential volume is the product of the three differential
lengths,
d υ = dlr dlφ dlz = r dr d φ dz. (3.44) Example 3-5: Cylindrical Area
These properties of the cylindrical coordinate system are
summarized in Table 3-1. Find the area of a cylindrical surface described by r = 5,
30◦ ≤ φ ≤ 60◦ , and 0 ≤ z ≤ 3 (Fig. 3-12).
Example 3-4: Distance Vector in
Cylindrical Coordinates
z
Find an expression for the unit vector of vector A shown in
Fig. 3-11 in cylindrical coordinates. z=3
r=5
Solution: In triangle OP1 P2 ,
−→ −→
OP2 = OP1 +A.

Hence,
−→ −→
A = OP2 − OP1 = r̂r0 − ẑh,
y
and 60°
A r̂r0 − ẑh
â = =q . 30°
|A| r02 + h2
x
We note that the expression for A is independent of φ0 . This
implies that all vectors from point P1 to any point on the circle Figure 3-12 Cylindrical surface of Example 3-5.
defined by r = r0 in the x–y plane are equal in the cylindrical
3-2 ORTHOGONAL COORDINATE SYSTEMS 141

Module 3.1 Points and Vectors Examine the relationships between Cartesian coordinates (x, y) and cylindrical
coordinates (r, φ ) for points and vectors.

Solution: The prescribed surface is shown in Fig. 3-12. Use distance from the origin to the point, describes a sphere of
of Eq. (3.43a) for a surface element with constant r gives radius R centered at the origin. The zenith angle θ is measured
Z 60◦ Z 3 from the positive z axis and it describes a conical surface with
π /3 3 5π its apex at the origin, and the azimuth angle φ is the same
S=r dφ dz = 5φ z = .
φ =30◦ z=0 π /6 0 2 as in cylindrical coordinates. The ranges of R, θ , and φ are
0 ≤ R < ∞, 0 ≤ θ ≤ π , and 0 ≤ φ < 2π . The base vectors R̂,
Note that φ had to be converted to radians before evaluating θ̂θ, and φ̂φ obey the right-hand cyclic relations:
the integration limits.
θ = φ̂φ,
R̂ × θ̂ θ̂θ × φ̂
φ = R̂, φ̂φ × R̂ = θ̂θ. (3.45)
Exercise 3-6: A circular cylinder of radius r = 5 cm is
concentric with the z axis and extends between z = −3 cm
A vector with components AR , Aθ , and Aφ is written as
and z = 3 cm. Use Eq. (3.44) to find the cylinder’s volume.
Answer: 471.2 cm3 . (See EM
.) A = â|A| = R̂AR + θ̂θAθ + φ̂φAφ , (3.46)
and its magnitude is
√
+
q
3-2.3 Spherical Coordinates |A| = A · A = + A2R + A2θ + A2φ . (3.47)
In the spherical coordinate system, the location of a point The position vector of point P(R1 , θ1 , φ1 ) is simply
in space is uniquely specified by the variables R, θ , and φ −→
(Fig. 3-13). The range coordinate R, which measures the R1 = OP = R̂R1 , (3.48)
142 CHAPTER 3 VECTOR ANALYSIS

As shown in Fig. 3-14, the differential lengths along R̂, θ̂θ,

z and φ̂φ are

dlR = dR, dlθ = R d θ , dlφ = R sin θ d φ . (3.49)

Hence, the expressions for the vector differential length dl, the
ˆ
R vector differential surface ds, and the differential volume d υ
φ̂ are
P = (R1, θ1, φ1)
R1 θˆ dl = R̂ dlR + θ̂θ dlθ + φ̂φ dlφ
θ = θ1
conical θ1
surface = R̂ dR + θ̂θR d θ + φ̂φR sin θ d φ , (3.50a)
y dsR = R̂ dlθ dlφ = R̂R2 sin θ d θ d φ (3.50b)
φ1 (θ –φ spherical surface),
dsθ = θ̂θ dlR dlφ = θ̂θR sin θ dR d φ (3.50c)
ˆ
φ
(R–φ conical surface),
x
dsφ = φ̂φ dlR dlθ = φ̂φR dR d θ (R–θ plane), (3.50d)
Figure 3-13 Point P(R1 , θ1 , φ1 ) in spherical coordinates.
d υ = dlR dlθ dlφ = R2 sin θ dR d θ d φ . (3.50e)

These relations are summarized in Table 3-1.

z
Example 3-6: Surface Area in Spherical
R sin θ dφ Coordinates

dν = R2 sin θ dR dθ dφ
The spherical strip shown in Fig. 3-15 is a section of a sphere
of radius 3 cm. Find the area of the strip.
dR R dθ
R
z
θ
dθ
y 60 o
30 o
φ
dφ

y
x 3c
m
Figure 3-14 Differential volume in spherical coordinates.

