Chapter 1 Vector Analysis 1.1.

1 Vector Operations (II)

1.1 Vector Algebra: 1.1.1 Vector Operations (I)
Vectors: Quantities have both magnitude and direction, (i) Addition of two vectors:
denoted by boldface (A, B, and so on). Place the tail of B at the head of A.
Scalars: Quantities have magnitude but no direction denoted Commutative: A+B=B+A
by ordinary type. Associative: (A+B)+C=A+(B+C)

In diagrams, vectors are denoted by arrows: the length of A-B=A+(-B)

the arrow is proportional to the magnitude of the vector, B -B
and the arrowhead indicates its direction.
Minus A (-A) is a vector with the same magnitude as A but A B+A A
A+B A A-B
of opposite direction.
Vectors have magnitude and direction but not location.
B
1 2

1.1.1 Vector Operations (III) 1.1.1 Vector Operations (IV)

(ii) Multiplication by a scalar:
Multiplies the magnitude but leaves the direction unchanged.
Distributive: a(A+B)= aA+ aB
(iii) Dot product of two vector (scalar product):
The dot product of two vectors is defined by A·B≡AB cosθ,
where θ is the angle they form when placed tail-to-tail.
(iv) Cross product of two vector (vector product):
The cross product of two vectors is defined by
A×B≡AB sinθ n̂ , where n̂ is a unit vector pointing
perpendicular to the plane of A and B.
A hat is used to designate the unit vector and its direction is
determined by the right-hand rule.

Commutative: A·B=B·A Distributive: A×(B+C)=A×B+A×C A

A
not commutative: A×B=-B×A θ
Distributive: A·(B+C)=A·B+A·C θ B
B
3 4
 1.1.2 Vector Algebra: Component form (I)
Let x̂, ŷ and ẑ be unit vectors parallel to the x, y, and z
axes, respectively. An arbitrary vector A can be expressed in
terms of these basis vectors.
A = Ax x̂ + Ay ŷ + Az ẑ
The numbers Ax, Ay, and Az are called components.
1.1.2 Vector Algebra: Component form (II)
Reformulate the four vector operations as a rule for
manipulating components:
(i) To add vectors, add like components.
A + B = ( Ax x̂ + Ay ŷ + Az ẑ) + ( Bx x̂ + B y ŷ + Bz ẑ)
= ( Ax + Bx )x̂ + ( Ay + B y )ŷ + ( Az + Bz )ẑ
(ii) To multiply by a scalar, multiply each component.
aA = a ( Ax x̂ + Ay ŷ + Az ẑ)
= aAx x̂ + aAy ŷ + aAz ẑ

5 6

1.1.2 Vector Algebra: Component form (III) 1.1.3 Triple Products (I)
(iii) To calculate the dot product, multiply like components, Since the cross product of two vectors is itself a vector, it
and add. can be dotted or crossed with a third vector to form a
A ⋅ B = ( Ax x̂ + Ay ŷ + Az ẑ) ⋅ ( Bx x̂ + B y ŷ + Bz ẑ) triple product.

= Ax Bx + Ay B y + Az Bz (i) Scalar triple product: A·(B×C). Geometrically,

|A·(B×C)| is the volume of a parallelepiped generated by
(iv) To calculate the cross product, form the determinant these three vectors as shown below.
whose first row is x̂, ŷ and ẑ , whose second row is A
(in component form), and whose third row is B. A ⋅ (B × C) = B ⋅ (C × A ) = C ⋅ ( A × B)
x̂ ŷ ẑ ( Ay Bz − Az B y ) x̂ In component form Ax Ay Az
A × B = Ax Ay Az = + ( Az Bx − Ax Bz ) ŷ A ⋅ ( B × C) = B x By Bz
Bx By Bz + ( Ax B y − Ay Bx )ẑ Cx Cy Cz
7 8
 1.1.4 Position, Displacement, and Separation
1.1.3 Triple Products (II)
Vectors (I)
(ii) Vector triple product: A×(B×C). The vector triple Position vector: The vector to that point from the origin.
product can be simplified by the so-called BAC-CAB
rule. r ≡ xxˆ + yyˆ + zzˆ
A × (B × C) = B( A ⋅ C) − C( A ⋅ B)
Its magnitude (the distance from the origin)
Notice that ( A × B) × C ≠ A × (B × C) r = r ⋅ r ≡ x2 + y2 + z 2
( A × B) × C = −C × ( A × B) = − A (B ⋅ C) + B( A ⋅ C) Its direction unit vector (pointing radially outward)
r xxˆ + yyˆ + zzˆ
rˆ = =
Problem 1.6 Under what conditions does r x2 + y2 + z 2
( A × B) × C = A × (B × C) ? The infinitesimal displacement vector, from (x, y, z)
Ans: Either A is parallel to C, to (x+dx, y+dy, z+dz), is
or B is perpendicular to A and C 9
dl = dxxˆ + dyyˆ + dzzˆ 10

