Contents

1 Introduction 3

1.1 Mathematical Tools . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.1.1 Coordinate systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.1.1.1 Cylindrical Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.1.1.2 Spherical Polar Coordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

1.1.1.3 Differential operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

1.1.1.4 Theorems involving integrals of differential operators . . . . . . . . . . . . . . . 15

1.1.2 Overview and Revision of Vector Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

1.1.2.1 Fields . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

1.1.2.2 Electrostatics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

1.1.2.3 Magnetostatics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

1.1.2.4 Electromagnetism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

1
PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

PHAS0038: Term 1 Introduction 2

Chapter 1

Introduction
1.1 Mathematical Tools

The understanding and practice of mathematical tools - particularly vector algebra and operators - is vital to

understanding electromagnetic theory.

1.1.1 Coordinate systems

Along with the Cartesian coordinate system (x, y, z), we will be using both the cylindrical coordinate system

and the spherical polar coordinate system. Note that throughout the course we will use symbols such as î and

ei to denote unit vectors. (In the classroom, these may be written on the whiteboard using a tilde underneath,

e.g. i , ei ). The cylindrical and spherical polar coordinate systems are summarized here:
∼ ∼

1.1.1.1 Cylindrical Coordinates

The following relations hold for transformation between the illustrated Cartesian and cylindrical coordinates:

x &= R \cos \phi \\ y &= R \sin \phi \\ z &= z \\ R &= (x^2+y^2)^{1/2} \\ \tan \phi &= {y\over x}

(1.1e)

3
PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

Figure 1.1: Cylindrical coordinate system (R, φ, z). R is a perpendicular distance, measured from the z axis
to the point of interest, P. φ is the azimuthal angle associated with the position P, measured in the xy plane. z
is the ‘vertical displacement’ of P above (positive) or below (negative) the xy plane.

and, for an infinitesimal volume,

\delta \Vol = R \, \delta R \, \delta \phi \, \delta z (1.2)

1.1.1.2 Spherical Polar Coordinates

The following relations hold for spherical coordinates:

x &= r \sin \theta \cos \phi \\ y &= r \sin \theta \sin \phi \\ z &= r \cos \theta \\ r &= (x^2+y^2+z^2)^{1/2} \\ \tan \phi &= {y\over x} \\ \cos \theta &= {z\over (x^2+y^2+z^2)^{1/2}} = {z\over r}

(1.3f)

and, for an infinitesimal volume,

\delta \Vol = (\delta r) \times (r\delta \theta ) \times (r\sin \theta \delta \phi ) = r^2 \, \sin \theta \, \delta r \, \delta \theta \, \delta \phi (1.4)

PHAS0038: Term 1 Introduction 4

PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

Figure 1.2: Cylindrical coordinate system unit vectors (eR , eφ , ez ). Note that each unit vector points in the
local direction along which each coordinate increases most rapidly. Thus, for example, eR points in the direction
of increasing R when keeping φ and z fixed.

1.1.1.3 Differential operators

Meaning and Representation We will become familiar with ‘formulae’ for differential operators in different

coordinate systems. However, it is worth spending some time and thought regarding how they are used, in their

invariant form, to tell us something about how the scalar and vector fields of interest to us change from point to

point in space.

• grad (gradient) or ∇ Consider firstly a scalar function of position in three-dimensional space denoted

f (l). This is a mapping between position l and the value of some scalar quantity at that point in space

- for example, temperature, density, electric potential. The ‘grad’ (‘gradient’) operator, represented by

the symbol ∇, is defined such that, if we consider moving through an infinitesmally small increment in

position, from l to l + δl, then the value of the function f changes by an amount:

\delta f = \nabla f \, \cdot \, \delta \boldsymbol {l} (1.5)

(· denotes dot or scalar product of two vectors, and the vector δl is a displacement in position, whose

components have units of physical distance).

This is an important definition. Without yet mentioning what ∇ looks like in a specific coordinate system,

PHAS0038: Term 1 Introduction 5

PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

Figure 1.3: Spherical polar coordinate system: r, θ, φ. Note that, this time, r represents a radial distance,
measured from the origin O. θ, the ‘polar angle’, is the angle between the position vector of the point of interest
and the z axis. φ is an azimuthal angle, measured in the xy plane, and is the angle between the x axis and the
projection of the position vector of P onto the xy plane.

we can note the following important points:

→ The ∇ operator acts on a scalar function, and returns a vector.

→ the magnitude of the vector ∇f determines how rapidly the value of the function f changes as we move

from point to point in space.

→ at any point in space, the vector ∇f points in the direction of maximum increase of the function f .

(Think about what direction δl must have to give the dot product its maximum possible value).