while keeping in mind that R̂ is implicitly dependent on θ1 Figure 3-15 Spherical strip of Example 3-6.
and φ1 .
3-3 TRANSFORMATIONS BETWEEN COORDINATE SYSTEMS 143

Solution: Use of Eq. (3.50b) for the area of an elemental section, we shall establish the relations between the variables
spherical area with constant radius R gives (x, y, z) of the Cartesian system, (r, φ , z) of the cylindrical sys-
Z 60◦ Z 2π
tem, and (R, θ , φ ) of the spherical system. These relations will
then be used to transform expressions for vectors expressed in
S = R2 sin θ d θ dφ
θ =30◦ φ =0 any one of the three systems into expressions applicable in the
60◦ 2π other two.
= 9(− cos θ ) φ (cm2 )
30◦ 0
3-3.1 Cartesian to Cylindrical Transformations
= 18π (cos30◦ − cos60◦) = 20.7 cm2 .
Point P in Fig. 3-16 has Cartesian coordinates (x, y, z) and
cylindrical coordinates (r, φ , z). Both systems share the coor-
dinate z, and the relations between the other two pairs of
Example 3-7: Charge in a Sphere coordinates can be obtained from the geometry in Fig. 3-16.
They are
y
A sphere of radius 2 cm contains a volume charge density ρv
p
r = + x2 + y2 , φ = tan−1 , (3.51)
given by x
ρv = 4 cos2 θ (C/m3 ). and the inverse relations are
Find the total charge Q contained in the sphere. x = r cos φ , y = r sin φ . (3.52)
Solution: Next, with the help of Fig. 3-17, which shows the directions of
the unit vectors x̂, ŷ, r̂, and φ̂φ in the x–y plane, we obtain the
Z
Q= ρv d υ relations:
υ
Z 2π Z π Z 2×10−2
r̂ · x̂ = cos φ , r̂ · ŷ = sin φ , (3.53a)
= (4 cos2 θ )R2 sin θ dR d θ d φ
φ =0 θ =0 R=0 φ · x̂ = − sin φ ,
φ̂ φ̂φ · ŷ = cos φ . (3.53b)
Z 2π Z π  3  2×10−2
R To express r̂ in terms of x̂ and ŷ, we write r̂ as
=4 sin θ cos2 θ d θ d φ
0 0 3 0 r̂ = x̂a + ŷb, (3.54)
π
cos3 θ
Z 2π  
32
= × 10−6 − dφ where a and b are unknown transformation coefficients. The
3 0 3 0 dot product r̂ · x̂ gives
64 2π Z
= × 10−6 dφ r̂ · x̂ = x̂ · x̂a + ŷ · x̂b = a. (3.55)
9 0
128π
= × 10−6 = 44.68 (µ C).
9 z
Note that the limits on R were converted to meters prior to
evaluating the integral on R.
P(x, y, z)
3-3 Transformations between Coordinate z
Systems y
3

φ r
2

x = r cos φ
1

The position of a given point in space of course does not 123

depend on the choice of coordinate system. That is, its location y = r sin φ
is the same irrespective of which specific coordinate system is x
used to represent it. The same is true for vectors. Nevertheless,
certain coordinate systems may be more useful than others in Figure 3-16 Interrelationships between Cartesian coordinates
solving a given problem, so it is essential that we have the tools (x, y, z) and cylindrical coordinates (r, φ , z).
to “translate” the problem from one system to another. In this
144 CHAPTER 3 VECTOR ANALYSIS

and, conversely,
y
Ax = Ar cos φ − Aφ sin φ , (3.59a)
ϕ ϕ̂ Ay = Ar sin φ + Aφ cos φ . (3.59b)
r
yˆ
ϕ rˆ The transformation relations given in this and the following
−ϕˆ two subsections are summarized in Table 3-2.
xˆ
x
Example 3-8: Cartesian to Cylindrical
Figure 3-17 Interrelationships between base vectors (x̂, ŷ) and Transformations
(r̂, φ̂φ).
Given point P1 = (3, −4, 3) and vector A = x̂2 − ŷ3 + ẑ4
defined in Cartesian coordinates, express P1 and A in cylin-
Comparison of Eq. (3.55) with Eq. (3.53a) yields a = cos φ . drical coordinates and evaluate A at P1 .
Similarly, application of the dot product r̂ · ŷ to Eq. (3.54) gives Solution: For point P1 , x = 3, y = −4, and z = 3. Using
b = sin φ . Hence, Eq. (3.51), we have
p y
r = + x2 + y2 = 5, φ = tan−1 = −53.1◦ = 306.9◦,
r̂ = x̂ cos φ + ŷsin φ . (3.56a) x
and z remains unchanged. Hence, P1 = P1 (5, 306.9◦, 3) in
Repetition of the procedure for φ̂φ leads to cylindrical coordinates.
The cylindrical components of vector A = r̂Ar + φ̂φAφ + ẑAz
can be determined by applying Eqs. (3.58a) and (3.58b):
φ = −x̂ sin φ + ŷ cos φ .
φ̂ (3.56b)
Ar = Ax cos φ + Ay sin φ = 2 cos φ − 3 sin φ ,
Aφ = −Ax sin φ + Ay cos φ = −2 sin φ − 3 cos φ ,
The third base vector ẑ is the same in both coordinate systems.
By solving Eqs. (3.56a) and (3.56b) simultaneously for x̂ Az = 4.
and ŷ, we obtain the following inverse relations:
Hence,