1.1.4 Position, Displacement, and Separation 1.2 Differential Calculus

Vectors (II) 1.2.1 “Ordinary” Derivatives
In electrodynamics one frequently encounters problems Suppose we have a function of
involving two points: one variable, f(x). What does the
A source point, r′, where an electric field is located derivative, df/dx, do for us?
A field point, r, at which you are calculating the electric field

A short-hand notation for the Ans: It tells us how rapidly the function f(x) varies when we
separation vector from the source change the argument x by a tiny amount, dx.
point to the field point is  df 
df =  dx
r ≡ r − r′, magnitude r = r − r′  dx 
In words, if we change x by an amount dx, then, f changes
r r − r′
unit vector in the direction form r′ to r is rˆ = = by an amount df.
r r − r′ The derivative df/dx is the slope of the graph of f versus x.
11 12
 1.2.2 Gradient (I) 1.2.2 Gradient (II)
Suppose we have a function of three A mountain hill Geometrical interpretation: Like any vector, the gradient
variables. What does the derivative H ( x, y , z ) has magnitude and direction.
mean in this case? A dot product in abstract form is: dH = ∇H ⋅ dl = ∇H dl cos θ
A theorem on partial derivatives states that where θ is the angle between ∇H and dl.
∂H ∂H ∂H
dH = dx + dy + dz
∂x ∂y ∂z The gradient ∇H points in the direction of maximum
∂H ∂H ∂H increase of the function H.
=( xˆ + yˆ + zˆ ) ⋅ (dxxˆ + dyyˆ + dzzˆ )
∂x ∂y ∂z
= (∇H ) ⋅ (dl )
Analogous to the derivative of one variable, a vanishing
The gradient of H is a vector quantity, with three components. derivative signals a maximum, a minimum, or an inflection.
∂H ∂H ∂H
∇H = xˆ + yˆ + zˆ
∂x ∂y ∂z 13 14

Example 1.3 & Problem 1.13 1.2.3 The Operator ∇ (I)

Example 1.3 Find the gradient of r = x2 + y2 + z 2 The gradient has the formal appearance of a vector, ∇,
“multiplying”, a scalar H.
∂r ∂r ∂r xxˆ + yyˆ + zzˆ r
Ans : ∇r = xˆ + yˆ + zˆ = = = rˆ ∂ ∂ ∂
∂x ∂y ∂z x2 + y2 + z 2 r ∇H = (xˆ + yˆ + zˆ ) H
∂x ∂y ∂z
Problem 1.13 Let r ≡ ( x − x ' ) xˆ + ( y − y ' ) yˆ + ( z − z ' )zˆ
del
Show that ∇ is a vector operator that acts upon H, not a vector that
multiplies H.
(a) ∇r 2 = ? ∇r 2 = ∇[( x − x' ) 2 + ( y − y ' ) 2 + ( z − z ' ) 2 ] ∂ ∂ ∂
∇ = xˆ + yˆ + zˆ
∂x ∂y ∂z
= 2( x − x' )xˆ + 2( y − y ' )yˆ + 2( z − z ' )zˆ = 2r
− ∇r − ∇ ( x − x ' ) 2 + ( y − y ' ) 2 + ( z − z ' ) 2 ∇ mimics the behavior of an ordinary vector in virtually
(b) ∇(1 r ) == ? = every way, if we translate “multiply” by “act upon”.
r2 ( x − x' ) 2 + ( y − y ' ) 2 + ( z − z ' ) 2
rˆ It is a marvelous piece of notational simplification.
= − [2( x − x' )xˆ + 2( y − y ' )yˆ + 2( z − z ' )zˆ ] / r = − 2 15
1
2
3
16
r
 1.2.3 The Operator ∇ (II) 1.2.4 The Divergence
An ordinary vector A can be multiply in three ways: Divergence of a vector v is:
∂ ∂ ∂
1. Multiply a scalar a : aA ∇ ⋅ v = (xˆ + yˆ + zˆ ) ⋅ (v x xˆ + v y yˆ + v z zˆ )
∂x ∂y ∂z
2. Multiply another vector (dot product): A·B
∂v ∂v y ∂v z
3. Multiply another vector (cross product): A×B = x+ +
∂x ∂y ∂z
∇·v is a measure of how much the vector v spread out
Correspondingly, there are three ways the operator ∇ can act: from the point in question.
1. On a scalar function H: ∇H (Gradient 梯度)
2. On a vector function (dot product): ∇·v (divergence 散度)
3. On a vector function (cross product): ∇× v (curl 旋度)

