Chapter 1 Vector Analysis

Vector Algebra
Vector Operations
Displacements, straight line segments
going from one point to another, have
direction as well as magnitude (length),
Such objects are called vectors: velocity,
acceleration, force and momentum are other examples.

Quantities that have magnitude but no direction are called scalars: examples
include mass, charge, density, and temperature.

The magnitude of a vector A is written |A| or, more simply, A.

−A is a vector with the same magnitude as A but of opposite direction.

Define 4 vector operations: addition and 3 kinds

of multiplication:
(i) Addition of 2 vectors
A + B= B+A commutative
(A + B)+ C = A +(B + C) associative
A − B = A +(− B) subtraction
(ii) Multiplication by a scalar: Multiplication of a vector by a
positive scalar a multiplies the magnitude but leaves the direction
unchanged.
a (A + B)= a A + a B distributive
If a is negative, the direction is reversed.
(iii) Dot product of 2 vectors: A⋅B ≡ A B cos θ a scalar ⇐ scalar product
A⋅B = B⋅A commutative
A⋅(B + C)= A⋅B + A⋅C distributive
Geometrically, A⋅B is the product of A(B) times the projection of B(A)
along A(B).

If A∥B, then A⋅B=AB.

For any vector A, A⋅A=A2.

If A⊥B, then A⋅B=0.

Example 1.1

(iv) Cross product of 2 vectors:

A × B ≡ A B sin θ n^
where n^ is a unit vector ⊥ the plane of A and B
 There are 2 directions ⊥ any plane: “in” and “out.”
The ambiguity is resolved by the right-hand rule: let your fingers point in the
direction of the 1st vector and curl around (via the smaller angle) toward the 2nd;
then your thumb indicates the direction of n ^.

A×B is itself a vector, ie, vector product:

A ×(B + C)= A × B + A ×C distributive
A × B =− B × A anti - commutative
Geometrically, |A×B| is the area of the
parallelogram generated by A and B. n^

If 2 vectors are parallel, their cross product is 0.

A×A =0 for any vector A.

Vector Algebra: Component Form
It is often easier to set up Cartesian coordinates x, y, z and work
with vector components.

Let x^ , y^ , and z^ be unit vectors parallel to the x, y, and z axes.

An arbitrary vector A can be expanded in terms of these
basis vectors: A = A x x^ + A y y^ + A z z^

Ax, Ay, Az, are the “components” of A; geometrically,

they are the projections of A along the 3 coordinate axes.
A x = A⋅x^ , A y = A⋅y^ , A z = A⋅z^
Reformulate the vector operations:
A + B =( A x x^ + A y y^ + A z z^ )+( B x x^ + B y y^ + B z z^ )
=( A x + B x ) x^ +( A y + B y ) y^ + ( A z + B z ) z^
Rule (i): To add vectors, add like components.
a A =(a A x ) x^ +(a A y ) y^ +(a A z ) z^
Rule (ii): To multiply by a scalar, multiply each component.
x^ ⊥ y^ ⊥ z^ ⇒ x⋅ ^ y^ = z^ ⋅z^ = 1 , x⋅
^ x^ = y⋅ ^ y^ = x⋅^ z^ = y⋅^ z^ =0
⇒ A⋅B =( A x x^ + A y y^ + A z z^ )⋅( B x x^ + B y y^ + B z z^ )= A x B x + A y B y + A z B z
Rule (iii): To calculate the dot product, multiply like components, and add.
Let x^ 1 = x^ , x^ 2 = y^ , x^ 3 = z^ , A1 = A x , A 2 = A y , A3 = A z
3

⇒ A = A x x^ + A y y^ + A z z^ = A1 x^ 1 + A2 x^ 2 + A3 x^ 3 = ∑A
k =1
k x^ k

x^ i⋅x^ j = δ i j ⇐ δ i j = 1 , for i = j Kronecker delta

[
0 , for i ≠ j

⇒ A⋅B = ∑ A x^ ⋅∑ B x^ = ∑ ∑ A B
i
i i
j
j j
i j
i j x^ i⋅x^ j
3

= ∑ ∑ A B δ =∑ A B = A
i j
i j i j
i
i i 1 B1 + A 2 B 2 + A 3 B 3

One can get rid of the annoying summation symbol ∑ by allowing

one index up and the same index down to represent the summation.
3

For example, Ai B i = ∑ A B =A
i
i i 1 B 1 + A2 B2 + A3 B 3

This is called the Einstein notation. We will only use it in Chapter 12.
A⋅A = A2x + A2y + A2z ⇒ A = A2x + A 2y + A2z
√
x^ × x^ = y^ × y^ = z^ × z^ = 0 ,
x^ × y^ =− y^ × x^ = z^ , y^ × z^ =− z^ × y^ = x^ , z^ × x^ =− x^ × z^ = y^
⇒ A × B =( A x x^ + A y y^ + A z z^ )×( B x x^ + B y y^ + B z z^ )
x^ y^ z^

Example 1.2
|
=( A y B z − A z B y ) x^ +( A z B x − A x B z ) y^ + ( A x B y − A y B x ) z^ = A x A y A z
Bx By Bz |
 + 1 if (i , j , k ) is (1,2,3) , (2,3,1) , or (3,1,2)

3
[
Levi-Civita symbol: ϵ i j k = −1 if (i , j , k ) is (1,3,2) , (3,2,1) , or (2,1,3)
0 if i = j , or j = k , or k =i

∑ ϵ x^
⇒ x^ i × x^ j =
k =1
i jk k

⇒ A × B = ∑ A x^ × ∑ B i i j x^ j = ∑A B i j x^ i × x^ j
i j i,j

x^ 1 x^ 2 x^ 3
= ∑ A B ∑ϵ
i, j
i j
k
i jk x^ k = ∑ϵ
i , j ,k
i jk Ai B j x^ k = A1
| B1 B2 B3
A2 A3
|
⇒ (A × B)i = ∑ϵj,k
i jk A j Bk

By the way, if M is a 3× 3 matrix, det M =|M|= ∑ϵ

i, j,k
i jk M1i M 2 j M3 k
Triple Products

(i) Scalar triple product: A⋅(B ×C )

Geometrically, |A⋅(B×C)| is the volume of the

parallelepiped generated by A, B, and C, since |B×C|
is the area of the base, and |A cos θ| is the altitude.
Ax A y Az
A⋅(B ×C )= B⋅(C × A)=C⋅( A × B) ⇐ cyclic
A⋅(C × B)= B⋅(A ×C )=C⋅( B × A)=− A⋅( B × C)

(ii) Vector triple product:

⇒ A⋅( B × C)= B x B y B z
Cx Cy Cz | |
The vector triple product can be simplified by the so-called BAC-CAB rule:

A ×(B × C)= B (A⋅C )−C ( A⋅B) (*)