A = r̂(2 cos φ − 3 sin φ ) − φ̂φ(2 sin φ + 3 cos φ ) + ẑ4.

x̂ = r̂ cos φ − φ̂φ sin φ , (3.57a)
ŷ = r̂ sin φ + φ̂φ cos φ . (3.57b) At point P, φ = 306.9◦, which gives

A = r̂ 3.60 − φ̂φ 0.20 + ẑ 4.

The relations given by Eqs. (3.56a) to (3.57b) are not only
φ ), and
useful for transforming the base vectors (x̂, ŷ) into (r̂, φ̂ 3-3.2 Cartesian to Spherical Transformations
vice versa, they also can be used to transform the components
of a vector expressed in either coordinate system into its From Fig. 3-18, we obtain the following relations between
corresponding components expressed in the other system. For the Cartesian coordinates (x, y, z) and the spherical coordinates
example, a vector A = x̂Ax + ŷAy + ẑAz in Cartesian coordi- (R, θ , φ ):
nates can be described by A = r̂Ar + φ̂φAφ + ẑAz in cylindrical p
coordinates by applying Eqs. (3.56a) and (3.56b). That is, R = + x2 + y2 + z2 , (3.60a)
"p #
+ 2
x + y2
θ = tan−1 , (3.60b)
Ar = Ax cos φ + Ay sin φ , (3.58a) z
y
Aφ = −Ax sin φ + Ay cos φ , (3.58b) φ = tan−1 . (3.60c)
x
3-3 TRANSFORMATIONS BETWEEN COORDINATE SYSTEMS 145

Table 3-2 Coordinate transformation relations.

Transformation Coordinate Variables Unit Vectors Vector Components
p
Cartesian to r = + x2 + y2 r̂ = x̂ cos φ + ŷ sin φ Ar = Ax cos φ + Ay sin φ
cylindrical φ = tan−1 (y/x) φ = −x̂ sin φ + ŷ cos φ
φ̂ Aφ = −Ax sin φ + Ay cos φ
z=z ẑ = ẑ Az = Az
Cylindrical to x = r cos φ x̂ = r̂ cos φ − φ̂φ sin φ Ax = Ar cos φ − Aφ sin φ
Cartesian y = r sin φ ŷ = r̂ sin φ + φ̂φ cos φ Ay = Ar sin φ + Aφ cos φ
z=z ẑ = ẑ Az = Az
p
Cartesian to R = + x2 + y2 + z2 R̂ = x̂ sin θ cos φ AR = Ax sin θ cos φ
spherical p + ŷ sin θ sin φ + ẑ cos θ + Ay sin θ sin φ + Az cos θ
θ = tan−1 [ +
x2 + y2 /z] θ = x̂ cos θ cos φ
θ̂ Aθ = Ax cos θ cos φ
+ ŷ cos θ sin φ − ẑ sin θ + Ay cos θ sin φ − Az sin θ
φ = tan−1 (y/x) φ = −x̂ sin φ + ŷ cos φ
φ̂ Aφ = −Ax sin φ + Ay cos φ
Spherical to x = R sin θ cos φ x̂ = R̂ sin θ cos φ Ax = AR sin θ cos φ
Cartesian + θ̂θ cos θ cos φ − φ̂φ sin φ + Aθ cos θ cos φ − Aφ sin φ
y = R sin θ sin φ ŷ = R̂ sin θ sin φ Ay = AR sin θ sin φ
+ θ̂θ cos θ sin φ + φ̂φ cos φ + Aθ cos θ sin φ + Aφ cos φ
z = R cos θ ẑ = R̂ cos θ − θ̂θ sin θ Az = AR cos θ − Aθ sin θ
√
+
Cylindrical to R = r2 + z2 R̂ = r̂ sin θ + ẑ cos θ AR = Ar sin θ + Az cos θ
spherical θ = tan−1 (r/z) θ = r̂ cos θ − ẑ sin θ
θ̂ Aθ = Ar cos θ − Az sin θ
φ =φ φ = φ̂φ
φ̂ Aφ = Aφ
Spherical to r = R sin θ r̂ = R̂ sin θ + θ̂θ cos θ Ar = AR sin θ + Aθ cos θ
cylindrical φ =φ φ = φ̂φ
φ̂ Aφ = Aφ
z = R cos θ ẑ = R̂ cos θ − θ̂θ sin θ Az = AR cos θ − Aθ sin θ