17 positive zero positive 18

Example 1.4 1.2.5 The Curl

Curl of a vector v is:
xˆ yˆ zˆ
∂ ∂ ∂ ∂v ∂v y ∂v ∂v ∂v y ∂v x
∇× v = = xˆ ( z − ) + yˆ ( x − z ) + zˆ ( − )
∂x ∂y ∂z ∂y ∂z ∂z ∂x ∂x ∂y
Example 1.4 Suppose the functions in above three figures vx vy vz
are v a = xxˆ + yyˆ + zzˆ , v b = zˆ , v c = zzˆ . Calculate their
divergences. ∇×v is a measure of how much the vector v curl around
∂x ∂y ∂z the point in question.
Ans : ∇ ⋅ v a = + + = 3,
∂x ∂y ∂z
∂ 0 ∂ 0 ∂1
∇ ⋅ vb = + + = 0,
∂x ∂y ∂z
∂ 0 ∂ 0 ∂z
∇ ⋅ vc = + + = 1.
∂x ∂y ∂z 19 20
 Example 1.5 1.2.6 Product Rules (I)
The sum rule:

d df dg
( f + g) = + ∇ ( f + g ) = ∇f + ∇g
dx dx dx
Example 1.5 Suppose the functions in above two figures ∇ ⋅ ( A + B) = ∇ ⋅ A + ∇ ⋅ B ∇ × ( A + B) = ∇ × A + ∇ × B
are v a = − yxˆ + xyˆ , v b = xyˆ Calculate their curls.

∂ 0 ∂x ∂ (− y ) ∂ 0 ∂x ∂ (− y ) The rule for multiplying by a constant:

Ans : ∇ × v a = xˆ ( − ) + yˆ ( − ) + zˆ ( − ) = 2zˆ
∂y ∂z ∂z ∂x ∂x ∂y
d df
∂0 ∂0 ∂0 ∂0 ∂x ∂ 0 (kf ) = k ∇(kf ) = k∇f
∇ × v b = xˆ ( − ) + yˆ ( − ) + zˆ ( − ) = zˆ dx dx
∂y ∂z ∂z ∂x ∂x ∂y
∇ ⋅ (kA ) = k∇ ⋅ A ∇ × (kA ) = k∇ × A

21 22

1.2.6 Product Rules (II) 1.2.6 Product Rules (III)

 scalar : fg
The product rule:  The quotient rule:
vector : fA  f A
d df dg g
df
−f
dg scalar : vector :
( fg ) = g +f ∇( fg ) = g∇f + f∇g d f  g g
dx dx dx ( ) = dx 2 dx
dx g g
∇ ⋅ ( fA ) = ∇f ⋅ A + f (∇ ⋅ A ) ∇ × ( fA ) = ∇f × A + f (∇ × A)
f g∇f − f∇g
 scalar : A ⋅ B ∇( ) =
 g g2
vector : A × B
∇( A ⋅ B) = A × (∇ × B) + B × (∇ × A ) + ( A ⋅ ∇)B + (B ⋅ ∇) A A g (∇ ⋅ A) − A ⋅ ∇g
∇⋅( ) =
Chaps. g g2
∇ ⋅ ( A × B) = B ⋅ (∇ × A) − A ⋅ (∇ × B)
8 and 10
A g (∇ × A ) − (∇g × A) g (∇ × A) + A × ∇g
∇ × ( A × B) = (B ⋅ ∇) A − ( A ⋅ ∇)B + A(∇ ⋅ B) − B(∇ ⋅ A ) ∇×( ) = =
g g2 g2
23 24
 1.2.7 Second Derivatives (I) 1.2.7 Second Derivatives (II)
∂ ∂ ∂ ∂T ∂T ∂T
By applying ∇ twice, we can construct five species of (1) ∇ ⋅ (∇T ) = (xˆ + yˆ + zˆ ) ⋅ (xˆ + yˆ + zˆ )
second derivatives. ∂x ∂y ∂z ∂x ∂y ∂z
Three first derivatives ∇T , ∇ ⋅ v, ∇ × v ∂ 2T ∂ 2T ∂ 2T
=
+ 2 + 2 = ∇ 2T the Laplacian of T
∂x 2
∂y ∂z
(1) Divergence of gradient : ∇ ⋅ (∇T ) very important The Laplacian of a vector is similar:
(2) Curl of gradient : ∇ × (∇T ) always zero (∇ ⋅ ∇) v ≡ ∇ 2 (xˆ v x + yˆ v y + zˆ v z ) = xˆ ∇ 2 v x + yˆ ∇ 2 v y + zˆ ∇ 2 v z
(3) Gradient of divergence : ∇(∇ ⋅ v ) Chaps. 8 and 10
(2) ∇ × (∇T ) ≠ (∇ × ∇)T
(4) Divergence of curl : ∇ ⋅ (∇ × v ) always zero The proof hinges on the equality of cross derivatives:
∂ ∂ ∂ ∂T ∂T ∂T
(5) Curl of curl : ∇ × (∇ × v ) ∇ × (∇T ) = (xˆ + yˆ + zˆ ) × (xˆ + yˆ + zˆ )=0
reduce to others ∂x ∂y ∂z ∂x ∂y ∂z
∂ ∂T ∂ ∂T ∂ ∂T ∂ ∂T ∂ ∂T ∂ ∂T
( ) = ( ), ( ) = ( ), ( )= ( )
25 ∂x ∂y ∂y ∂x ∂y ∂z ∂z ∂y ∂z ∂x ∂x ∂z 26