(A × B )×C =−C ×( A × B)=− A (B⋅C )+ B (A⋅C ) entirely different vector

All higher vector products can be similarly reduced, usually by repeated (*)
( A × B)⋅(C × D)=( A⋅C) (B⋅D)−(A⋅D) (B⋅C )
A ×[B ×(C × D)]= B [A⋅(C × D)]−( A⋅B) (C × D )
∑ϵ
k
i jk ϵm n k = δ i m δ j n − δi n δ j m , ∑ϵ j ,k
i jk ϵ ℓ j k = 2 δi ℓ , ∑ϵ
i , j ,k
i jk ϵi j k = 6

 ∑δ k
ik δ j k = δi j , ∑δ k
kk =3

A⋅(B ×C)= ∑ A (B ×C ) = ∑ A ∑ ϵ
i
i i
i
i
j,k
i jk B j Ck

A1 A2 A3
= ∑ϵ
i , j ,k
i jk

C1 C2 C3 |
A i B j C k = B 1 B2 B 3 = B⋅(C × A)= C⋅(A × B )
|
A ×(B ×C )= ∑ ϵ A (B × C) x^ = ∑ ϵ A x^ ∑ ϵ B C
i, j,k
i jk i j k
i, j,k
i jk i k
m,n
jmn m n

= ∑ ϵ ϵ A B C x^ =− ∑ ϵ ϵ
i jk jmn A B C x^
i m n k ik j mn j i m n k
i , j ,k ,m ,n i , j ,k ,m ,n

=− ∑ (δ δ − δ δ ) A B C x^ = ∑ ( A C B x^ − A B C x^ )
im kn in km i m n k i i k k i i k k
i ,k ,m ,n i ,k

=(∑ A C ) ∑ B x^ −( ∑ A B ) ∑ C x^ =( A⋅C) B −(A⋅B) C

i i k k i i k k
i k i k
Position, Displacement, and Separation Vectors
The vector to the location of a point from the origin O
is called the position vector: r = x x^ + y y^ + z z^

Its magnitude, r = √ x 2 + y2 + z 2
r x x^ + y y^ + z z^
and the unit vector is r = =
^
r √ x
2
+ y
2
+ z
2

The infinitesimal displacement vector, from

(x, y, z) to (x+dx, y+dy, z+dz), is
d ℓ = d r = d x x^ + d y y^ + d z z^
In electrodynamics, one frequently encounters problems involving 2 points—
typically, a source point r', where an electric charge is located, and a field
point r, at which you are calculating the electric or magnetic field.

A short-hand for the separation vector from the source point to the field point
г r −r '
г ≡ r −r ' ⇒ г =|r − r '| , г^ = =
г |r − r '|
г =( x − x ' ) x^ + ( y − y ' ) y^ +( z − z ' ) z^
г = √ ( x − x ' )2 +( y − y ' )2 +( z − z ' )2
⇒
( x − x ' ) x^ +( y − y ' ) y^ +( z − z ' ) z^
г^ =
√( x − x ' )2 +( y − y ' )2 + (z − z ' )2
How Vectors Transform
The definition of a vector as “a quantity with a
magnitude and direction” is not satisfactory.

should transform properly when you change

A vector
coordinates.
The coordinate frame we use to describe positions in
space is arbitrary, but there is a specific geometrical transformation law for
converting vector components from one frame to another.

Let the x , y , z system is rotated by angle ϕ, relative to x, y, z, about the

common x = x axes:
A y = A cos θ , A z = A sin θ
A y = A cos θ = A cos (θ −ϕ)= A (cos θ cos ϕ + sin θ sin ϕ)= A y cos ϕ + A z sin ϕ
⇒
A z = A sin θ = A sin (θ − ϕ)= A (sin θ cos ϕ − cos θ sin ϕ)=− A y sin ϕ + A z cos ϕ
Ay Ay
= cos ϕ sin ϕ
⇒
[ ][
Az −sin ϕ cos ϕ ][ ]
Az Ax Rx x Rx y R x z Ax

⇒ Ai =
3

∑
j=1
Ri j A j
[][
For rotation about an arbitrary axis in 3D: A y = R y x R y y R y z
Az R z x R z y Rz z ][ ]
Ay
Az
 Formally, a vector is any set of 3 components that transforms in the same manner
as a displacement when you change coordinates. As always, displacement is the
model for the behavior of all vectors.

A (2nd-rank) tensor is a quantity with 9 components, Tx x, Tx y, Tx z, Ty x, ..., Tz z,

which transform with 2 factors of R:
T x x = Rx x ( Rx x T x x + Rx y T x y + R x z T x z) 3 3

+ Rx y ( Rx x T y x + Rx y T y y + Rx z T y z) ⇒ Ti j=∑∑ R
k =1 ℓ =1
ik Rjℓ Tkℓ
+ Rx z ( Rx x T z x + Rx y T z y + Rx z T z z) , ⋯
In general, an nth-rank tensor has n indices and 3 n components, and transforms
with n factors of R.

A vector is a tensor of rank 1, and a scalar is a tensor of rank 0.

Differential Calculus
“Ordinary” Derivatives
d f
If we have f(x), what does do for us? It tells us how rapidly f(x) varies when
dx
d f
we change x by a tiny amount, dx ⇒ d f = dx
dx
If we increment x by dx, then f changes by df; the derivative is the
proportionality factor.

If f varies slowly with x, and the derivative is correspondingly small. If f

increases rapidly with x, and the derivative is large.
d f
Geometrical Interpretation: The derivative is the slope of the graph of f
versus x.
dx
Gradient
If we have a function of 3 variables, T(x,y,z), we want to generalize the notion
of “derivative” to functions like T, which depend not on one but on 3 variables.

A derivative is supposed to tell us how fast the function varies for a little
distance, and on what direction we move.
∂T ∂T ∂T
dT= d x+ d y+ dz
∂x ∂y ∂z
This tells us how T changes when we alter all 3 variables by dx, dy, dz.
∂T ∂T ∂T
Rewrite d T =( x+
^ y+
^ z^ )⋅(d x x^ + d y y^ + d z z^ )= ∇ T⋅d r
∂x ∂y ∂z
∂T ∂T ∂T
where ∇ T ≡ x^ + y^ + z^ ⇐ gradient of T
∂x ∂y ∂z
∇T is a vector quantity, a generalized derivative, with 3 components.

Geometrical Interpretation of the Gradient

Like any vector, the gradient has magnitude and direction:
d T = ∇ T⋅d r =|∇ T||d r| cos θ
magnitude |dr| and search around in various directions (vary θ), the
If we fix the
maximum change in T evidentally occurs when θ=0 (for then cos θ=1).
 For a fixed |dr|, dT is greatest when moving in the same direction as ∇ T.

The gradient ∇T points in the direction of maximum increase of the function T.