The converse relations are

z
x = R sin θ cos φ , (3.61a)
θ
zˆ y = R sin θ sin φ , (3.61b)
ˆ
R
z = R cos θ . (3.61c)
(π/2 – θ)
R r̂ The unit vector R̂ lies in the r̂–ẑ plane. Hence, it can be
expressed as a linear combination of r̂ and ẑ as
θ
z = R cos θ
R̂ = r̂a + ẑb, (3.62)
y
where a and b are transformation coefficients. Since r̂ and ẑ
φ x = r cos φ are mutually perpendicular,
r
φ̂ R̂ · r̂ = a, (3.63a)
y = r sin φ
R̂ · ẑ = b. (3.63b)
rˆ
x
From Fig. 3-18, the angle between R̂ and r̂ is the complement
Figure 3-18 Interrelationships between (x, y, z) and (R, θ , φ ). of θ and that between R̂ and ẑ is θ . Hence, a = R̂ · r̂ = sin θ
146 CHAPTER 3 VECTOR ANALYSIS

and b = R̂ · ẑ = cos θ . Upon inserting these expressions for a Similarly,

and b in Eq. (3.62) and replacing r̂ with Eq. (3.56a), we have
Aθ = (x + y) cos θ cos φ + (y − x) cos θ sin φ − z sin θ ,
R̂ = x̂ sin θ cos φ + ŷ sin θ sin φ + ẑ cos θ . (3.64a) Aφ = −(x + y) sin φ + (y − x) cos φ ,

A similar procedure can be followed to obtain the expression and following the procedure used with AR , we obtain
for θ̂θ:
Aθ = 0, Aφ = −R sin θ .

θ = x̂ cos θ cos φ + ŷcos θ sin φ − ẑ sin θ .

θ̂ (3.64b) Hence,

A = R̂AR + θ̂θAθ + φ̂φAφ = R̂R − φ̂φR sin θ .

Finally φ̂φ is given by Eq. (3.56b) as

φ = −x̂ sin φ + ŷ cos φ .

φ̂ (3.64c) 3-3.3 Cylindrical to Spherical Transformations
Transformations between cylindrical and spherical coordinates
Equations (3.64a) through (3.64c) can be solved simultane- can be realized by combining the transformation relations
θ, φ̂
ously to give the expressions for (x̂, ŷ, ẑ) in terms of (R̂, θ̂ φ): of the preceding two subsections. The results are given in
Table 3-2.
x̂ = R̂ sin θ cos φ + θ̂θ cos θ cos φ − φ̂φ sin φ , (3.65a)
3-3.4 Distance between Two Points
ŷ = R̂ sin θ sin φ + θ̂θ cos θ sin φ + φ̂φ cos φ , (3.65b)
In Cartesian coordinates, the distance d between two points
ẑ = R̂ cos θ − θ̂θ sin θ . (3.65c) P1 = (x1 , y1 , z1 ) and P2 = (x2 , y2 , z2 ) is given by Eq. (3.12) as

Equations (3.64a) to (3.65c) also can be used to transform

Cartesian components (Ax , Ay , Az ) of vector A into their spher- d = |R12 | = [(x2 − x1 )2 + (y2 − y1 )2 + (z2 − z1 )2 ]1/2 .
ical counterparts (AR , Aθ , Aφ ), and vice versa, by replacing
θ, φ̂
(x̂, ŷ, ẑ, R̂, θ̂ φ) with (Ax , Ay , Az , AR , Aθ , Aφ ). (3.66)
Upon using Eq. (3.52) to convert the Cartesian coordinates of
P1 and P2 into their cylindrical equivalents, we have
Example 3-9: Cartesian to Spherical
Transformation
d = (r2 cos φ2 − r1 cos φ1 )2


1/2
Express vector A = x̂(x + y) + ŷ(y − x) + ẑz in spherical + (r2 sin φ2 − r1 sin φ1 )2 + (z2 − z1 )2
coordinates. 1/2
= r22 +r12 − 2r1 r2 cos(φ2 −φ1 )+(z2 −z1 )2

. (3.67)
Solution: Using the transformation relation for AR given in
(cylindrical)
Table 3-2, we have
AR = Ax sin θ cos φ + Ay sin θ sin φ + Az cos θ
A similar transformation using Eqs. (3.61a) through (3.61c)
= (x + y) sin θ cos φ + (y − x) sin θ sin φ + z cos θ . leads to an expression for d in terms of the spherical coordi-
nates of P1 and P2 :
Using the expressions for x, y, and z given by Eq. (3.61c), we
have
d = R22 + R21 − 2R1R2 [cos θ2 cos θ1

AR = (R sin θ cos φ + R sin θ sin φ ) sin θ cos φ
1/2
+ (R sin θ sin φ −R sin θ cos φ ) sin θ sin φ + R cos2 θ + sin θ1 sin θ2 cos(φ2 − φ1 )] . (3.68)