1.3 Integral Calculus

1.2.7 Second Derivatives (III)
1.3.1 Line, Surface, and Volume (I)
∂ ∂v ∂v y ∂ ∂v ∂v ∂ ∂v y ∂v x
(4) ∇ ⋅ (∇ × v ) = xˆ (xˆ ( z − )) + yˆ (yˆ ( x − z )) + zˆ (zˆ ( − )) In electrodynamics, the line (or path) integrals, surface
∂x ∂y ∂z ∂y ∂z ∂x ∂z ∂x ∂y
integrals (or flux), and volume integrals are the most
∂ ∂v ∂v y ∂ ∂v ∂v ∂ ∂v y ∂v x important integrals.
= ( z− )+ ( x − z )+ ( − )
∂x ∂y ∂z ∂y ∂z ∂x ∂z ∂x ∂y
(a) Line integrals: a line integral is an expression of the
=0 always zero b
∫
form
v ⋅ dl ,
(5) ∇ × (∇ × v ) Can we use the following vector identity? aP

C × ( A × B) = A (B ⋅ C) − B( A ⋅ C) Where v is a vector function, dl is the infinitesimal

displacement vector, and the integral is to be carried out
∂ ∂ ∂ ∂v ∂v y ∂v ∂v ∂v y ∂v x
∇ × (∇ × v ) = (xˆ + yˆ + zˆ ) × (xˆ ( z − ) + yˆ ( x − z ) + zˆ ( − )) along a prescribed path P from point a to point b.
∂x ∂y ∂z ∂y ∂z ∂z ∂x ∂x ∂y
= … = ∇(∇ ⋅ v ) − ∇ 2 v Put a circle on the integral, in the path in question
We will encounter this derivative when dealing with forms a closed loop.
the vector potential (magnetism).
27
∫ v ⋅ dl 28
 1.3.1 Line, Surface, and Volume (II)
The value of a line integral depends critically on the
particular path taken from a to b, but there is an important
special class of vector functions for which the line integral
is independent of the path, and is determined entirely
by the end points, e.g. b
W = ∫ F ⋅ dl
aP
A force that has this property is called conservative.
29
1.3.1 Line, Surface, and Volume (III)
(b) Surface integrals: a line integral is an expression of
the form
∫ v ⋅ d a,
S
where v is a vector function, and da is
the infinitesimal patch of area, with
direction perpendicular to the surface.
The value of a surface integral depends on the particular
surface chosen, but there is a special class of vector
functions for which it is independent of the surface, and is
determined entirely by the boundary.
31
Example 1.6 Calculate the line integral of the function
v = y 2 xˆ + 2 x( y + 1)yˆ , from the point a=(1,1,0) to the point
b=(2,2,0), along the paths (1) and (2) in Fig.1.21. What is
the loop integral that goes from a to b along (1) and
returns to a along (2)?
The strategy here is to get everything in
terms of one variable.
30
Example 1.7 Calculate the surface integral of the function
v = 2 xzxˆ + (2 + x)yˆ + y ( z 2 − 3)zˆ over five sides of the
cubical box. Let ”upward and outward” be the positive
direction, as indicated by the arrow.
Sol : Taking the sides one at a time :
(1) x = 2, da = dydzxˆ , v ⋅ da = 2 xzdydz = 4 zdydz
2 2
∫ v ⋅ da = 4∫ dy ∫ zdz = 16
0 0
(5) z = 2, da = dxdyzˆ , v ⋅ da = y ( z 2 − 3)dxdy = ydxdy
2 2
∫ v ⋅ da = ∫ dx ∫ ydy = 4
0 0
32
 1.3.1 Line, Surface, and Volume (IV)
(c) Volume integrals: a line integral is an expression of Example 1.8 Calculate the volume integral of the function
the form
T = xyz 2 over the prism in Fig. 1.24.
∫ Tdτ ,
v
where T is a scalar function, and dτ is an infinitesimal Sol : Let' s do z first (0 to 3); then y from 0 to 1 - x;
volume element. In Cartesian coordinates, dτ=dxdydz
finally x from 0 to 1.
For example, if T is a density of a substance, then the
volume integral would give the total mass.
2 2
0
3
∫∫∫ xyz dxdydz = ∫ z dz ∫ x(∫ { 1