The magnitude |∇T| gives the slope (increase rate) along this maximal direction.
The direction of steepest ascent is the direction of the gradient.

The direction of max descent is opposite to the direction of max ascent, while at
right angles (θ= 90˚) the slope is 0 (the gradient ⊥ the contour lines).

If ∇ T=0 at (x, y, z), then dT=0 for small displacements about the point (x, y, z).
This is, then, a stationary point of T(x, y, z).

It could be an extremum, ie, maximum (a summit), a minimum (a valley), a

saddle point (a pass), or a “shoulder.”

If you want to locate the extrema of a function of 3 variables, set its gradient
equal to 0.

Example 1.3
The Del Operator

) T , the term in parentheses is called del:

∂ ∂ ∂
∇ T =( x^ + y^ + z^
∂x ∂y ∂z
3
∂ ∂ ∂
∇ = x^ + y^ + z^ = x^ ∂ x + y^ ∂ y + z^ ∂ z = ∑ x^ k ∂ k
∂x ∂y ∂z k =1

∇ is not a vector, but a vector operator that acts upon T (a function).

There are 3 ways the operator ∇ can act:

1. On a scalar function T: ∇ T (the gradient);

2. On a vector function v, via the dot product: ∇⋅v (the divergence);

3. On a vector function v, via the cross product: ∇×v (the curl).

The Divergence
3
∂ vx ∂ v y ∂ vz ∂ vk
∇⋅v = ( x^ )⋅(v x x^ + v y y^ + v z z^ )=
∂ ∂ ∂
+ y^ + z^ + + =∑
∂x ∂y ∂z ∂ x ∂ y ∂ z k =1 ∂ xk
The divergence of a vector function v is itself a scalar ∇⋅v.

Geometrical Interpretation: ∇⋅v is a measure of how much the vector v spreads

out (diverges) from the point in question.

The vector function in 1st figure has a large (positive) divergence (if the arrows
pointed in, it would be a negative divergence), the 2nd has 0 divergence, and the
3rd again has a positive divergence.

A point of positive divergence is a source, or “faucet”; a point of negative

divergence is a sink, or “drain.”
v c = z z^
Example 1.4 v b = z^

v a = x x^ + y y^ + z z^
 1
∇⋅A ≡ lim ∮ A⋅d a
Δτ0 Δτ S

∮ A⋅d a = ∫ +∫ +∫ +∫ +∫ +∫
S [ front
face
back
face
right
face
left
face
top
face
bottom
face
] A⋅d a

∫ front
A⋅d a = A front⋅Δ a front = A front⋅Δ y Δ z x^
face face face face

= A x ( x 0 + Δ x /2 , y 0 , z 0 ) Δ y Δ z
Δx Δ x ∂ Ax
Ax ( x0 ± , y 0 , z0 )= A x ( x 0 , y 0 , z 0 )±
2
+ O ((Δ x ) )
2 2 ∂x ( x 0 , y 0 , z0 )

Δx
∫ back A⋅d a = A back⋅Δ a back = A back⋅Δ y Δ z (− x^ )=− A x ( x 0 − , y0 , z 0 ) Δ y Δ z
face face face face 2
∂ Ax
[ ∫ front
face
+∫ back
face
] A⋅d a =
∂x [ 2
+ O (( Δ x ) )
]
( x 0 , y 0 , z0 )
ΔxΔyΔz

∂ Ay
⇒
[ ∫ right
face
+∫ left
face
] A⋅d a =
∂y [
+ O ((Δ y)2 )
]
( x 0 , y0 , z0 )
ΔxΔyΔz

∂ Az
[ ∫ top
face
+∫ bottom
face
] A⋅d a =
[ ∂z
+ O ((Δ z) )
2
]
( x 0 , y 0 , z 0)
ΔxΔyΔz
Δ τ =Δ x Δ y Δ z
3
∂ Ax ∂ A y ∂ Az
⇒ ∮ A⋅d a = ( )
2
+ + Δ τ + ∑ O ((Δ x i ) ) Δ τ
S ∂x ∂y ∂ z (x , y , z ) i=1 0 0 0

∂ Ax ∂ Ay ∂ A z
⇒ ∇⋅A = + + as Δ τ  0 ⇐ Δ x i  0
∂x ∂y ∂z

1
∇⋅A ( x i , y i , z i )= ∮ A⋅d a i ⇒ ∇⋅A ( x i , y i , z i ) d τ i = ∮ A⋅d a i
d τi Si Si

⇒ ∫ ∇⋅A d τ = ∮ A⋅d a
V S
The Curl
x^ y^ z^
∇ ×v =
| |
∂ ∂
∂x ∂y ∂z
∂

v x v y vz
=(
∂ vz ∂ vy
∂y
−
∂z

The curl of a vector function

) x^ + (
∂ vx ∂ vz
∂z
−

v is a vector.
∂x
) y^ + (
∂ vy ∂ v x
∂x
−
∂y
) z^

Geometrical Interpretation: ∇×v is a measure of how much the vector v swirls

around the point in question.

The 3 functions in the above have 0 curl, whereas the functions shown have a
substantial curl, pointing in the z direction (with the right-hand rule).

Example 1.5
v a =− y x^ + x y^ v b = x y^
 n^

1
∇ × A ≡ lim ( n^ ∮ A⋅d ℓ )
Δ a0 Δ a C max
1 1
⇒ (∇ × A)i = lim ∮ A⋅d ℓ ⇒ (∇ × A) x = lim ∮ A⋅d ℓ
Δ a  0 Δ ai ΔyΔz0 Δ y Δ z
sides
i C i 1,2,3,4
Δy
side 1/3: d ℓ =± Δ z z^ ⇒ A⋅d ℓ =± A z ( x 0 , y 0 ± , z0 ) Δ z
2
Δy Δ y ∂ Az
Az ( x 0 , y0 ± , z 0 ) = A z ( x 0 , y 0 , z 0 )± + O ((Δ y )2 )
2 2 ∂ y (x , y , z )0 0 0

Δ y ∂ Az
⇒ ∫
Sides 1/3 [
A⋅d ℓ = A z ( x 0 , y 0 , z 0 )±
2 ∂ y (x 0 , y0 , z 0 ) ]
+O ((Δ y)2 ) (± Δ z)
 ∂ Az

⇒
∫ Sides A⋅d ℓ = +
1&3
∂y[ ( x 0 , y0 , z 0 )
2

]
+ O ((Δ y) ) Δ y Δ z
⇒ ( ∇ × A )x =
∂ Az
∂y
−
∂ Ay
∂z
∂ Ay
∫ Sides
2&4
[
A⋅d ℓ = −
∂ z (x
0 , y0 , z0 ) ]
+ O ((Δ z)2 ) Δ y Δ z