→ The invariant definition is itself a clue as to how to calculate ∇ in your favourite coordinate

system. For example, in Cartesian coordinates (x, y, z), the increment in function value for infinitestimal

change in coordinates is, from calculus:

\delta f = {\partial f\over \partial x}\, \delta x + {\partial f\over \partial y}\, \delta y + {\partial f\over \partial z}\, \delta z. \notag

Since δl = δx x̂ + δy ŷ + δz ẑ in the Cartesian system, we can simply use the basic defining equality

δf = ∇f · δl to read off the components of ∇f :

\nabla f = {\partial f\over \partial x}\, \boldsymbol {\hat {x}} + {\partial f\over \partial y}\, \boldsymbol {\hat {y}} + {\partial f\over \partial z}\, \boldsymbol {\hat {z}} . \notag

PHAS0038: Term 1 Introduction 6

PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

Figure 1.4: Spherical polar coordinate system unit vectors (er , eθ , eφ ). As before, each unit vector points in
the direction of increasing value of the corresponding coordinate, keeping the other coordinates fixed. In the
figure above, the red curve is part of a circle of constant θ lying on a sphere of radius r, which is centred on the
origin and passes through P. Similarly, the blue curve is part of a circle of constant φ on the same sphere.

If we decide to work in, for example, spherical polar coordinates (r, θ, φ), however, we would note that:

\delta f = {\partial f\over \partial r}\, {\delta r} + {\partial f\over \partial \theta }\, {\delta \theta } + {\partial f\over \partial \phi }\, {\delta \phi }. \notag

And for given infinitesimal changes in the spherical polar coordinates, the corresponding displacement in

position (distance) is:

\delta \boldsymbol {l} = \delta r \, {\boldsymbol {\hat {e_r}}} + r \delta \theta \, {\boldsymbol {\hat {e_\theta }}} + r \, \sin \theta \, \delta \phi \, {\boldsymbol {\hat {e_\phi }}} \notag

So, to conform with the defining equality δf = ∇f · δl, the required form of ∇ must be:

\nabla f = {\partial f\over \partial r}\, {\boldsymbol {\hat {e_r}}} + {1\over r}{\partial f\over \partial \theta }\, {\boldsymbol {\hat {e_\theta }}} + {1\over r \sin \theta }{\partial f\over \partial \phi }\, {\boldsymbol {\hat {e_\phi }}}. \notag

We can use this type of reasoning to derive the form of ∇ in a general curvilinear coordinate system defined

by a set of coordinate directions (unit vectors) at each point in space.

→ Note that, regardless of what coordinate system is used to define the function, the result of calculating

∇f always gives a vector with the same magnitude and direction. For example, consider the function:

f(x,y,z) = x^2 + y^2 + z^2. \notag

An alternative form in spherical polar coordinates, giving the same function value everywhere in space,

would be:

f(r,\theta ,\phi ) = r^2. \notag

PHAS0038: Term 1 Introduction 7

PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

We see that, in the Cartesian form:

∂f ∂f ∂f
∇f = x̂ + ŷ + ẑ
∂x ∂y ∂z

= 2x x̂ + 2y ŷ + 2z ẑ
x y z
= 2r x̂ + ŷ + ẑ
r r r

= 2reˆr ,

where we have used the definition of squared radial distance r2 = x2 + y 2 + z 2 , and the equality eˆr =

x y z
r x̂ + r ŷ + r ẑ (both sides involve vectors of unit length, pointing in the radial direction i.e. from the origin

to the point in question). The final line is simply ∇ in spherical polar form for the function we have defined

in this simple example, which depends only on r.

• curl or ∇× Curl acts on a vector function, and also produces a vector result. An invariant definition of

∇× involves the line integral of a vector function of spatial position, F (l), around the perimeter (bounding

curve Γ) of a flat surface element in space, ∆S. Specifically, ∇× may be defined by the following equality

in the limit as the surface area ∆S approaches zero.

\nabla \times \boldsymbol {F} \cdot \boldsymbol {\hat {n}} = \lim _{{\Delta S }\to 0} {1\over \Delta S} \, \oint _{\Gamma } \boldsymbol {F} \cdot \boldsymbol {dl}. (1.6)

This definition actually gives us the component of ∇ × F in the direction specified by the unit vector n̂. n̂

is orthogonal to the plane defined by the surface element ∆S, and is directed in a right-handed sense with

respect to the direction of integration in which we move around Γ. That is, if the fingers of a right hand

curl around in the direction of integration, then the thumb will point in the direction of n̂.

We will list coordinate-specific forms of ∇× later. For now, we illustrate the use of our definition by

deriving the z component, in a Cartesian system, of a vector function F (x, y, z). For simplicity, we choose

a rectangular surface element of dimensions 2∆x × 2∆y in the xy plane, and we choose n̂ = ẑ, the z

direction. A diagram of our surface element and corresponding direction of integration is shown below.