= R sin2 θ (cos2 φ + sin2 φ ) + R cos2 θ (spherical)

= R sin2 θ + R cos2 θ = R.
3-4 GRADIENT OF A SCALAR FIELD 147

Example 3-10: Vector Component Concept Question 3-8: Why is it that the base vectors
(x̂, ŷ, ẑ) are independent of the location of a point, but r̂
and φ̂φ are not?
At a given point in space, vectors A and B are given in
cylindrical coordinates by
Concept Question 3-9: What are the cyclic relations
for the base vectors in (a) Cartesian coordinates, (b) cylin-
A = r̂2 + φ̂φ3 − ẑ, drical coordinates, and (c) spherical coordinates?
B = r̂ + ẑ.
Concept Question 3-10: How is the position vector of
Determine (a) the scalar component of B, or projection, in the a point in cylindrical coordinates related to its position
direction of A, (b) the vector component of B in the direction vector in spherical coordinates?
of A, and (c) the vector component of B perpendicular to A.
√
Exercise 3-7: Point P = (2 3, π /3, −2) is given in cylin-
drical coordinates. Express P in spherical coordinates.
B
Answer: P = (4, 2π /3, π /3). (See EM
.)
D
Exercise 3-8: Transform vector
A
C
A = x̂(x + y) + ŷ(y − x) + ẑz
Figure 3-19 Vectors A, B, C, and D of Example 3-10.
from Cartesian to cylindrical coordinates.
φr + ẑz. (See
Answer: A = r̂r − φ̂ EM
.)
Solution: (a) Let us denote the scalar component of B in the
direction of A as C, as shown in Fig. 3-19. Thus,

C = B · â = B ·
A (r̂2 + φ̂φ3 − ẑ) 2 − 1
= (r̂ + ẑ) · √ = √ = 0.267.
3-4 Gradient of a Scalar Field
|A| 4+9+1 14
When dealing with a scalar physical quantity whose magnitude
(b) The vector component of B in the direction of A is given depends on a single variable, such as the temperature T as
by the product of the scalar component C and the unit vector â: a function of height z, the rate of change of T with height
can be described by the derivative dT/dz. However, if T is
A (r̂2 + φ̂φ3 − ẑ) also a function of x and y, its spatial rate of change becomes
C = âC = C= √ × 0.267 more difficult to describe because we now have to deal with
|A| 14
three separate variables. The differential change in T along x,
= r̂0.143 + φ̂φ0.214 − ẑ0.071. y, and z can be described in terms of the partial derivatives
of T with respect to the three coordinate variables, but it is
(c) The vector component of B perpendicular to A is equal not immediately obvious as to how we should combine the
to B minus C: three partial derivatives so as to describe the spatial rate of
change of T along a specified direction. Furthermore, many
D = B − C = (r̂ + ẑ) − (r̂0.143 + φ̂φ0.214 − ẑ0.071) of the quantities we deal with in electromagnetics are vectors;
therefore, both their magnitudes and directions may vary with
= r̂0.857 − φ̂φ0.214 + ẑ0.929. spatial position. To this end, we introduce three fundamental
operators to describe the differential spatial variations of
scalars and vectors: the gradient, divergence, and curl oper-
Concept Question 3-7: Why do we use more than one ators. The gradient operator applies to scalar fields and is the
coordinate system? subject of the present section. The other two operators, which
apply to vector fields, are discussed in succeeding sections.
148 CHAPTER 3 VECTOR ANALYSIS

Equation (3.71) then can be expressed as

z P2 = (x + dx, y + dy, z + dz)
dT = ∇T · dl. (3.73)
dy
dx The symbol ∇ is called the del or gradient operator and is
dz defined as
dl
P1 = (x, y, z) ∂ ∂ ∂
∇ = x̂ + ŷ + ẑ (Cartesian). (3.74)
∂x ∂y ∂z
y
x
◮ Whereas the gradient operator itself has no physical
Figure 3-20 Differential distance vector dl between points P1 meaning, it attains a physical meaning once it operates on
and P2 . a scalar quantity, and the result of the operation is a vector
with magnitude equal to the maximum rate of change of
the physical quantity per unit distance and pointing in the
direction of maximum increase. ◭
Suppose that T1 = T (x, y, z) is the temperature at point
P1 = (x, y, z) in some region of space, and
With dl = âl dl, where âl is the unit vector of dl, the directional
T2 = T (x + dx, y + dy, z + dz) derivative of T along âl is

is the temperature at a nearby point P2 = (x+dx, y+dy, z+dz)