∂ Az ∂ Ay ∂ Ax ∂ Az ∂ Ay ∂ Ax
⇒ ∇ × A = x^ ( − ) + y^ ( − ) + z^ ( − )
∂y ∂z ∂z ∂x ∂x ∂y

⇒ (∇ × A) j⋅d a j = ∮ A⋅d ℓ
Cj

⇒ ∫ ∇ × A⋅d a = ∮ A⋅d ℓ
S C
∇⋅F = 0 , ∇ × F ≠ 0 ∇⋅F ≠ 0 , ∇ × F =0 ∇⋅F = 0 , ∇ × F =0

∇⋅F = 0 , ∇ × F = 0 ∇⋅F = 0 , ∇ × F ≠0 ∇⋅F ≠ 0 , ∇ × F ≠0

Product Rules
d df dg d df
Sum rule: (f + g)= + Multiplying by a constant: ( k f )= k
dx dx dx dx dx
df dg
g −f
d dg df d f dx dx
Product rule: (f g)= f +g Quotient rule: =
dx dx dx dx g g2
Similar relations hold for the vector derivatives:
∇ ( f + g)= ∇ f + ∇ g , ∇⋅(A + B)= ∇⋅A + ∇⋅B , ∇ ×( A + B )= ∇ × A + ∇ × B
∇ (k f )= k ∇ f , ∇⋅( k A)= k ∇⋅A , ∇ ×(k A )= k ∇ × A
2 ways to construct a scalar as the product of 2 functions:
f g (product of 2 scalar functions )
A⋅B (dot product of 2 vector functions )
f A (scalar times vector )
2 ways to make a vector:
A × B (cross product of 2 vectors )
There are 6 product rules, 2 for gradients:

(i) ∇ (f g)= f ∇ g + g ∇ f
(ii) ∇ ( A⋅B)= A ×(∇ × B)+ B ×(∇ × A )+( A⋅∇ ) B +(B⋅∇ ) A
2 for divergences:

(iii) ∇⋅(f A )= ∇ f ⋅A + f ∇⋅A

(iv) ∇⋅(A × B)= B⋅( ∇ × A)− A⋅(∇ × B)
2 for curls:

(v) ∇ ×(f A )= ∇ f × A + f ∇ × A
(vi) ∇ ×(A × B)=(B⋅∇ ) A −( A⋅∇ ) B +( ∇⋅B) A −(∇⋅A ) B
The proofs come straight from the product rule for ordinary derivatives, eg,
∇⋅( f A )=∂ x ( f A x )+ ∂ y ( f A y )+ ∂ z ( f A z )
=( A x ∂ x f + f ∂ x A x )+ ( A y ∂ y f + f ∂ y A y )+( A z ∂ z f + f ∂ z A z )
= ∇ f ⋅A + f ∇⋅A
It is also possible to formulate 3 quotient rules:
f g∇ f−f ∇g A g ∇⋅A − A⋅∇ g A g ∇×A +A×∇ g
∇ = , ∇⋅ = , ∇× =
g g
2
g g
2
g g
2
 2nd Derivatives
By applying ∇ twice, we can construct 5 species of 2nd derivatives.

The gradient ∇T is a vector, so we can take the divergence and curl of it:
(1) Divergence of gradient: ∇⋅∇ T
(2) Curl of gradient: ∇ ×∇ T
The divergence ∇⋅v is a scalar—all we can do is take its gradient:

(3) Gradient of divergence: ∇ ( ∇⋅v )

The curl ∇×v is a vector, so we can take its divergence and curl:

(4) Divergence of curl: ∇⋅(∇ × v )

(5) Curl of curl: ∇ ×( ∇ × v )

2 2 2
2 T T T T T T
)⋅( x^ )=
∇ T =( x^ ∂ ∂ ∂ ∂ ∂ ∂ ∂ ∂ ∂
+ y^ + z^ + y^ + z^ + +
= ∇⋅∇ T ∂x ∂y ∂z ∂x ∂y ∂z ∂x
2
∂y
2
∂z
2

This object, which we write as ∇ T for short, is called the Laplacian of T .

2

2 2 2 3 2 3
2 ∂ + ∂ + ∂ = ∂ = 2
The Laplacian of a scalar is a scalar. ∇ ≡ 2 2 2 ∑ 2 ∑ ∂ k
∂x ∂y ∂z k =1 ∂ x k k =1
The Laplacian of a vector, ∇ 2v: ∇ 2 v ≡( ∇ 2 v x ) x^ +( ∇ 2 v y ) y^ +( ∇ 2 v z ) z^

The curl of a gradient is always 0: ∇ ×∇ T=0

∂ ∂T ∂ ∂T
Its proof hinges on the equality of cross derivatives: =
∂x ∂y ∂y ∂x
∇(∇⋅v) seldom occurs in physical applications, ∇ 2 v =( ∇⋅∇ ) v ≠ ∇ (∇⋅v )

The divergence of a curl, like the curl of a gradient, is always 0: ∇⋅(∇ × v )= 0

∇ ×( ∇ × v )= ∇ ( ∇⋅v )− ∇ 2 v
∂
Proof: Let ∂ i ≡
∂ xi
∇ ×( ∇ × v )= ∑ ϵ i j k x^ i ∂ j ( ∇ × v )k = ∑ ϵ i j k x^ i ∂ j ∑ ϵ k m n ∂ m v n
i, j,k i, j,k m ,n

= ∑ ϵi j k ϵm n k x^ i ∂ j ∂m v n = ∑ (δ i m δ j n −δ i n δ j m ) x^ i ∂ j ∂m v n
i , j ,k ,m ,n i , j ,m ,n

=( ∑ x^ i ∂i ) ( ∑ ∂ j v j )−( ∑ ∂ j ∂ j ) ∑ v i x^ i = ∇ (∇⋅v )− ∇ 2 v
i j j i
Integral Calculus
Line, Surface, and Volume Integrals
In electrodynamics, the most important integral are line
(or path) integrals, surface integrals (or flux), and
volume integrals.
b
Line Integrals:∫ v⋅d ℓ
a
the integral is to be carried out along a path P from point a to point b.

If the path forms a closed loop (ie, if b=a), it can be expressed as:∮ v⋅d ℓ
One example of a line integral is the work done by a force F: W =∫ F⋅d ℓ
Ordinarily, the value of a line integral depends critically on the path, but there
is an important special class of vector functions for which the line integral is
independent of path and is determined entirely by the end points.
Example 1.6
A force that has this property is called conservative.

Surface Integrals:∫ v⋅d a

S
the integral is over a specified surface S.

If the surface is closed:∮ v⋅d a

 For a closed surface, tradition dictates that “outward” is
positive, but for open surfaces it’s arbitrary.

If v describes the flow of a fluid (mass per unit area per

unit time), then∫ vd a represents the total mass per unit
time passing through the surface—hence “flux.”

Ordinarily, 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 line.