0
1− x

0
ydy )dx }
The volume integrals of vector functions:  1 1 
= 9∫ x( (1 − x) 2 )dx 
 
∫ vdτ = ∫ (v xˆ + v yˆ + v zˆ )dτ
0 2
x y z
1 1 3
= 9( )( ) =
= xˆ ∫ v dτ + yˆ ∫ v dτ + zˆ ∫ v dτ
x y z 2 12 8
33 34

1.3.2 The Fundamental Theorem of Calculus 1.3.3 The Fundamental Theorem for Gradients
Fundamental theorem of calculus: A scalar function of three variables T(x, y, z) changes by
b df b a small amount.
∫
a dx
dx = ∫ df = f (b) − f (a )
a
dT = (∇T ) ⋅ dl1

Geometrical Interpretation: two ways to determine the total The total change in T in going from a to b along the path
change in the function: selected is:
1. go step-by-step adding up all the tiny increments as you go b
2. subtract the values at the ends. ∫a
(∇T ) ⋅ dl = T (b) − T (a)

Fundamental theorem for gradient.

Geometrical Interpretation: Measure the high of a skyscraper.

1. Measure the high of each floor and add them all up.
The integral of a derivative over an interval is given by the 2. Place an altimeter at the top and the bottom, subtract the
value of the function at the end points (boundary). readings at the ends.
35 36
1.3.3 The Fundamental Theorem for Gradients (II) Potential Energy and Conservative Forces
b
∫
a
(∇T ) ⋅ dl = T (b) − T (a) the right side of this equation makes
Potential energy defined in terms of work done by the
no reference to the path---only to the end points. associated conservative force.
Thus gradients have special property that their line integrals B
are path independent. U B − U A = − ∫ Fc ⋅ ds
A

b
Corollary 1: ∫
a
(∇T ) ⋅ dl is independent of path taken from *Conservative forces tend to minimize the potential
a to b. energy within any system: It allowed to, an apple falls
Corollary 2: ∫ (∇T ) ⋅ dl = 0 , since the beginning and end to the ground and a spring returns to its natural length.
points are identical, and hence T(b)-T(a)=0.
Non-conservative force does not imply it is dissipative,
for example, magnetic force, and also does not mean it
A conservative force may be associated with a scalar
will decrease the potential energy, such as hand force.
potential energy function, whereas a non-conservative
force cannot.
37 38

Conservative Force and

Distinction Between
Potential Energy Function
Conservative and Non-conservative Forces
How can we find a conservative force if the associated
The distinction between conservative and non- potential energy function is given?
conservative forces is best stated as follows: A conservative force can be derived from a scalar
A conservative force may be associated with a scalar potential energy function.
potential energy function, whereas a non-conservative Fc = −∇U
force cannot.
The negative sign indicates that the force points in the
B direction of decreasing potential energy.
U B − U A = − ∫ Fc ⋅ ds
A
dU g
Gravity U g = mgy; Fy = − = − mg
Fc = −∇U dy
dU sp
1 2
Spring U sp = kx ; Fx = − = − kx
2 dx
39 40
 Supplementary
Gauss’s divergence theorem
1.3.4 The Fundamental Theorem for Divergences (Transformation between volume integrals and surface integrals)
The fundamental theorem for divergences states that:
∫v (∇ ⋅ v)dτ = ∫ v ⋅ nˆ da
∫ (∇ ⋅ v)dτ = ∫ v ⋅ da
v
S
Rough
S
v = vx xˆ + v y yˆ + vz zˆ and nˆ = cos α xˆ + cos β yˆ + cos γ zˆ
proof:
The integration of a derivative (in this case the divergence) where α , β , and γ are the angles between nˆ and x-, y -
over a region (in this case a volume) is equal to the value of
and z - axis, respectively.
the function at the boundary (in this case the surface that
bounds the volume) ∂vx ∂v y ∂vz
∫v (∇ ⋅ v ) dτ = ∫∫∫ ∂x + ∂y + ∂z )dxdydz
(
This theorem has at least three special names: Gauss’s v
theorem, Green’s theorem, or the divergence theorem.
= ∫∫ (vx dydz + v y dzdx + vz dxdy )
Geometrical Interpretation: Measure the total amount of S
fluid passing out through the surface, per unit time. = ∫∫ (vx cos α + v y cos β + vz cos γ )da
1. Count up all the faucets, recording how much each put out. S
2. Go around the boundary, measuring the flow at each point, Rigorous proof can be found in: Erwin Kreyszig, Advanced