(Fig. 3-20). The differential distances dx, dy, and dz are the dT
= ∇T · âl . (3.75)
components of the differential distance vector dl. That is, dl

dl = x̂ dx + ŷ dy + ẑ dz. (3.69) We can find the difference (T2 − T1 ), where T1 = T (x1 , y1 , z1 )

and T2 = T (x2 , y2 , z2 ) are the values of T at points
From differential calculus, the temperature difference between
points P1 and P2 , dT = T2 − T1 , is P1 = (x1 , y1 , z1 ) and P2 = (x2 , y2 , z2 )
∂T ∂T ∂T not necessarily infinitesimally close to one another, by inte-
dT = dx + dy + dz. (3.70)
∂x ∂y ∂z grating both sides of Eq. (3.73). Thus,

Because dx = x̂ · dl, dy = ŷ · dl, and dz = ẑ · dl, Eq. (3.70) can

Z P2
be rewritten as T2 − T1 = ∇T · dl. (3.76)
P1

∂T ∂T ∂T
dT = x̂ · dl + ŷ · dl + ẑ · dl
∂x ∂y ∂z

∂T ∂T ∂T
 Example 3-11: Directional Derivative
= x̂ + ŷ + ẑ · dl. (3.71)
∂x ∂y ∂z
Find the directional derivative of T = x2 + y2 z along direction
The vector inside the square brackets in Eq. (3.71) relates the x̂2 + ŷ3 − ẑ2 and evaluate it at (1, −1, 2).
change in temperature dT to a vector change in direction dl.
This vector is called the gradient of T (or grad T for short) Solution: First, we find the gradient of T :
and denoted ∇T : 
∂ ∂ ∂

∇T = x̂ + ŷ + ẑ (x2 + y2 z) = x̂2x + ŷ2yz + ẑy2 .
∂x ∂y ∂z
∂T ∂T ∂T
∇T = grad T = x̂ + ŷ + ẑ . (3.72)
∂x ∂y ∂z We denote l as the given direction,

l = x̂2 + ŷ3 − ẑ2.

3-4 GRADIENT OF A SCALAR FIELD 149

Its unit vector is Hence, the gradient operator in cylindrical coordinates can be
expressed as
l x̂2 + ŷ3 − ẑ2 x̂2 + ŷ3 − ẑ2
âl = =√ = √ .
|l| 2
2 +3 +22 2 17
∂ 1 ∂ ∂
Application of Eq. (3.75) gives ∇ = r̂ + φ̂φ + ẑ , (cylindrical)
∂r r ∂φ ∂z
 
dT 2 x̂2 + ŷ3 − ẑ2
= ∇T · âl = (x̂2x + ŷ2yz + ẑy ) · √ (3.82)
dl 17
4x + 6yz − 2y2 A similar procedure leads to the expression for the gradient in
= √ . spherical coordinates:
17
At (1, −1, 2), ∂ 1 ∂ 1 ∂
∇ = R̂ + θ̂θ + φ̂φ . (3.83)
∂R R ∂θ R sin θ ∂ φ
dT 4 − 12 − 2 −10
= √ =√ . (spherical)
dl (1,−1,2) 17 17

3-4.1 Gradient Operator in Cylindrical and 3-4.2 Properties of the Gradient Operator
Spherical Coordinates
For any two scalar functions U and V , the following relations
Even though Eq. (3.73) was derived using Cartesian coordi- apply:
nates, it should have counterparts in other coordinate systems.
To convert Eq. (3.72) into cylindrical coordinates (r, φ , z), we (1) ∇(U + V ) = ∇U + ∇V , (3.84a)
start by restating the coordinate relations (2) ∇(UV ) = U ∇V + V ∇U, (3.84b)
p y (3) ∇V n = nV n−1 ∇V , for any n. (3.84c)
r = x2 + y2 , tan φ = . (3.77)
x
From differential calculus,
∂T ∂T ∂r ∂T ∂φ ∂T ∂z Example 3-12: Calculating the Gradient
= + + . (3.78)
∂x ∂r ∂x ∂φ ∂x ∂z ∂x
Find the gradient of each of the following scalar functions and
Since z is orthogonal to x and ∂ z/∂ x = 0, the last term in then evaluate it at the given point.
Eq. (3.78) vanishes. From the coordinate relations given by
Eq. (3.77), it follows that (a) V1 = 24V0 cos (π y/3)sin (2π z/3) at (3, 2, 1) in Cartesian
coordinates,
∂r x (b) V2 = V0 e−2r sin 3φ at (1, π /2, 3) in cylindrical coordi-
=p = cos φ , (3.79a) nates,
∂x x + y2
2
(c) V3 = V0 (a/R)cos 2θ at (2a, 0, π ) in spherical coordinates.
∂φ 1
= − sin φ . (3.79b) Solution: (a) Using Eq. (3.72) for ∇,
∂x r
Hence, ∂ V1 ∂ V1 ∂ V1
∇V1 = x̂ + ŷ + ẑ
∂T ∂ T sin φ ∂ T ∂x ∂y ∂z
= cos φ − . (3.80)
∂x ∂r r ∂φ πy 2π z πy 2π z
= −ŷ8π V0 sin sin + ẑ16π V0 cos cos
This expression can be used to replace the coefficient of x̂ in 3 3 3 3
Eq. (3.72), and a similar procedure can be followed to obtain 
πy 2π z πy 2π z

an expression for ∂ T /∂ y in terms of r and φ . If, in addition, we = 8π V0 −ŷ sin sin + ẑ2 cos cos .
3 3 3 3
use the relations x̂ = r̂ cos φ − φ̂φ sin φ and ŷ = r̂ sin φ + φ̂φ cos φ
[from Eqs. (3.57a) and (3.57b)], then Eq. (3.72) becomes At (3, 2, 1),