Volume Integrals:∫ T d τ ⇐ d τ =d x d y d z
V
Cartesian coordinates
∫ v d τ =∫ (v x x^ + v y y^ + v z z^ ) d τ

= x^ ∫ v x d τ + y^ ∫ v y d τ + z^ ∫ v z d τ
because the unit vectors are constants, they come
outside the integral.

Example 1.7 Example 1.8

The Fundamental Theorem of Calculus
Let f(x) is a function of one variable, the fundamental theorem of calculus:
b b
d f d f
∫ d x = f (b)− f (a) ⇐ ∫ F d x = f (b)− f ( a) ⇐ F =
a dx a dx
d f
Geometrical Interpretation: d f = d x is the infinitesimal change in f when
dx
you go from x to x+dx. The fundamental theorem says that if you chop the
interval from a to b into many tiny pieces, dx, and add up the increments df from
each little piece, the result is equal to the total change in f: f(b)−f(a).

2 ways to determine the total change in the function:

either subtract the values at the ends or go step-by-step,
adding up all the tiny increments as you go. You’ll get
the same answer either way.

So the integral of a derivative over some region is given

by the value of the function at the end points (boundaries).

In vector calculus there are 3 species of derivative

(gradient, divergence, and curl), and each has its own
“fundamental theorem,” with essentially the same format.
The Fundamental Theorem for Gradients
Let we have a scalar function of 3 variables T(x,y,z).
Starting at a, we move a small distance dℓ1. T will
change by an amount d T = ∇ T⋅d ℓ 1

Now we move a little further, by an additional small

displacement dℓ2; the incremental change in T
will be ∇Tdℓ2. By proceeding by infinitesimal steps, we make the journey to b.

The total change in T in going from a to b (along the path selected) is

b
∫ ∇ T⋅d ℓ =T (b )−T (a ) ⇐ fundamental theorem Example 1.9
a for gradients
The integral (a line integral) of a derivative (the gradient)
is given by the value of T at the boundaries (a & b).

Line integrals ordinarily depend on the path from a to b. But the rhs of the eqn
makes no reference to the path—only to the end points.

Gradients have the property that the line integrals are path independent:
b
Corollary 1: ∫ ∇ T⋅d ℓ is independent of the path taken from a to b.
a
Corollary 2: ∮ ∇ T⋅d ℓ =0 , since the beginning and end points are identical,
and hence T(b)−T(a)=0.
The Fundamental Theorem for Divergences
The fundamental theorem for divergences states ∫ ∇⋅v d τ = ∮ v⋅d a
V S
It is called as Gauss’s theorem, Green’s theorem, the divergence theorem.

The integral of a derivative (the divergence) over a region (volume V) is equal to

the value of the function at the boundary (the surface S that bounds the volume).

If v represents the flow of an incompressible fluid, then the flux of v is the total
amount of fluid passing out through the surface, per unit time.

The divergence measures the “spreading out” of the vectors from a point—a
place of high divergence is like a “faucet,” pouring out liquid.

If we have a bunch of faucets in a region filled with

incompressible fluid, an equal amount of liquid will be
forced out through the boundaries of the region.

2 ways to determine how much is being produced:

(a) count up all the faucets, recording how much each
puts out, or
(b) measure the flow at each point of the boundary,
and add it all up:
∫ (faucets within the volume )= ∮ (flow out through the surface ) Example 1.10
The Fundamental Theorem for Curls
The fundamental theorem for curls,
∫ ∇ × v⋅d a = ∮ v⋅d ℓ
also called Stokes’ theorem:
S P

The integral of a derivative (the curl) over

a region (surface S) is equal to the value
of the function at the boundary (the
perimeter of the surface, P).

The curl measures the “twist” of v; a region of high curl is a whirlpool.

The integral of the curl over some surface (the flux of the curl
through that surface) represents the “total amount of swirl,” and we
can determine that by going around the edge and finding how much
the flow is following the boundary.

∮ v⋅d ℓ is sometimes called the circulation of v.

Consistency in Stokes’ theorem is given by the right-hand rule: if your fingers

point in the direction of the line integral, then your thumb is the direction of d a.

Ordinarily, a flux integral depends critically on what surface you integrate over,
but evidently this is not the case with curls.
 Stokes’ theorem says that∫ ∇ × v⋅d a is equal to the line integral of v around

the boundary, and the latter makes no reference to the surface you choose.

Corollary 1:∫ ∇ × v⋅d a depends only on the boundary line, not on the
particular surface used.

Corollary 2:∮ ∇ × v⋅d a = 0 for any closed surface, since the boundary line,
like the mouth of a balloon, shrinks down to a
point, and hence the line integral vanishes.

Example 1.11
 ∫ ∇⋅F d τ = ∮ F⋅d a ∫ ∇ × A⋅d a = ∮ A⋅d ℓ ∫ ∇ ϕ⋅d ℓ = ϕ2 −ϕ 1
volume surface surface curve curve

In general, ∫ d ω= ∫ ω
M ∂M
Integration by Parts
d dg d f
( f g)= f +g
dx dx dx
b b b b
d dg d f
⇒ ∫
a dx
b
a
|
( f g) d x =( f g) = ∫
a
f
dx
d x+ ∫
a
g
dx
dx

b b
dg d f
⇒ ∫
a
f
dx
d x =( f g) − ∫
a
| a
g
dx
dx Example 1.12

It applies to the situation in which you are called upon to integrate the product
of one function (f) and the derivative of another (g); it says you can transfer the
derivative from g to f , at the cost of a minus sign and a boundary term.