and add it all up. 41 = ∫∫ v ⋅ nˆ da Engineering Mathematics (John Wiley and Sons, New
York, 1993), 7th ed. Chap. 9, pp. 546-547. 42
S

Example 1.10 Check the divergence theorem using the 1.3.5 The Fundamental Theorem for Curls (I)
function v = y 2 xˆ + (2 xy + z 2 ) yˆ + ( 2 yz )zˆ
The fundamental theorem for curls---Stokes’ theorem---
and the unit cube situated at the origin. states that:
∫ (∇ × v) ⋅ da = ∫ v ⋅ dl
S
Sol : In this case ∇ ⋅ v = 2( x + y ) P
The integration of a derivative (here, the curl) over a region
1 1 1
∫ 2( x + y)dxdydz =2∫ dz ∫
v 0 ∫ ( x + y)dxdy
0 0
(here, a patch of surface) is equal to the value of the
function at the boundary (in this case the perimeter of the
1 1 1
= 2∫ ∫ ( 12 + y )dy = 2 ∫ ( 12 + y )dy = 2 patch).
0 0 0

∴ ∫ ∇ ⋅ v dτ = 2 Geometrical Interpretation:
v
Measure the “twist” of the
To evaluate the surface integral we must consider vectors v; a region of high
separately the six sides of the cube. The total flux is… curl is a whirlpool.

43 44
 Supplementary

1.3.5 The Fundamental Theorem for Curls (II) Stokes’ theorem

(Transformation between surface integrals and line integrals)
Ambiguity in Stokes’ theorem: Concerning the boundary
line integral, which way are we supposed to go around
(clockwise or counterclockwise)? The right-hand rule. ∫ (∇ × v) ⋅ da = ∫ v ⋅ dl
S
P

Corollary 1: ∫ (∇ × v) ⋅ da depends only on the boundary

lines, not on the particular surface used.
Corollary 2: ∫ (∇ × v) ⋅ da = 0 for any closed surface, since Rigorous proof can be found in:
the boundary line shrinks down to a point. Erwin Kreyszig, Advanced Engineering Mathematics
(John Wiley and Sons, New York, 1993),
7th ed. Chap. 9, pp. 556-559.

These corollaries are analogous to

those for the gradient theorem.
45 46

Comments: graduate level (reference only)

Example 1.11 Suppose v = (2 xz + 3 y 2 )yˆ + (4 yz 2 )zˆ
• Green’s theorems:
Check Stokes’ theorem for the square surface shown below.
Let v = f ∇g ⇒ ∇ ⋅ v = ∇ ⋅ ( f ∇g ) = f ∇ 2 g + ∇f ⋅∇g
v ⋅ nˆ = f (nˆ ⋅∇g )
Sol : ∇ × v = (4 z 2 − 2 x)xˆ + 2 zzˆ ; da = dydzxˆ
∂g
∫v ( f ∇ g + ∇f ⋅∇g )dτ = ∫
2
Green's first formula: f da 1 1 4
∂n
∫ ∇ × v ⋅ a = ∫0 ∫0 =
2
S ( ) d 4 z dydz
3
∂g ∂f
∫v ( f ∇ g − g ∇ 2 f )dτ = ∫(f
2
Green's second formula: − g )da The line integral of the four segments
∂n ∂n
S
• Green’s theorem in the plane as a special case of
Stokes’ theorem
Let v be a vector function in the xy -plane.
∂v y ∂vx ∂v y ∂vx
(∇ × v ) ⋅ nˆ =
∂x
−
∂y
⇒ ∫∫ ( ∂x −
∂y
)da = ∫ (vx dx + v y dy)
S P 47 48
 1.3.6 Integration by Parts
Homework #1
d
( fg ) = g
df
+f
dg ∇ ⋅ ( fA) = ∇f ⋅ A + f (∇ ⋅ A)
dx dx dx Integrate it over a volume and
Integrating both sides and inviking the divergence theorem. Problems: 1.5, 1.7, 1.13, 1.16, 1.32,
invoking the fundamental theorem
Left ∫ ∇ ⋅ ( fA)dτ = ∫ ( fA) ⋅ da
b d
Left ∫ ( fg )dx = fg a
b
a dx Right ∫ (∇f ⋅ A + f (∇ ⋅ A))dτ
dg df = ∫ (∇f ⋅ A)dτ + ∫ f (∇ ⋅ A)dτ
b b
Right ∫ f dx + ∫ g dx
a dx a dx
b df b df = ∫ (∇f ⋅ A)dτ + ∫ ( fA) ⋅ da − ∫ (∇f ⋅ A)dτ
= fg a + ∫ g dx − ∫ g
b
dx
dx dx
∫
= ( f A ) ⋅ da
a a