As the surface element becomes infinitesimally small, the value of the line integral approaches the following

sum, involving the values of the relevant component of F at the midpoints A, B, C, D, shown in the figure:

\oint _{ABCD} \boldsymbol {F} \boldsymbol {dl} \approx F_{x,A} (2 \, \Delta x) + F_{y,B} (2 \, \Delta y) - F_{x,C} (2 \, \Delta x)- F_{y,D} (2 \, \Delta y). \notag

PHAS0038: Term 1 Introduction 8

PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

C(xP, yP + ∆#)

P(xP, yP)
2 ∆# D(xP − ∆", yP) B(xP+∆", yP)

A(xP, yP − ∆#)

2 ∆"

Figure 1.5: Surface element lying in the xy Cartesian plane. The z direction points orthogonally out of the
plane of the page. The path of integration, as shown by arrowheads, is directed (in this view) anticlockwise
around the perimeter, tracing linear segments with midpoints A, B, C and D. The surface element decreases
towards zero, always containing the point of evaluation P, situated at coordinates (xP , yP , zP ). The x and y
coordinates of all points are shown.

In this ‘small size’ limit, we can use Taylor expansion to first order in the coordinate increments in order

to evaluate the components of F on the perimeter by adding appropriate terms to their evaluation at the

point of interest, P. Hence:

I
∂Fx ∂Fy
F dl → (Fx,P − ∆y) (2 ∆x) + (Fy,P + ∆x) (2 ∆y)
ABCD ∂y ∂x
∂Fx ∂Fy
−(Fx,P + ∆y) (2 ∆x) − (Fy,P − ∆x)(2 ∆y).
∂y ∂x
∂Fx ∂Fy
= (−2 +2 ) (2∆x ∆y).
∂y ∂x

Dividing this expression by the surface area 4∆x ∆y, we finally obtain:

\nabla \times \boldsymbol {F} \, \cdot \, \boldsymbol {\hat {z}} = \left ({\partial F_y\over \partial x} - {\partial F_x\over \partial y}\right ). \notag

Analogous expressions exist for the x and y components of curl. This derivation can be generalized to

other coordinate systems. An important point is related to line integrals around the bounding curves of a

surface of arbitrary shape and size. One can picture a large, ‘wavy’ surface as a collection of very many

infinitesimal surface elements. Some thought reveals that the line integral of a vector function around the

bounding curve Γ of this general surface S is equal to the sum of the tiny ‘loop’ integrals around each

of the constituent, infinitesimal surface elements (think of how adjacent edges of the tiny ‘loop’ integrals

PHAS0038: Term 1 Introduction 9

PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

would cancel each other, except at the very edge (perimeter) of the surface) - see diagram below:

dS
!

Figure 1.6: A general bounded surface S can be considered as the superposition of many infinitesimally small
surface elements dS. The arrows show the directions of integration of some vector function F within each element
‘loop’. The sum of all of these is simply the line integral of F around the perimeter Γ of the surface - this is
because the contributions to the line integrals of neighbouring loops are equal in magnitude and opposite in sign
for edges which they share.

Since, by definition, each tiny ‘loop’ integral equals the local value of curl multiplied by the area of the local

tiny loop, we have the following general result, Stokes’ Theorem – which equates the surface integral of the

curl of a vector function to the line integral of that same function around the boundary of the surface:

Stokes’ Theorem \iint _{S} \nabla \times \boldsymbol {F} \cdot \boldsymbol {dS} = \oint _{\Gamma } \boldsymbol {F} \cdot \boldsymbol {dl}.
(1.7)

• div (divergence) or ∇· Like curl, the div operator acts on a vector function. Unlike curl, it returns

a scalar value. This value is related to the surface integral, or flux , of the vector field through a closed

volume ∆V in space, bounded by a surface ∆S. An invariant definition for div involves the value of this

surface integral in the limit that the volume approaches zero.

Let’s take the simple example of a small cube in Cartesian space (see Figure 1.7), whose edges are all parallel

with one of the x, y or z directions. The length of each edge is 2∆l and the coordinates of the centre of the

cube, at point P, are (xP , yP , zP ). The six midpoints of each face have the following coordinates (some

shown in the diagram):

A: (xP , yP + ∆l, zP ), B: (xP , yP − ∆l, zP )

C: (xP + ∆l, yP , zP ), D: (xP − ∆l, yP , zP )

E: (xP , yP , zP + ∆l), F: (xP , yP , zP − ∆l)

PHAS0038: Term 1 Introduction 10

PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

E(xP, yP, zP+∆")

B(xP , yP-∆", zP)

2∆" C(xP+∆", yP, zP)

(
'

Figure 1.7: Small cube in Cartesian space. The cube edge length is 2∆l. The coordinates of the visible central
points of three of the faces are indicated. The arrows show directions orthogonal to the visible faces, pointing
outwards from the cube, as required for the surface integral.