∇⋅( f A)= f ∇⋅A + A⋅∇ f

⇒ ∫ ∇⋅( f A ) d τ = ∫ f ∇⋅A d τ + ∫ A⋅∇ f d τ =∮ f A⋅d a

⇒ ∫ f ∇⋅A d τ = ∮ f A⋅d a − ∫ A⋅∇ f d τ

V S V
The integrand is the product f and the derivative (the divergence) of A, and
integration by parts licenses us to transfer the derivative from A to f (a gradient),
at the cost of a minus sign and a boundary term (a surface integral).
Curvilinear Coordinates e^ θ '
y e^ r '
Derivation for a polar coordinate r⃗ ' = r⃗ + Δ r⃗
In a polar coordinate, the unit vectors are
e^ θ
e^ r ( radial direction) and e^ θ (tangential direction )
θ e^ r
The position vector can be written as r⃗ =r e^ r r⃗
d dr d x
and the velocity is u
⃗= r⃗ = e r
^ +r e^ r
dt dt dt
Since the derivatives of the unit vectors in a polar coordinate are
d d dθ e^ r ' Δ e^ r Δ e^ θ
e^ r = ω e^ θ , e^ θ =− ω e^ r ⇐ ω = θ
dt dt dt θ e^ θ e^ θ
e^ r e^ θ '
dr
Therefore, u
⃗ = ṙ e r + r ω e θ ⇐ ṙ =
^ ^
dt
e^ r
With r held constant u ⃗ = r ω e^ θ = v e^ θ
d d d
The acceleration is ⃗ a= u⃗ = r̈ e^ r + ṙ e^ r + ( ṙ ω + r ω̇) e^ θ + r ω e^ θ
dt dt dt
= r̈ e^ r + ṙ ω e^ θ + ( ṙ ω + r ω̇) e^ θ −r ω 2 e^ r
2
v
⇒ ⃗a =( r̈ − ) e^ r + (2 ṙ ω + r α) e^ θ ⇐ α ≡ ω̇ = θ̈
r
v2
a =− e^ r +r α e^ θ = ar e^ r + a t e^ θ
With r held constant ⃗
r
Spherical Coordinates
Sometimes it is more convenient to use spherical
coordinates (r, θ, ϕ) instead of Cartesian coordinates
(x, y, z); r is the distance from the origin, θ is called
the polar angle, and ϕ is the azimuthal angle.

x =r sin θ cos ϕ , y = r sin θ sin ϕ , z =r cos θ

3 unit vectors, r^ , θ ^ , point in the
^, ϕ direction of increase of the
corresponding coordinates. They form an orthogonal (mutually perpendicular)
basis set, and any vector A can be expressed as: A = A r r^ + A θ θ^ + A ϕ ϕ
^

r^ =sin θ cos ϕ x^ + sin θ sin ϕ y^ + cos θ z^

[ θ^ = cos θ cos ϕ x^ + cos θ sin ϕ y^ −sin θ z^

ϕ^ = − sin ϕ x^ + cos ϕ y^

Warning: r^ , θ^ , ϕ
^ are associated with a particular point, and they change
direction as the point moves around (compared with Cartesian coordinates).

One could take account of this by explicitly indicating the point of reference:
r^ (θ , ϕ) , θ^ (θ , ϕ) , ϕ
^ (θ , ϕ)
 2 2 2
r =√x + y + z

[
r^ r

[
x =r sin θ cos ϕ
y = r sin θ sin ϕ ⇒
z = r cos θ
θ = tan− 1

ϕ = tan
−1
√x
y
x
2

z
+y
2
⇒
[][ ]
θ^ = ϕ

ϕ^
r
^ × r^
z^ × r^
sin θ
⇒ r^ ⊥ θ^ ⊥ ϕ
^
 r^ sin θ cos ϕ sin θ sin ϕ cos θ x^
⇒
[][
θ^ = cos θ cos ϕ cos θ sin ϕ −sin θ
ϕ^ −sin ϕ cos ϕ 0 ][ ]
y^
z^
⇒ S^ = R D
^

x^ sin θ cos ϕ cos θ cos ϕ −sin ϕ r^

⇒
[] [
y^ = sin θ sin ϕ cos θ sin ϕ cos ϕ θ^
z^ cos θ − sin θ 0 ϕ^][ ] ^ = R T S^
⇒ D

z^

r^ sin θ θ^ cos θ
∂ r^ ∂ θ^ ∂ϕ^ 
=0 , =0 , =0 ^
ϕ r^ sin θ + θ^ cos θ
∂r ∂r ∂r
∂ r^ ^ ∂ θ^ ∂ ^
ϕ θ
⇒ =θ , =− r^ , =0
∂θ ∂θ ∂θ
∂ r^ ∂ θ^ ^ ∂ϕ^
=ϕ^ sin θ , = ϕ cos θ , =− r^ sin θ − θ^ cos θ
∂ϕ ∂ϕ ∂ϕ
 Do not naively combine the spherical components of vectors associated with
different points. Beware of differentiating a vector that is expressed in spherical
coordinates, since the unit vectors themselves are functions of position. And do
not take r^ , θ ^ outside an integral.
^ , and ϕ

An infinitesimal displacement in the r^ direction is: d ℓr = d r

An infinitesimal element of length in the θ^ direction is: d ℓθ = r d θ
An infinitesimal element of length in the ^ direction is: d ℓ =r sin θ d ϕ
ϕ ϕ

Thus the general infinitesimal displacement is: d ℓ = dr r^ + r d θ θ^ + r sin θ d ϕ ϕ

^

The infinitesimal volume element in spherical coordinates is the product of the

3 infinitesimal displacements: d τ = d ℓ d ℓ d ℓ =r 2 sin θ d r d θ d ϕ
r θ ϕ

Surface elements depend on the orientation of the surface. One has to analyze
the geometry for any given case.

For the surface of a sphere, r=const, whereas θ and ϕ change, so

d a 1 = d ℓ θ d ℓϕ r^ =r 2 sin θ d θ d ϕ r^
 If the surface lies in the xy plane,, so that θ=π/2
while r and ϕ vary, then d a 2 = d ℓ r d ℓ ϕ θ
^ =r d r d ϕ θ^
r ∈ [ 0 , ∞ ) , θ ∈[0 , π ] , ϕ ∈[0 , 2 π ]
Example 1.13

Now I would like to “translate” the vector derivatives

(gradient, divergence, curl, and Laplacian) into r, θ, ϕ notation.
∂T ∂T ∂T
Since ∇ T = x^ + y^ + z^ , one can do it in the hard way by translate
∂x ∂y ∂z
∂ ∂r ∂ ∂θ ∂ ∂ϕ ∂ ∂ ∂
= + + , =⋯ , =⋯
∂x ∂x ∂r ∂x ∂θ ∂x ∂ϕ ∂y ∂z
x^ = x^ ( r^ , θ^ , ϕ)
^ , y^ = y^ ( r^ , θ^ , ϕ)
^ , z^ = z^ ( r^ , θ^ , ϕ)
^ , ⋯

We can do it in an easier way by

ℓ r ≡r , ℓ θ ≡r θ [with r fixed ] , ℓ ϕ ≡(r sin θ ) ϕ [with r , θ fixed ]

⇒ hr = 1 , h θ =r , h ϕ =r sin θ ⇐ h i : metric coefficients
∂T ^ ∂T ^ ∂T ∂ T θ^ ∂ T ^
ϕ ∂T
⇒ ∇ T = r^ +θ +ϕ = r^ + +
∂ ℓr ∂ ℓθ ∂ ℓϕ ∂ r r ∂ θ r sin θ ∂ ϕ
 ^ ∂ ^ T ^ ∂T ^ ∂T
) T =r
Gradient: ∇ T =( r^ ∂ θ ϕ ∂ ∂ θ ϕ
+ + ^ + +
∂ r r ∂ θ r sin θ ∂ ϕ ∂ r r ∂ θ r sin θ ∂ ϕ