= fg a
b

not a rigorous prove

49 51

1.4 Curvilinear Coordinates

1.4.1 Spherical Polar Coordinates (II)
1.4.1 Spherical Polar Coordinates (I)
The spherical (polar) coordinates (r, θ, φ) of a point P are The direction of the coordinates: the unit vector r̂, θˆ, φˆ
defined below; They constitute an orthogonal (mutually perpendicular)
r: the distance from the origin (the magnitude of the basis set (just like x
ˆ , yˆ , zˆ ).
position vector).
θ: the angle down from the z-axis (called polar angle). So any vector A can be expressed in terms of them:
φ: The angle around from the x-axis (call the azimuthal
angle). A = Ar rˆ + Aθ θˆ + Aφφˆ
In terms of Cartesian unit vector
 x = r sin θ cos φ

 y = r sin θ sin φ
 z = r cos θ

Murray R Spiegel, Vector Analysis
(McFRAW-Hill, New York, 1989), 6th ed. Chap. 7. 52 53
 1.4.1 Spherical Polar Coordinates (III)
Warning: r̂ , θˆ, φˆ are associated with particular point P,
and they change direction as P moves around.
For example, r̂ always points radially outward, but “radially
outward” can be the x direction, the y direction, or any other
direction, depending on where you are.
Notice: Since the unit vectors are function of position, we
must handle the differential and integral with care.
1. Differentiate a vector that is expressed in spherical
coordinates.
2. Do not take the unit vectors outside an integral.
54
1.4.1 Spherical Polar Coordinates (V)
The vector derivatives in spherical coordinates:
56
1.4.1 Spherical Polar Coordinates (IV)
The general infinitesi mal displacement :
dl = drrˆ + rdθθˆ + r sin θdφφˆ
The infinitesimal surface element da for the surface
of a sphere.
da = (dlθ )(dlφ ) rˆ = r 2 sin θdθdφ rˆ
The infinitesimal volume element dτ
dτ = (dlr )(dlθ )(dlφ ) = r 2 sin θdrdθdφ 55
1.4.2 Cylindrical Coordinates (I)
The cylindrical coordinates (s, φ, z) of a point P are defined
below: x = s cos φ , y = s sin φ , z = z
s: the distance from the z axis.
φ: the same meaning as in spherical coordinates.
z: the same as Cartesian.
The unit vectors are
The infinitesi mal displacement :
dl = dssˆ + sdφφˆ + dzzˆ
57
 1.4.2 Cylindrical Coordinates (II) 1.5 The Dirac Delta Function
1.5.1 The Divergence of rˆ / r
2

The vector derivatives in cylindrical coordinates:

Consider a vector function v = rˆ / r 2
The divergence of this vector function is:
1 ∂ 2 1 1 ∂
∇⋅v = (r 2 ) = 2 (1) = 0
r ∂r
2
r r ∂r
The surface integral of this function is:
π 2π 1 2
∫ v ⋅ da = ∫ 0 ∫
0
(
r2
r sin θ )dθdφ
π 2π
= ∫ sin θdθ ∫ dφ = 4π ≠ ∫ (∇ ⋅ v )dτ
0 0 v

The divergence theorem is false?