The corresponding unit vectors orthogonal to each face of the cube, and pointing outwards from P, are:

nˆA = ŷ, nˆB = −ŷ

nˆC = x̂, nˆD = −x̂

nˆE = ẑ, nˆF = −ẑ

Thus for an infinitesimal cube, the closed surface integral of the vector function G over the faces of the

cube is well approximated by:

ZZ
G · n̂dS
S(CU BE)

→ (4∆l2 ) (Gy,A − Gy,B + Gx,C − Gx,D + Gz,E − Gz,F )

= (8∆l3 ) ((Gy,A − Gy,B )/(2 ∆l) + (Gx,C − Gx,D )/(2 ∆l) + (Gz,E − Gz,F )/(2 ∆l))

We note that the quantity (∆V = 8∆l3 ) is the volume of the cube. The divergence of the vector field is

the value of the closed surface integral divided by ∆V, in the limit as the cubic volume approaches zero. In

this limit, we can replace some of the terms involving differences in the vector components by derivatives.

Hence:

ZZ
1
∇ · G = lim G · n̂dS = (1.8)
∆V→0 ∆V S(CU BE)
∂Gx ∂Gy ∂Gz
+ + .
∂x ∂y ∂z

PHAS0038: Term 1 Introduction 11

PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

Analogous formulae for div could be derived for other coordinate systems using the invariant definition as

a starting point.

In a similar way to which we informally demonstrated Stokes’ Theorem above, we could also imagine some

arbitrary volume V, bounded by a closed surface S, as consisting of many infinitesimal volume elements

such as the ‘cube’ in Figure 1.7. The surface integral of a vector function G through the surface of each

volume element could be determined, and all the results added. Since the contributions to this integral

would have equal magnitude and opposite sign at the ‘faces’ shared by neighbouring cubes, the result we

obtain is that the sum of all the ‘cube surface integrals’ is simply equal to the surface integral of this vector

function through the bounding surface S. Now, we know that the value of each cube surface integral (in

the limit of small cube size) is just the local value of div multiplied by the volume of that cube. Thus, in

the limit of infinitesimal cube size, we have:

Divergence Theorem \iiint _{\Vol } \nabla \cdot \boldsymbol {G} \, {d\Vol } = \iint _{S} \boldsymbol {G}\cdot \boldsymbol {n}\,dS.
(1.9)

PHAS0038: Term 1 Introduction 12

PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

Implementations of Operators in Different Coordinate Systems

Differential operators transform vectors and scalars as follows (note that r is used here to represent spatial

position):

\mbox {Grad $\bm {\nabla }$ scalar to vector } \rightarrow \, &\bm {\nabla } \varphi (\bm {r}) \\ &=\bigl ( \frac {\partial \varphi }{\partial x} \bm {\hat {i}} + \frac {\partial \varphi }{\partial y} \bm {\hat {j}} + \frac {\partial \varphi }{\partial z} \bm {\hat {k}} \bigr ) \\ \mbox {Div $\bm {\nabla \cdot }$ vector to scalar } \rightarrow \, & \bm {\nabla \cdot } \bm {F} (\bm {r}) \\ &= \frac {\partial F_x}{\partial x} + \frac {\partial F_y}{\partial y} + \frac {\partial F_z}{\partial z} \\ \mbox {Curl $\bm {\nabla \times }$ vector to vector } \rightarrow \, &\bm {\nabla } \times \bm {F} (\bm {r}) \\ &= \Bigg | \begin {matrix} \bm {\hat {i}} & \bm {\hat {j}} & \bm {\hat {k}} \\ \frac {\partial }{\partial x} & \frac {\partial }{\partial y} & \frac {\partial }{\partial z} \\ F_x & F_y & F_z \end {matrix} \Bigg | \\ \mbox {Laplacian: scalar to scalar} \rightarrow \nabla ^2 \varphi &=\biggl ( \frac {\partial ^2 \varphi }{\partial x^2} + \frac {\partial ^2 \varphi }{\partial y^2} + \frac {\partial ^2 \varphi }{\partial z^2} \biggr ) \\ \mbox {(Laplacian can also be vector to vector)}

(1.10h)

Note: in subsequent parts, the ‘hat’ on the unit vectors may be omitted. We list below, for convenience, the

explicit forms of div, grad and curl in different coordinate systems:

• Cartesian: (x, y, z), Volume element dV = dx dy dz

\bm { \nabla }\varphi &= \bm {i}\frac {\partial \varphi }{\partial x}+ \bm {j}\frac {\partial \varphi }{\partial y}+ \bm {k}\frac {\partial \varphi }{\partial z} \\ \bm {\nabla } \cdot \bm {F} &= \frac {\partial F_x}{\partial x}+ \frac {\partial F_y}{\partial y}+ \frac {\partial F_z}{\partial z} \\ \bm {\nabla } \times \bm {F} &= \bigg | \begin {matrix} \bm {i} & \bm {j} & \bm {k} \\ \frac {\partial }{\partial x} & \frac {\partial }{\partial y} & \frac {\partial }{\partial z} \\ F_x & F_y & F_z \end {matrix} \bigg | \\ \bm {\nabla } \cdot \bm {\nabla } \varphi = \nabla ^2 \varphi &= \frac {\partial ^2\varphi }{\partial x^2}+ \frac {\partial ^2\varphi }{\partial y^2}+ \frac {\partial ^2\varphi }{\partial z^2}

(1.11d)

PHAS0038: Term 1 Introduction 13

PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

• Cylindrical polar: (R, z, φ), Volume element dV = R dR dz dφ

\bm { \nabla }\varphi &= \bm {e}_R\frac {\partial \varphi }{\partial R}+ \bm {e}_{\phi }\frac {1}{R}\frac {\partial \varphi }{\partial \phi }+ \bm {e}_z\frac {\partial \varphi }{\partial z} \\ \bm {\nabla } \cdot \bm {F} &= \frac {1}{R}\frac {\partial (R F_R)}{\partial R}+ \frac {1}{R}\frac {\partial F_{\phi }}{\partial \phi }+ \frac {\partial F_z}{\partial z} \\ \bm {\nabla } \times \bm {F} &= \frac {1}{R} \bigg | \begin {matrix} \bm {e}_R & R\bm {e}_{\phi } & \bm {e}_z \\ \frac {\partial }{\partial R} & \frac {\partial }{\partial \phi } & \frac {\partial }{\partial z} \\ F_R & R F_{\phi } & F_z \end {matrix} \bigg | \\ \nabla ^2 \varphi &= \frac {1}{R}\frac {\partial }{\partial R}\biggl ( R \frac {\partial \varphi }{\partial R}\biggr )+ \frac {1}{R^2}\frac {\partial ^2 \varphi }{\partial \phi ^2}+ \frac {\partial ^2 \varphi }{\partial z^2}

(1.12d)

• Spherical polar: (r, θ, φ), Volume element dV = r2 sin θ dr dθ dφ

\bm { \nabla }\varphi &= \bm {e}_r\frac {\partial \varphi }{\partial r}+ \bm {e}_{\theta }\frac {1}{r}\frac {\partial \varphi }{\partial \theta }+ \bm {e}_{\phi }\frac {1}{r\sin \theta }\frac {\partial \varphi }{\partial \phi } \\ \bm {\nabla } \cdot \bm {F} &= \frac {1}{r^2}\frac {\partial (r^2F_r)}{\partial r}+ \frac {1}{r\sin \theta }\frac {\partial (\sin \theta F_{\theta })}{\partial \theta }+ \frac {1}{r\sin \theta }\frac {\partial F_{\phi }}{\partial \phi } \\ \bm {\nabla } \times \bm {F} &= \frac {1}{r^2 \sin \theta } \bigg | \begin {matrix} \bm {e}_r & r\bm {e}_{\theta } & r\sin \theta \bm {e}_{\phi } \\ \frac {\partial }{\partial r} & \frac {\partial }{\partial \theta } & \frac {\partial }{\partial \phi } \\ F_r & rF_{\theta } & r\sin \theta F_{\phi } \end {matrix} \bigg | \\ \nabla ^2 \varphi &= \frac {1}{r^2}\frac {\partial }{\partial r}\biggl ( r^2 \frac {\partial \varphi }{\partial r}\biggr )+ \frac {1}{r^2\sin \theta }\frac {\partial }{\partial \theta }\biggl ( \sin \theta \frac {\partial \varphi }{\partial \theta }\biggr )+ \frac {1}{r^2\sin ^2 \theta }\frac {\partial ^2 \varphi }{\partial \phi ^2} \\ \frac {1}{r^2}\frac {\partial }{\partial r}\biggl ( r^2 \frac {\partial \varphi }{\partial r}\biggr ) &= \frac {1}{r}\frac {\partial ^2 (r\varphi )}{\partial r^2} = \frac {\partial ^2 \varphi }{\partial r^2}+ \frac {2}{r}\frac {\partial \varphi }{\partial r}

(1.13e)

These should be hopefully familiar from previous courses in the first and second years.