Divergence: ∇⋅v = 1 ∂ (r 2 v r )+ 1 ∂ 1 ∂ vϕ
( v θ sin θ )+
r ∂r
2
r sin θ ∂ θ r sin θ ∂ ϕ
1 2
= 2 [∂r (v r r sin θ )+ ∂θ (v θ r sin θ )+ ∂ϕ (v ϕ r )]
r sin θ
r^ ∂ vθ θ^ 1 ∂ vr
Curl: ∇ ×v =
r sin θ [ ∂
∂θ
(v ϕ sin θ)−
∂ϕ ]
+
[ −
r sin θ ∂ ϕ ∂ r
∂
(r v ϕ )
]
r
^ r ^ r sin θ ϕ ^

| |
v θ
^ 1
+
ϕ
r [ ∂
∂r
(r v θ )−
∂
∂θ ]
r
= 2
r sin θ
∂r ∂θ ∂ϕ
v r r v θ r sin θ v ϕ
2
1 T 1 T 1 T
( ) ( )
Laplacian: ∇ T =
2 ∂ 2 ∂ ∂ ∂ ∂
r + sin θ +
r ∂r ∂r
2 2 2 2 2
r sin θ ∂ θ ∂θ r sin θ ∂ ϕ
∇⋅v = ( r^ )⋅(v r r^ + v θ θ^ + v ϕ ϕ)
∂ ^ ∂ ^ ∂ ^
+θ +ϕ
∂ ℓr ∂ ℓθ ∂ ℓϕ
∂ 1 ∂
= r⋅
^ (v r r + v θ θ + v ϕ ϕ)+ θ⋅
^ ^ ^ ^ (v r r^ + v θ θ^ + v ϕ ϕ)
^
∂r r ∂θ
1 ∂
^
+ ϕ⋅ ( v r r^ + v θ θ^ + v ϕ ϕ^)
r sin θ ∂ ϕ
∂ vr ∂ r^ ∂ θ^ ^
)
∂ϕ
= + 0 + 0 + r^ ⋅( v r + vθ + vϕ
∂r ∂r ∂r ∂r
1 ∂ vθ θ^ ∂ r^ ∂ θ^ ^
)
∂ϕ
+ 0 + + 0 + ⋅( v r + vθ + vϕ
r ∂θ r ∂θ ∂θ ∂θ
1 ∂ vϕ ^ ∂ r^ ∂ θ^ ^
)
ϕ ∂ϕ
+ 0 + 0 + + ⋅( v r + vθ + vϕ
r sin θ ∂ ϕ r sin θ ∂ϕ ∂ϕ ∂ϕ
∂ vr 1 ∂ vθ vr 1 ∂ v ϕ v r v θ cos θ
= + + +0+ + + +0
∂r r ∂θ r r sin θ ∂ ϕ r r sin θ
∂ v r 2 v r 1 ∂ v θ v θ cos θ 1 ∂ vϕ
= + + + +
∂r r r ∂ θ r sin θ r sin θ ∂ ϕ
1 ∂ 2 1 ∂ 1 ∂ vϕ
= 2 (r v r )+ (v θ sin θ) +
r ∂r r sin θ ∂ θ r sin θ ∂ ϕ
 For Curvilinear Coordinates

d ℓ = n^ 1 h1 d u 1 + n^ 2 h 2 d u2 + n^ 3 h 3 d u 3 ⇒ d τ = h 1 h2 h 3 d u1 d u 2 d u 3

⇒ d a 1 = n^ 1 h 2 h 3 d u 2 d u 3 , d a 2 = n^ 2 h 3 h1 d u3 d u 1 , d a 3 = n^ 3 h1 h2 d u1 d u 2

3
n^ 1 ∂ n^ 2 ∂ n^ 3 ∂ n^ 1 n^ 2 n^ 3 n^ j ∂
∇= + + = ∂u + ∂u + ∂u = ∑
h 1 ∂ u 1 h 2 ∂ u 2 h 3 ∂ u 3 h1 h2 h3 i =1 h j ∂ u j
1 2 3

∂ u ( A1 h 2 h 3 ) + ∂ u ( h 1 A2 h 3 ) + ∂ u ( h 1 h 2 A3 )
1 2 3
∇⋅A =
h 1 h 2 h3

h1 n^ 1 h 2 n^ 2 h 3 n^ 3
∇ ×A=
1
h 1 h 2 h3
| ∂u 1
∂u
h 1 A1 h 2 A2 h 3 A3
∂u 2
=
1
| ∑
h 1 h2 h 3 i , j , k
3
i jk
ϵ hi n^ i
∂ hk Ak
∂ uj

∂u V ∂u V ∂u V
∇2 V =
1
= ∇⋅∇ V h 1 h2 h3
∂u (
h1 [
h2 h 3 )+ ∂u ( h 1
1
h2
1
h3 ) + ∂u ( h1 h2
h3 2
) 2

3
3

]
Cylindrical Coordinates
In the cylindrical coordinates (s, ϕ, z), ϕ has the
same meaning as in spherical coordinates, and z is
the same as Cartesian; s is the distance to P from
the z axis.

x = s cos ϕ , y = s sin ϕ , z =z

s^ = cos ϕ x^ + sin ϕ y^

[ ϕ^ = −sin ϕ x^ + cos ϕ y^
z^ = z^

d ℓ s = d s , d ℓϕ = s d ϕ , d ℓ z = d z
⇐ h s =1 , h ϕ = s , h z =1
⇒ d ℓ = d s s^ + s d ϕ ϕ ^ + d z z^ , d τ = s d s d ϕ d z

s ∈ [ 0 , ∞ ) , ϕ ∈[0 , 2 π ] , z ∈(− ∞ , ∞)
^ ∂ ∂T ϕ^ ∂T ∂T
) T = s^
Gradient: ∂ ϕ ∂
∇ T =( s^ + + z^ + + z^
∂s s ∂ϕ ∂z ∂s s ∂ϕ ∂z
∂ vϕ ∂ v z 1
Divergence: ∇⋅v = 1 ∂ (s v s ) + 1 + = [∂s ( s v s )+ ∂ϕ v ϕ + ∂ z (s v z )]
s ∂s s ∂ϕ ∂z s
 1 ∂ v z ∂ vϕ ∂ vs ∂ vz 1 ∂ ∂ vs
Curl: ∇ × v =(
s ∂ϕ ∂z
− ) s^ + (
∂z
−
∂s
) ϕ+
^
[
s ∂s
(s v ϕ )−
∂ϕ ]z^

s^ s ϕ^ z^
=
1
s |
∂s ∂ϕ ∂ z
v s s vϕ v z |
2 2
1 T 1 T T
(s )+ 2
Laplacian: ∇ T =
2 ∂ ∂ ∂ ∂
+
s ∂s ∂s s ∂ϕ
2
∂z
2
The Dirac Delta Function
2
The Divergence of r^ /r
v = r^ / r 2 is directed radially outward; it is likely to have a
large positive divergence from it. But
1 ∂ 2 1 1 ∂
∇⋅v = ( r )= (1)= 0 ? (?)
r ∂r r ∂r
2 2 2
r
If we integrate over a sphere of radius R, centered at the origin, the surface
integral is r^ π 2π
∫ v⋅d a = ∫ 2 ⋅R sin θ d θ d ϕ r^ = ∫ sin θ d θ ∫ d ϕ = 4 π ($)
2

R 0 0
But the volume integral, ∫ ∇⋅v d τ = 0, if we are really to believe Eq. (?). Does
this mean that the divergence theorem is false?

