Ed Book
Ed Book
P UBLISHED BY P UBLISHER
1 Mathematical Preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.1 Vector Analysis 5
1.1.1 Vector Addition, Subtraction and Multiplication . . . . . . . . . . . . . . . . . . . . . 5
1.1.2 Gradient of a Scalar Field . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.1.3 Divergence of a Vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
1.1.4 Curl of a vector . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.2 Co–ordinate systems 12
1.2.1 Catesian co–ordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
1.2.2 Cylindrical co–ordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
1.2.3 Spherical Polar Co–ordinates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
1.3 Mathematical Symbols 15
1.3.1 Kronecker Delta and Levi–Civita Symbols . . . . . . . . . . . . . . . . . . . . . . . . . . 15
1.3.2 Uses of Kronecker Delta and Levi–Civita symbol . . . . . . . . . . . . . . . . . . . . 15
1.4 Dirac Delta Function 16
2 Electrostatics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
2.1 Conducting Sphere in a Uniform Electric Field 21
2.2 Green’s Function Method 25
7 Electromagnetic Waves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39
Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43
Books 43
Articles 43
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
1. Mathematical Preliminaries
here, x̂, ŷ, and ẑ are the unit vector along the x, y, and z axis of the Cartesian plane. It is
important to note that x̂, ŷ, and ẑ are not a variables. However while performing vector
6 Chapter 1. Mathematical Preliminaries
additions or subtraction one should treat them as one. The following are the properties of
the vector additions/subtraction.
• Vector addition is commutative: ⃗A + ⃗B = ⃗B + ⃗A
• Vector addition is associative: (⃗A + ⃗B) + C
⃗ = ⃗A + (⃗B) + C
⃗
⃗ ⃗
• Vector subtraction in form of vector addition: A − B = A + (−⃗B)
⃗
• Vector addition multiplied by a scalar number is distributive: k(⃗A + ⃗B) = k⃗A + k⃗B
Example 1.1: Find a vector ⃗A directed from (2,-4,1) to (0,-2,0) in cartesian coordinates
and find the unit vector along ⃗A.
Answer:
⃗A = (0 − 2)x̂ + {−2 − (−4)}ŷ + (0 − 1)ẑ = −2x̂ + 2ŷ − ẑ
Answer: Here the direction of the forces are very important as this is an example of
tug-of-war the force are moving along a line outwards, so are the unit vectors. Thus, the
resultant force will be -
Fresultant = Fteam−1 + Fteam−2 + Fteam−3
Now,
4 3
Fteam−1 = Force magnitude × Force direction = 5 × ( aˆx + aˆy ) = 4aˆx + 3aˆy
5 5
√ √
= (−3 3 − 4)aˆx√− (3 + 4 3)aˆy
Similarly, Fteam−2 √
−12 3+16
and Fteam−3 = 5 ˆx + 12−16
a√ 5
3
aˆy √
−27 3+16
Thus, Fresultant = 5 aˆx + 12−4
5
3
aˆy
If ⃗A and ⃗B are multiplied, then the result will be either a scalar or a vector. It depends
on how multiplications of those are made. The two types of vector multiplication are:
1.1 Vector Analysis 7
Note that ⃗A · ⃗B is a scalar and not a vector. For example force and displacement are two
vectors, but their product is the work done by the force due to the displacement and is a
scalar.
Example 1.3: Show that ⃗A = 4x̂ − 2ŷ − ẑ and ⃗B = x̂ + 4ŷ − 4ẑ are perpendicular.