Some useful equalities / identities involving differential operators:

\int _\Vol \bm {\nabla } \cdot \bm {F} \, d\Vol = \oint _S \bm { F } \cdot \bm {n} dS
(1.15)

• Stokes’ Theorem: For a surface S bounded by a closed curve C:

\int _S \bm { \nabla } \times \bm {F} \cdot \bm {n} dS =\oint _C \bm {F} \cdot d\bm {l}
(1.16)

• Line Integral Evaluation: It’s useful to understand how a line integral works by considering the basic

definition in terms of small steps, going from a point ‘a’ to another point ‘b’ along the curve of integration:

{\int _a ^b} _C \bm {F} \cdot d\bm {l} = \displaystyle {\lim _{N \to \infty }} \sum _{i=1} ^N \bm {F}_i \cdot d\bm {l}_i ,
(1.17)

PHAS0038: Term 1 Introduction 15

PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

where C is the curve we’re integrating along, which we are approximating as a series of small tangential linear

segments dli . In other words, at each segment along the curve we take the dot product between the local value

of the vector function we’re integrating and the vector of length |dli | which is tangent to the curve. We then

sum over all these dot products, which are evaluated at at every one of the N segments. It should be easy to see

that, in general, the value of the line integral will depend on the curve chosen.

There are two standard ways of working out a Cartesian line integral in two dimensions. Consider a vector

function F (x, y) = Fx (x, y)i + Fy (x, y)j along some curve defined by g(x, y) = k, where k is a constant. Along

the curve:

∂g ∂g dy ∂g ∂g
dg = 0 = dx + dy → =−
∂x ∂y dx ∂x ∂y

then we can replace every occurrence of y and dy in the integral below with some functions of x (found from

g(x, y) and its derivatives), and integrate:

\int _C \bm {F} \cdot d\bm {l} = \int _C [F_x(x,y)\,dx + F_y(x,y)\,dy] = \int _x \left [F_x (x, y(x))\,dx + F_y(x, y(x))\frac {dy}{dx}dx\right ].
(1.18)

The second way is using a parametric form. This is possible if the curve being used for the integral is given in

terms of a parameter (e.g. angle around a circle). Then we have a curve l(t) = x(t)i + y(t)j which depends on

a single parameter t. So we write:

\int _C \bm {F} \cdot d\bm {l} = \int _a ^b \bm {F}(\bm {l}(t)) \cdot \frac {d\bm {l}}{dt}dt = \int _a ^b \biggl [ F_x (x(t),y(t))\frac {dx}{dt}dt + F_y(x(t),y(t))\frac {dy}{dt}dt\biggr ] .
(1.19)

Consider the example of F (x, y) = −y 2 i + x2 j, and the unit half-circle defined by g(x, y) = x2 + y 2 = 1 for y ≥ 0.
dy
√
In this case, dx = −x/y = −x/ 1 − x2 . The line integral is:

\int _C \bm {F} \cdot d\bm {l} = \int _C [-y^2\,dx + x^2\,dy] = \int _C \left [(x^2-1)\,dx + x^2\,\frac {-x}{\sqrt {1-x^2}}\,dx\right ]
(1.20)

The limits of integration are from x = 1 to x = −1 (anti-clockwise around the circle). The second term does not

contribute to the final integral as it is an odd function of x. Hence the final integral is [ 13 x3 − x]−1 4
1 = 3.

Alternatively, we could parameterise the half-circle as l(γ) = cos γ i + sin γ j with γ running from 0 to π. In the

PHAS0038: Term 1 Introduction 16

PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

parametric form, l′ (γ) = − sin γ i + cos γ j and the integral is:

\int _C \bm {F} \cdot d\bm {l} = \int _{\gamma =0}^{\pi } [\sin ^3\gamma +\cos ^3\gamma ]\, d\gamma
(1.21)

Only the sin3 function has the correct ‘symmetry’ to contribute a non-zero integral value on this interval, which

turns out to be 43 , as for the other method of calculation.

• Line Integral of a Gradient: A very important result arises when line-integrating the gradient of a scalar

function between two points a and b:

\int _a ^b \bm {\nabla }\varphi \cdot d\bm {l} = \int _a ^b d\varphi = \varphi (b) - \varphi (a),
(1.22)

so the line integral of a gradient is independent of path.

• Surface Integral Evaluation: Surface integrals can be evaluated in a similar way to the above methods for

line integration, with the added requirement that we need to specify the surface appropriately as we are now

working in three dimensions. For example, the positions of all points on the surface could be specified by using

two appropriate parameters u and v:

\bm {r}(u, v) = x(u, v)\bm {i} + y(u, v)\bm {j} + z(u, v)\bm {k} . (1.23)

∂r ∂r
The vector derivatives ∂u and ∂v must lie in a plane which is locally tangential to the surface. It then follows

∂r ∂r
that their cross product, ∂u × ∂v is orthogonal or ‘normal’ to the surface.

For a plane passing through a given point r0 in space, the normal to the plane can be defined by a unit vector

n̂. Thus all points r which lie on the plane must satisfy the following orthogonality condition:

\bm {\hat {n}}\cdot (\bm {r} - \bm {r}_0) = 0 (1.24)

\rightarrow (a,b,c)\cdot (x-x_0,y-y_0,z-z_0) = 0, (1.25)

where the components of n̂ are a, b, c and a2 + b2 + c2 = 1. Clearly the same plane can be specified by using

any set of values λa, λb, λc, where λ is a real, non-zero constant. It follows that a plane whose equation is

Ax + By + Cz = D, where A, B, C, D are constants, has a normal vector given by (A, B, C) (which may have

non-unit length, depending on the values of these coefficients).