The source of the problem is at r=0, where v blows up. It is true that ∇⋅v=0
everywhere except the origin, but right at the origin the situation is complicated.

The surface integral ($) is independent of R; if the divergence theorem is right

(and it is), we should get ∫ ∇⋅v d τ = 4 π for any sphere centered at the origin,
no matter how small. So the entire contribution must come from the point r=0.

Thus, ∇⋅v has the bizarre property that it vanishes everywhere except at one
point, and yet its integral (over any volume containing that point) is 4π.

This is where the Dirac delta function comes in.

The One-Dimensional Dirac Delta Function
The 1d Dirac delta function, δ(x), can be pictured as an 1
infinitely high, infinitesimally narrow “spike,” ie, [δ ( x)]=
∞ L
δ ( x )= 0 if x ≠0 and ∫ δ ( x ) d x =1
[ ∞ if x =0 −∞
δ(x) is not a function at all, since its value is not finite at
x=0; so it is a generalized function, or distribution.

It is the limit of a sequence of functions, such as rectangles Rn(x), of height

n and width 1/n, or isosceles triangles
Tn(x), of height n and base 2/n.

If f(x) is some “ordinary” function

thus continuous, then the product
f(x)δ(x) is 0 everywhere except
at x=0,
∞ ∞
⇒ f ( x) δ ( x )= f (0) δ ( x ) ⇒ ∫ f ( x ) δ ( x ) d x = f (0) ∫ δ ( x) d x = f (0)
−∞ −∞
Under an integral, the delta function “picks out” the value of f(x) at x=0.

The integral need not run from −∞ to +∞; it is sufficient that the domain
extend across the delta function, and −ϵ to +ϵ would do as well.
 we can shift the spike from x=0 to some
other point, x=a:
∞
δ ( x − a)= 0 if x ≠ a with ∫ δ ( x − a) d x =1
[
∞ if x = a −∞

⇒ f ( x) δ ( x − a)= f (a) δ ( x − a)
∞
⇒ ∫ f ( x ) δ ( x − a) d x = f (a)
−∞

Although δ isn’t a legitimate function, integrals over δ are perfectly acceptable.

It is best to think of the delta function as something that is always intended for
use under an integral sign.

In particular, 2 expressions involving delta functions (say, D1(x) and D2(x)) are
considered equal if, for all (“ordinary”) functions f(x),
∞ ∞
∫ f ( x ) D 1 ( x) d x = ∫ f ( x ) D 2 ( x) d x
−∞ −∞

Example 1.14, Example1.15

The Three-Dimensional Delta Function
It is easy to generalize the delta function to 3d: δ 3 (r )= δ ( x ) δ ( y ) δ ( z)
This 3d delta function is 0 everywhere except at (0, 0, 0), where it blows up. Its
volume integral is 1:
∞ ∞ ∞
∫ δ (r) d τ = ∫ ∫ ∫
3
δ ( x ) δ ( y) δ ( z) d x d y d z =1
all space −∞ −∞ −∞
1
⇒ ∫
3 3
f (r ) δ (r − a ) d τ = f (a ) ⇐ [δ (r )]= 3
all space L
Integration with δ picks out the value of f at the location of the spike.
r^
The divergence of 2
is 0 everywhere except at the origin, and yet its integral
r
over any volume containing the origin is a constant (4π). These are precisely the
defining conditions for the Dirac delta function;
r^ 3 3 г^
⇒ ∇⋅ 2
= 4 π δ (r ) ⇒ ∇⋅ 2
= 4 π δ (г ) ⇐ г = r − r '
r г
2 1 3 1 г 2 1
⇒ ∇ =− 4 π δ (г ) ⇐ ∇ =− 2 ⇒ ∇ =− 4 π δ 3 (r − r ' )
г г г |r − r '|
Example 1.16
Useful in solving the problems with various boundaries.

q
V (∞)  0 ⇐ boundary condition

q
⇒ V (r )= ⇒ ∇ 2 V (r )=− 4 π q δ 3 (r )
r

V ' ( R)=V 0 ⇐ boundary condition

q
V ' (r)= V (r )+ F (r ) ⇐ ∇ 2 F (r )= 0

⇒ ∇ 2 V ' (r )=− 4 π q δ 3 (r )

In general, ∇ 2 G (r , r ' )=− 4 π q δ 3 (r − r ' ) ⇐ G : Green function

The Theory of Vector Fields
The Helmholtz Theorem
Maxwell reduced the entire EM theory to 4 equations, specifying respectively
the divergence and the curl of the electric field E and the magnetic field B.

To what extent is a vector function determined by its divergence and curl? Let
∇ ⋅ F=D ⇒ Can F be determined?
∇ × F =C ⇐ ∇⋅C = 0
To solve a differential equation you must also be supplied with appropriate
boundary conditions.

In electrodynamics we typically require that the fields go to 0 “at infinity”.

With the extra information, the Helmholtz theorem guarantees that the field
is uniquely determined by its divergence and curl.
Potentials
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

Theorem 1
Curl-less (or “irrotational”) fields. The following conditions are equivalent
(that is, F satisfies one if and only if it satisfies all the others):

(a) ∇ × F =0 everywhere.
b
(b) ∫ F⋅d ℓ is independent of path, for any given end points.
a

(c) ∮ F⋅d ℓ = 0 for any closed loop.

(d) F is the gradient of some scalar function: F =− ∇ V .

The potential is not unique—any constant can be added to V with impunity,

since this will not affect its gradient.

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
Theorem 2
Divergence-less (or “solenoidal”) fields. The following conditions are
equivalent:

(a) ∇⋅F = 0 everywhere.

(b) ∫ F⋅d a is independent of surface, for any given boundary line.

(c) ∮ F⋅d a = 0 for any closed surface.

(d) F is the curl of some vector function: F = ∇ × A .

The vector potential is not unique—the gradient of any scalar function can be
added to A without affecting the curl, since the curl of a gradient is 0.

In all cases (whatever its curl and divergence may be) a vector field F can be
written as the gradient of a scalar plus the curl of a vector:

F =− ∇ V + ∇ × A + C (always) ⇐ C = constant vector

Selected problems: 6, 8, 13, 43, 47, 56