Answer:
⃗A · ⃗B = (4)(1) + (−2)(4) + (−1)(−4) = 0
Vector product (or Cross product): The vector or cross product of two vectors ⃗A and
⃗B, denoted by [ ⃗A × ⃗B ( ⃗A cross ⃗B)], is defined by
⃗A × ⃗B = AB sinθ n̂ (1.3)
where θ represents the angle between the vectors, n̂ is the unit vector in the direction
normal to the plane of the two vectors. The magnitude of cross product is defined as
|⃗A × ⃗B| = AB sineθ and the unit vector is defined as n̂ = (⃗A × ⃗B)/|⃗A × ⃗B|. Here, ⃗A × ⃗B is
neatly written down in component form as
x̂ ŷ ẑ
⃗A × ⃗B = Ax Ay Az
(1.4)
Bx By Ay
Example 1.4: Given ⃗A = 2x̂ + 4ŷ and ⃗B = 6ŷ − 4ẑ, find the smaller angle between them
using (a) the cross product, (b) the dot product.
Answer:
(a)
x̂ ŷ ẑ
⃗A × ⃗B = 2 4 0 = −16x̂ + 8ŷ + 12ẑ
0 6 −4
q
|⃗A|= (2)2 + (4)2 + (0)2 = 4.47
q
|⃗B|= (0)2 + (6)2 + (−4)2 = 7.21
8 Chapter 1. Mathematical Preliminaries
q
|⃗A × ⃗B|= (−16)2 + (8)2 + (12)2 = 21.54
Then, since |⃗A × ⃗B|= |⃗A||⃗B|sin θ ,
21.54
sin θ = = 0.668 or θ = 41.9◦
(4.47)(7.21)
(b)
⃗A · ⃗B = (2)(0) + (4)(6) + (0)(−4) = 24
⃗A.⃗B 24
cos θ = = = 0.745 or θ = 41.9◦
⃗ ⃗
|A.B| (4.47)(7.21)
Then
x̂ ŷ ẑ
(⃗A × ⃗B) × C
⃗ = 2 −2 −1 = −2ŷ + 4ẑ
0 2 1
Parentheses are not needed in the scalar triple product since it has meaning only when
the cross product is taken first. In general, it can be shown that
⃗ ⃗y ⃗ z
A A A
x
⃗A · ⃗B × C
⃗ = B
⃗ ⃗y
B ⃗ z
B
x
C⃗x ⃗y
C ⃗z
C
As long as the vectors appear in the same cyclic order the result is the same. The scalar
triple products not in this cyclic order have a change in sign.
1.1 Vector Analysis 9
where n̂ is naturally the unit vector perpendicular to the surface. Note, however, that there
are two shortcomings of this way of defining a flux through a closed surface:
10 Chapter 1. Mathematical Preliminaries
´
Figure 1.1: Calculating the integral A · dl in a closed loop in the y − z plane.
Technically, of course, this is not quite correct! ∇ × A is a vector quantity and the right
hand side is NOT a vector. We must, therefore, write three of these equations, one for each
1.1 Vector Analysis 11
We can evaluate the first of the above expressions, and calculate the x−component of the
curl. To set this up, refer to Fig.(1.1). Let ABCD be the closed curve of infinitesimal¸ dimen-
sions enclosing an infinitesimal area ∆Sx along which we will evaluate lim∆S→0 ∆S1 x C A·dl.
Thus, we get, taking a middle–value of the vector A for each segment,
ˆ
∆y
For AB : A · dl = A(y + , z)∆y
2
ˆ
∆z
For BC : A · dl = A(y + ∆y, z + )∆z
2
ˆ
∆y
For CD : A · dl = −A(y + , z + ∆z)∆y
2
ˆ
∆z
For DA : A · dl = −A(y, z + )∆z (1.15)
2
Adding the first and third lines, we get
∆y ∆y ∂ Ay (z)
AB and CD : A(y + , z) − A(y + , z + ∆z) ∆y = − ∆z∆y (1.16)
2 2 ∂z
Similarly, adding the second and fourth lines in Eqs.(1.15), we get
∆z ∆z ∂ Az (y)
BC and DA : A(y + ∆y, z + ) − A(y, z + ) = ∆y∆z (1.17)
2 2 ∂y
Noting that ∆Sx = ∆y∆z, we get, finally, for the curl of A,
˛
1 ∂ Az ∂ Ay
[∇ × A]x = lim A · dl = − (1.18)
∆S→0 ∆Sx C ∂y ∂z
Answer: Let’s compute the divergence first and there isn’t much to do other than run
through the formula.