PHAS0038: Term 1 Introduction 17

PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

1.1.2 Overview and Revision of Vector Fields

1.1.2.1 Fields

• In a vacuum, the basic fields are the electric field E and the magnetic field B. E is the force that would

act on a unit positive electric charge at the specified point in space; B is related to the force which would

act on a unit length of a wire carrying a unit current.

• When these fields interact with matter, changes may take place as, for example, charged moving particles

respond to both types of field.

• We then have the related field fields of electric displacement D and magnetic intensity (or magnetizing

field∗ ) H, defined by D = ϵ0 E +P and B = µ0 (H +M ) where P and M are, respectively, the polarization

and magnetization vectors.

• ∗
The term ‘magnetic field’ has been used for both B and H. B is also referred to as the magnetic

induction, or magnetic flux density. In this course, we will usually refer to B as the ‘magnetic field’ and

H as the ‘magnetizing field’ or the ‘magnetic intensity’.

The electric and magnetic fields have SI units of Newtons per Coulomb (or Volts per metre) and Tesla (equivalent

to kilograms per Coulomb second!). Be careful with units: Gaussian units are quite different. It is good practice

to be able to recognize and work in different systems of units, although for this course we will usually work in

SI, or sometimes in what physicists refer to as ‘dimensionless units’. Matter responds to the presence of these

fields: the atoms of molecules polarize in an electric field, and respond in varied ways to a magnetic field (both

diminishing and amplifying it). The fields D and H reflect this behaviour, as we shall see.

1.1.2.2 Electrostatics

This sub-branch of electricity and magnetism is concerned with the electric fields generated by static (non-

moving) systems of charge, with no explicit time dependence.

PHAS0038: Term 1 Introduction 18

PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

• Force between charges For two charges, q1 and q2 at rest at points r1 and r2 , the force felt by q2 due

to q1 , and the electric potential at the location of q2 , are:

Force Field

q1 q2 q1
F (r2 ) = r̂12 E(r2 ) = r̂12 (1.26)
4πϵ0 |r2 − r1 |2 4πϵ0 |r2 − r1 |2

The corresponding potential energy of the charge q2 , and the value of the electric potential at its location,

are:

Energy Potential

q1 q2 q1
U (r2 ) = ϕ(r2 ) = (1.27)
4πϵ0 |r2 − r1 | 4πϵ0 |r2 − r1 |

Note that r̂12 is a unit vector directed from charge q1 to charge q2 . Thus if both charges are of the same sign,

the electrostatic force is repulsive. If the charges are of opposite sign, the force between them is attractive. Both

force and field are directed along a line joining the two charges. The constant ϵ0 is known as the permittivity of

free space (absolute permittivity), and its numerical value is ϵ0 = 8.854 × 10−12 F m−1 in SI units. An alternative

unit for ϵ0 is C 2 N −1 m−2 .

• For a charged, continuous medium with charge density ρ(r), we have:

First Maxwell Equation: Gauss’ Law

\bm {\nabla } \cdot \bm {E}(\bm {r}) = \frac {\rho (\bm {r})}{\epsilon _0} (1.28)

The divergence of the electric field is thus related to the local density of charge at that point in space.

• The same rule can be applied in integral form, for example, to a collection of discrete charges qi inside a

surface S:
\oint _S \bm {E} \cdot \bm {n} dS = \frac {1}{\epsilon _0} \sum _i q_i
(1.29)

• The integral and differential forms of are linked by the Divergence Theorem.

• The total charge in a volume V of a continuous medium is

R
V
ρ dV

PHAS0038: Term 1 Introduction 19

PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

1.1.2.3 Magnetostatics

• Consider an element of a current loop, dl, carrying current I (along its direction) at position r ′ , the ‘source

point’. The contribution to the magnetic field produced by this current element at another point r (the

‘test point’ or the ‘field point’) is:

d \bm {B} (\bm {r}) = \frac {\mu _0I}{4\pi }\frac {d\bm {l} \times (\bm {r}-\bm {r}^{\prime })}{|\bm {r}-\bm {r}^{\prime }|^3} (1.30)

• We can perform a loop integral to get the total magnetic field at the test point:

\bm {B} (\bm {r}) = \frac {\mu _0I}{4\pi }\oint _C \frac {d\bm {l} \times (\bm {r}-\bm {r}^{\prime })}{|\bm {r}-\bm {r}^{\prime }|^3} (1.31)

• We shall see later that, because of the form of this integral, the magnetic field has zero divergence (∇ · B =

0). It follows that there exists a magnetic vector potential A such that B = ∇ × A. Hence:

Second Maxwell Equation (No Magnetic Monopoles∗ )

\bm {\nabla } \cdot \bm {B} = 0 (1.32)

(∗ ... that we have observed)

• What is µ0 and what are its units?