2π 2π

∂T 1 ∂T ∂T
∇T = r̂ + φ̂φ + ẑ . (3.81) ∇V1 = 8π V0 −ŷ sin2 + ẑ2 cos2 = π V0 [−ŷ6 + ẑ4] .
∂r r ∂φ ∂z 3 3
150 CHAPTER 3 VECTOR ANALYSIS

Module 3.2 Gradient Select a scalar function f (x, y, z), evaluate its gradient, and display both in an appropriate 2-D plane.

(b) The function V2 is expressed in terms of cylindrical At (2a, 0, π ), R = 2a and θ = 0, which yields
variables. Hence, we need to use Eq. (3.82) for ∇:
V0
 
∂ 1 ∂ ∂ ∇V3 = −R̂
∇V2 = r̂ + φ̂φ + ẑ V0 e−2r sin 3φ 4a
.
∂r r ∂φ ∂z
= −r̂2V0 e−2r sin 3φ + φ̂φ(3V0 e−2r cos 3φ )/r
 
3 cos 3φ
= −r̂2 sin 3φ + φ̂φ V0 e−2r .
r
Exercise 3-9: Given V = x2 y + xy2 + xz2 , (a) find the
At (1, π /2, 3), r = 1 and φ = π /2. Hence, gradient of V , and (b) evaluate it at (1, −1, 2).
 
3π 3π Answer: (a) ∇V = x̂(2xy + y2 + z2 ) + ŷ(x2 + 2xy)
∇V2 = −r̂2 sin + φ̂φ3 cos V0 e−2 = r̂2V0 e−2 = r̂0.27V0.
2 2 + ẑ2xz, (b) ∇V (1,−1,2) = x̂3 − ŷ + ẑ4. (See EM .)
(c) As V3 is expressed in spherical coordinates, we apply
Eq. (3.83) to V3 :
   
∂ 1 ∂ 1 ∂ a
∇V3 = R̂ + θ̂θ + φ̂φ V0 cos 2θ
∂R R ∂θ R sin θ ∂ φ R Exercise 3-10: Find the directional derivative of
V = rz2 cos 2φ along the direction of A = r̂2 − ẑ and
V0 a 2V0 a
= −R̂ cos 2θ − θ̂θ 2 sin 2θ evaluate it at (1, π /2, 2).
R2 R √
V0 a Answer: (dV/dl) (1,π /2,2) = −4/ 5 . (See EM .)
= −[R̂ cos 2θ + θ̂θ2 sin 2θ ] 2 .
R
3-4 GRADIENT OF A SCALAR FIELD 151

Technology Brief 5:
Global Positioning System
The Global Positioning System (GPS), initially devel-
oped in the 1980s by the U.S. Department of Defense
as a navigation tool for military use, has evolved into
a system with numerous civilian applications, including
vehicle tracking, aircraft navigation, map displays in au-
tomobiles and hand-held cell phones (Fig. TF5-1), and
topographic mapping. The overall GPS comprises three
segments. The space segment consists of 24 satellites
(Fig. TF5-2), each circling Earth every 12 hours at an
orbital altitude of about 12,000 miles and transmitting
continuous coded time signals. All satellite transmitters
broadcast coded messages at two specific frequencies:
1.57542 GHz and 1.22760 GHz. The user segment
consists of hand-held or vehicle-mounted receivers that
determine their own locations by receiving and process-
ing multiple satellite signals. The third segment is a
network of five ground stations distributed around the
world that monitor the satellites and provide them with
updates on their precise orbital information.

◮ GPS provides a location inaccuracy of about

30 m both horizontally and vertically, but it can be Figure TF5-1 iPhone map feature.
improved to within 1 m by differential GPS. (See
final section.) ◭