∂ 2 ∂ ∂
∇ · ⃗F = 3x − z3 + 4y2 = 2xy
x y +
∂x ∂y ∂z
x̂ ŷ ẑ
∂ ∂ ∂
∇ × ⃗F =
∂x ∂y ∂z
x y 3x − z 4y2
2 3
12 Chapter 1. Mathematical Preliminaries
∂ ∂ 2 ∂ ∂ 2 ∂ ∂
4y2 x̂ + 3x − z3 ẑ − 4y2 ŷ − 3x − z3 x̂
= x y ŷ + x y ẑ −
∂y ∂z ∂x ∂y ∂x ∂z
= 8y + 3z2 x̂ + 3 − x2 ẑ
x3 y2
Example 1.8: Compute ∇ · ⃗F and ∇ × ⃗F for ⃗F = 3x + 2z2 x̂ +
ŷ − (z − 7x) ẑ.
z
Answer: Let’s compute the divergence first.
x3 y2 2x3 y
∂ ∂ ∂
∇ · ⃗F = ∇ ⃗F = 3x + 2z2 + + (7x − z) = 2 +
∂x ∂y z ∂z z
x̂ ŷ ẑ
∂ ∂ ∂
curl ⃗F = ∇ × ⃗F =
∂x ∂y ∂z
x 3 y2
3x + 2z2 7x − z
z
x3 y2 ∂ x3 y2
∂ ∂ ∂ ∂ ∂
3x + 2z2 ŷ+ 2
= (7x − z) x̂+ ẑ− 3x + 2z ẑ− (7x − z) ŷ− x̂
∂y ∂z ∂x z ∂y ∂x ∂z z
3x2 y2 x3 y2
= 4zŷ + ẑ − 7ŷ + 2 x̂
z z
x3 y2 3x2 y2
= x̂ + (4z − 7) ŷ + ẑ
z2 z
A = Ax x̂ + Ay ŷ + Az ẑ. (1.19)
We always use r to denote the position vector and ri to indicate either x, y or z, so that
The expression for the ∇2 operator, operating on a scalar field, which we will discuss later
in the chapter, is particularly simple (this operator can operate on vectors as well and in
that case, the scalar can simply be replaced by the vector):
∂ 2ϕ ∂ 2ϕ ∂ 2ϕ
∇2 ϕ = + 2 + 2. (1.22)
∂ x2 ∂y ∂z
The notation for the components and unit vectors in this system is
A = Aρ ρ̂ + Aφ ϕ̂ + Az ẑ. (1.24)
The volume element in cylindrical coordinates is d 3 r = ρdρdφ dz. The gradient or ‘del’
operator in this system is
∂ ϕ̂ ∂ ∂
∇ = ρ̂ + + ẑ . (1.25)
∂ρ ρ ∂φ ∂z
1 ∂ (ρAρ ) 1 ∂ Aφ ∂ Az
∇·A = + + (1.26)
ρ ∂ρ ρ ∂φ ∂z
1 ∂ Az ∂ Aφ ∂ Aρ ∂ Az 1 ∂ (ρAφ ) ∂ Aρ
∇×A = − ρ̂ + − ϕ̂ + − ẑ (1.27)
ρ ∂φ ∂z ∂z ∂ρ ρ ∂ρ ∂φ
1 ∂ 2 Aφ ∂ 2 Az
2 1 ∂ ∂ Aρ
∇ A = ρ + 2 + . (1.28)
ρ ∂ρ ∂ρ ρ ∂φ2 ∂ z2
14 Chapter 1. Mathematical Preliminaries
Example 1.9: Let ⃗r(x, y, z) = x⃗i + y⃗j + z⃗k be the position vector field on R3 . Then
∥⃗r(x, y, z)∥2 =⃗r ·⃗r = x2 + y2 + z2 is a real-valued function. Find
Answer:
(a) ∇∥⃗r∥2 = 2x⃗i + 2y⃗j + 2z⃗k = 2⃗r
∂ ∂ ∂
(b) ∇ ·⃗r = (x) + (y) + (z) = 1 + 1 + 1 = 3
∂x ∂y ∂z
î
jˆ k̂
∂ ∂ ∂
(c) ∇ ×⃗r = = 0î − 0 jˆ + 0k̂ = ⃗0
∂x ∂y ∂ z
x y z
∂2 2 ∂2 2 ∂2 2
(d) ∇∥⃗r∥2 = (x + y2
+ z2
) + (x + y2
+ z2
) + (x + y2 + z2 ) = 2 + 2 + 2 = 6
∂ x2 ∂ y2 ∂ z2
1.3 Mathematical Symbols 15
These are particularly useful since expressions in cartesian and even other coordinate
systems can be easily written out as well as evaluated using the above definitions, as
discussed next.