Remember that the Biot-Savart law is empirical: there is no underlying theory stating that there are

no magnetic monopoles in the universe. But we haven’t definitively found any yet! The result derived

from the Biot-Savart law is the second Maxwell equation. µ0 is the permeability of free space, and is

4π × 10−7 T m A−1 (which is equivalent to kilogram metres per square Coulomb).

1.1.2.4 Electromagnetism

• For a surface S bounded by a loop C,

\oint _C \bm {B} \cdot d\bm {l} = \mu _0\,I ,
(1.33)

where I is the total current (charge per unit time) passing through the surface S.

• We can write I as
R
S
J · ndS, where n is a local unit vector normal to the surface, the surface element

dS is thus perpendicular to n, and the vector field J is known as the current density. The dot product

J · ndS represents the current flowing locally through (across) the surface element in the general case that

the direction of current flow (represented by J ) is not parallel to n.

PHAS0038: Term 1 Introduction 20

PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

• Using Stokes’ theorem we find:

\oint _C \bm {B} \cdot d\bm {l} = \int _S \bm {\nabla } \times \bm {B} \cdot \bm {n} dS = \mu _0I =\mu _0 \int _S \bm {J} \cdot \bm {n} dS
(1.34)

For the two surface integrals above to always be equal for an arbitrary surface, it follows that the two

integrands must be equal at every point in space. Thus we have:

Third Maxwell Equation: Ampere’s Law (‘Static’ Case)

\bm {\nabla } \times \bm {B} = \mu _0 \, \bm {J} (1.35)

• We will consider why this form of Ampère’s Law is ‘incomplete’ later in the lectures, though you should

hopefully already have seen this and understood why. This law will form our third Maxwell equation when

presented in its complete form to include time-dependent effects.

• If a conducting circuit, C, is intersected by a B field, then the magnetic flux through any surface S bounded

by C, is given by:

\Phi _C = \int _S \bm {B} \cdot \bm {n} dS \label {eq:magflux} (1.36)

• The EMF induced around the circuit is generally defined by:

\mathcal {E} = -\frac {d\Phi }{dt} = \oint _C \bm {E^{\prime }} \cdot d\bm {l} \label {eq:emf} (1.37)

where the flux may change with time because of explicit time-dependence of B, and / or the circuit C itself

moving with time. E ′ represents the electric field in an appropriate reference frame (i.e the rest frame of

the conductor, whether it is static or moving in our frame of reference).

• If we assume, for simplicity, a static circuit in our frame of reference, then the electric field we observe is

E and we can apply Stokes’ Theorem as follows:

\int _S \frac {d\bm {B}}{dt} \cdot \bm {n} dS = \int _S \frac {\partial \bm {B}}{\partial t} \cdot \bm {n} dS = -\oint _C \bm {E} \cdot d\bm {l} = -\int _S \bm {\nabla } \times \bm {E} \cdot \bm {n} dS
(1.38)

and thus derive:

Fourth Maxwell Equation: Faraday’s Law

\bm {\nabla } \times \bm {E} = - \frac {\partial \bm {B}}{\partial t} (1.39)

As you can see, the derivation is almost trivial: substitute Eq. (1.36) into Eq. (1.37), and then apply

Stokes’ theorem to the loop integral of E. This is the final Maxwell equation.

PHAS0038: Term 1 Introduction 21

PHAS0038: Electromagnetic Theory Chapter 1 - Introduction

• Ampère’s law as described above is incomplete: it needs to account for time-varying electric fields

• When we do this, we can write the final set of equations governing the evolution of electric and magnetic

fields (in a vacuum):

Maxwell’s Equations:

\bm {\nabla } \cdot \bm {E} &= \frac {\rho }{\epsilon _0} \\ \bm {\nabla } \cdot \bm {B} &= 0 \\ \bm {\nabla } \times \bm {B} &= \mu _0 \bm {J} + \mu _0 \epsilon _0 \frac {\partial \bm {E}}{\partial t} \\ \bm {\nabla } \times \bm {E} &= - \frac {\partial \bm {B}}{\partial t}

(1.40d)

• To complete our set of equations, we have the force on a charge q moving with velocity v:

Lorentz Force
\bm {F} = q(\bm {E} +\bm {v} \times \bm {B}) (1.41)

Once Maxwell’s equations and the Lorentz force law have been specified, classical electromagnetism is

essentially complete: the basic physics has not changed, though the details of the interaction of the fields

with matter are still being understood.

PHAS0038: Term 1 Introduction 22