Principle of Operation
The triangulation technique allows the determination
of the location (x0 , y0 , z0 ) of any object in 3-D space
from knowledge of the distances d1 , d2 , and d3 between
that object and three other independent points in space
of known locations (x1 , y1 , z1 ) to (x3 , y3 , z3 ). In GPS, the
distances are established by measuring the times it
takes the signals to travel from the satellites to the GPS
receivers, and then multiplying them by the speed of
light c = 3 × 108 m/s. Time synchronization is achieved
by using atomic clocks. The satellites use very precise
clocks, accurate to 3 nanoseconds (3 × 10−9 s), but
receivers use less accurate, inexpensive, ordinary quartz
clocks. Consequently, the receiver clock may have an Figure TF5-2 GPS nominal satellite constellation. Four
satellites in each plane with 20,200 km altitudes and a 55◦
unknown time offset error t0 relative to the satellite
inclination.
clocks. To correct for the time error of a GPS receiver,
a signal from a fourth satellite is needed.
152 CHAPTER 3 VECTOR ANALYSIS

similar process applies to satellites 2 to 4, leading to four

equations:
SAT 2
SAT 3 (x2, y2, z2)
(x3, y3, z3) d3 d12 = (x1 − x0 )2 + (y1 − y0 )2 + (z1 − z0 )2 = c2 [(t1 + t0 )]2 ,
d2
SAT 4 d22 = (x2 − x0 )2 + (y2 − y0 )2 + (z2 − z0 )2 = c2 [(t2 + t0 )]2 ,
(x4, y4, z4) SAT 1
(x1, y1, z1) d32 = (x3 − x0 )2 + (y3 − y0 )2 + (z3 − z0 )2 = c2 [(t3 + t0 )]2 ,
d4
d42 = (x4 − x0 )2 + (y4 − y0 )2 + (z4 − z0 )2 = c2 [(t4 + t0 )]2 .
d1
The four satellites report their coordinates (x1 , y1 , z1 ) to
(x4 , y4 , z4 ) to the GPS receiver, and the time delays t1 to
t4 are measured directly by the receiver. The unknowns
Time delay (x0, y0, z0) are (x0 , y0 , z0 ), the coordinates of the GPS receiver, and
the time offset of its clock t0 . Simultaneous solution of the
four equations provides the desired location information.

Receiver Code Differential GPS

The 30 m GPS position inaccuracy is attributed to several
Satellite Code
factors, including time-delay errors (due to the differ-
ence between the speed of light and the actual signal
speed in the troposphere) that depend on the receiver’s
Figure TF5-3 Automobile GPS receiver at location location on Earth, delays due to signal reflections by tall
(x0 , y0 , z0 ).
buildings, and satellites’ locations misreporting errors.

◮ Differential GPS, or DGPS, uses a stationary

reference receiver at a location with known coordi-
The GPS receiver of the automobile in Fig. TF5-3 is at nates. ◭
distances d1 to d4 from the GPS satellites. Each satellite
sends a message identifying its orbital coordinates
(x1 , y1 , z1 ) for satellite 1, and so on for the other satellites, By calculating the difference between its location on the
together with a binary-coded sequence common to all basis of the GPS estimate and its true location, the
satellites. The GPS receiver generates the same binary reference receiver establishes coordinate correction
sequence (Fig. TF5-3), and by comparing its code with factors and transmits them to all DGPS receivers in the
the one received from satellite 1, it determines the time area. Application of the correction information usually
t1 corresponding to travel time over the distance d1 . A reduces the location inaccuracy down to about 1 m.
3-5 DIVERGENCE OF A VECTOR FIELD 153

Exercise 3-11: The power density radiated by a star where x is measured in kilometers and x = 0 is the sea–
(Fig. E3.11(a)) decreases radially as S(R) = S0 /R2 , where land boundary. (a) In which direction does ∇T point and
R is the radial distance from the star and S0 is a constant. (b) at what value of x is it a maximum?
Recalling that the gradient of a scalar function denotes the
maximum rate of change of that function per unit distance T
and the direction of the gradient is along the direction of
T2
maximum increase, generate an arrow representation of
∇S.
T1
S x
Sea Land

(a)

∆
T
(a)

x
Sea Land

(b)
∆ Figure E3.12
S

Answer: (a) +x̂; (b) at x = 0.

T2 − T1
(b) T (x) = T1 + ,
e−x + 1
∂T e−x (T2 − T1 )
∇T = x̂ = x̂ .
∂x (e−x + 1)2

(See EM
.)
Figure E3.11

Answer: ∇S = −R̂ 2S0 /R3 (Fig. E3.11(b)). (See EM

.)
3-5 Divergence of a Vector Field
From our brief introduction of Coulomb’s law in Chapter 1, we
know that an isolated, positive point charge q induces an elec-
tric field E in the space around it with the direction of E being
Exercise 3-12: The graph in Fig. E3.12(a) depicts a gen- outward away from the charge. Also, the strength (magnitude)
tle change in atmospheric temperature from T1 over the of E is proportional to q and decreases with distance R from
sea to T2 over land. The temperature profile is described the charge as 1/R2 . In a graphical presentation, a vector field is
by the function usually represented by field lines, as shown in Fig. 3-21. The
T2 − T1 arrowhead denotes the direction of the field at the point where
T (x) = T1 + , the field line is drawn, and the length of the line provides a
e−x + 1
qualitative depiction of the field’s magnitude.