58
No Î The Dirac delta function 59

1.5.2 The One-Dimensional Dirac Delta Function 1.5.2 The One-Dimensional Dirac Delta Function (II)
The 1-D Dirac delta function can be pictured as If f(x) is some “ordinary” function (let’s say that it is
an infinitely high, infinitesimally narrow “spike”, with area just 1. continuous), then the product f(x)δ(x) is zero everywhere
except at x=0. It follows that f(x)δ(x)=f(0)δ(x). In particular,
 0 if x ≠ 0 +∞
δ ( x) = 
∞ if x = 0
with ∫-∞
δ ( x)dx = 1
∫
+∞

-∞
+∞
f ( x)δ ( x)dx = f (0) ∫ δ ( x)dx = f (0)
-∞

Technically, δ(x) is not a function at all, since its value is We can shift the spike from x=0 to some other point x=a.
not finite at x=0. Such function is called the generalized
 0 if x ≠ a +∞
function, or distribution. δ ( x − a) = 
∞ if x = a
with ∫
-∞
δ ( x − a )dx = 1

A generalized integration equation:

+∞ +∞

60
∫-∞
f ( x)δ ( x − a )dx = f (a ) ∫ δ ( x)dx = f (a )
-∞
61
1.5.2 The One-Dimensional Dirac Delta Function (III) 1
Example 1.15 Show that δ (kx) = δ ( x)
k
Although δ(x) is not a legitimate function, integrals over δ(x) where k is any (nonzero) constant.
are perfectly acceptable.
Sol : Consider the integral for an arbitrary test function f ( x),
It is best to think of the delta function as something that is ∞
always intended for use under an integral sign. ∫−∞
f ( x)δ (kx)dx
In particular, two expressions involving delta function are Let y ≡ kx, so that x ≡ y k , dx ≡ 1 k dy
considered equal if:
 positive : the integration runs from - ∞ to ∞
+∞ +∞ k =
∫-∞
f ( x) D1 ( x)dx = ∫
-∞
f ( x) D2 ( x)dx negative : the integration runs from ∞ to - ∞
∞ 1 ∞ 1
for all (" ordinary" ) function of f ( x).
∫−∞
f ( x)δ (kx)dx = ± ∫ f ( y / k )δ ( y )dy =
k −∞ k
f ( 0)

3
∫ x 3δ ( x − 2)dx 1
Example 1.14 Evaluate the integral (a)
3 0
So δ (kx) serves the same purpose as δ ( x) and δ (− x) = δ ( x) .
(b) ∫ x 3δ ( x − 4)dx 62
k
63
0

Prob. 1.45 1.5.3 The three-Dimensional Dirac Delta Function

d The generalized 3D delta function

(a) x (δ ( x)) = −δ ( x)
dx δ 3 (r ) = δ ( x)δ ( y )δ ( z )
where r is the position vector. It is zero everywhere
(b) Let θ ( x) be the step function : except at (0,0,0), where it blows up.
1, if x > 0 Its volume integral is:
θ ( x) = 
0, if x < 0 +∞ +∞ +∞

Show that dθ dx = δ ( x)
∫ all space
δ 3 (r )dx = ∫
-∞ ∫ ∫
-∞ -∞
δ ( x)δ ( y )δ ( z )dxdydz = 1

As in the 1-D case, the integral with delta function picks

out the value of the function at the location of the spike.

64
∫all space
f (r )δ 3 (r − a)dx = f (a)
65
 1.6 The Theory of Vector Fields
1.5.3 The three-Dimensional Dirac Delta Function (II)
1.6.1 The Helmholtz Theorem
We found that the divergence of rˆ / r is zero everywhere
2
To what extent is a vector function F determined by its
except at the origin, and yet its integral over any volume divergence and curl?
containing the origin is a constant of 4π. The Dirac delta The divergence of F is a specified scalar function D,
function can be defined as:
rˆ ∇⋅F = D
∇⋅( 2
) = 4πδ 3 (r )
r and the curl of F is a specified vector function C,
More generally, ∇×F = C with ∇ ⋅ (∇ × F ) = ∇ ⋅ C = 0
rˆ
∇ ⋅ ( 2 ) = 4πδ 3 (r )
r Can you determine the function F?
where r is the separation vector r =r-r′. Note that the
differentiation here is with respect to r, while r′ is held Helmholtz theorem guarantees that the field F is uniquely
constant. determined by the divergence and curl with appropriate
1 1 rˆ boundary conditions.
∇ 2 ( ) = ∇ ⋅ (∇( )) = ∇ ⋅ (− 2 ) = −4πδ 3 (r )
r r r 66 67

1.6.2 Potentials (simple example)

Homework #2
If the curl of a vector field (F) vanishes (everywhere), then
F can be written as the gradient of a scalar potential (V):

∇×F = 0 ⇒ F = −∇V Problems: 1.37, 1.39, 1.42, 1.45, 1.48

conventional

If the divergence of a vector field (F) vanishes (everywhere),

then F can be expressed as the curl of a vector potential (A):

∇⋅F = 0 ⇒ F = ∇× A

68 69