Let us now use the above for evaluating a couple of vector formulae.
Example 1.10: Prove vector identity ∇ × (∇ × ⃗A) = ∇(∇ · ⃗A) − ∇2⃗A using Levi–Civita
symbol.
Answer: We’ll write the ith Cartesian component of the gradient operator ∇ as ∂i (cf. ∂∂xi ).
Let’s simplify ∇ × (∇ × ⃗A(x)). We start by considering the ith component and then we use
our expression for the cross product (working from the outside in):
∇ × (∇ × ⃗A)i = εi jk ∂ j (∇ × ⃗A)k
Next we replace the remaining cross product, making sure to introduce new dummy
summation variables l and m:
(The partial derivatives act only on the components of A, so we can pull out the ε’s.) We
rotated the indices in one of the ε’s in the last step so that we can now directly apply our
very useful identity (and simplify)
Thus we arrive at
In other words, we need a function that "chooses" a point and nothing else; i.e. it should
have a large value at one point only, and should be zero everywhere else, such that it is
normalized, i.e.
ˆ ∞
δ (x − x0 )dx = 1. (1.53)
−∞
1.4 Dirac Delta Function 17
Eq.(1.52) is a filtering property and can be used to prove the following identities
1
δ (ax) = δ (x) for a ̸= 0 (1.54)
|a|
ˆ ∞
d ′
d
dx f (x) δ x − x = − f (x) (1.55)
−∞ dx dx ′
ˆ ∞ x=x
′ 1 ′
dkeik(x−x )
δ x−x = (1.56)
2π −∞
1
δ [g(x)] = ∑ ′ δ (x − xn ) where g(xn ) = 0, g′ (xn )(1.57)
̸= 0.
n |g (x n )|
The dirac delta function is particularly useful in dealing with point charges, where the
charge density is huge at a single point and zero everywhere else. Since we are going to
deal with point charges in 3 dimensions, it makes sense to define the dirac delta function
in 3 dimensions as well, in the following way
ˆ (
1, if r ′ ∈ V
d 3 r f (r)δ (r − r ′ ) = (1.58)
V 0, if r ′ ∈
/V
Problems:
1. Under what circumstances are the vectors (⃗A × ⃗B) × C
⃗ and ⃗A × (⃗B × C)
⃗ equal?
⃗
2. If the vector ⃗A has constant magnitude, show that the vectors ⃗A and ddtA are perpen-
⃗
dicular, provided ddtA ̸= 0.
3. If V is a scalar function of x and y, show that the divergence of the vector field
⃗F = aˆz × grad V is always zero.
4. The direction of a vector ⃗A is radially outwards from the origin, and its magnitude is
krn , where
r2 = x2 + y2 + z2
Evaluate the line integral of A around the contour in the z = 0 plane bounded by +
x-axes and + y-axes and the arc of the circle of radius 1 unit. Check the answer by
performing the appropriate surface integral of curl ⃗A.
6. A point P moves so that its position vector r relative to another point O satisfies the
equation
d⃗r
⃗ ×⃗r
=ω
dt
where ω⃗ is a constant vector. Prove that P describes a circle with constant velocity.
7. Given ⃗A = 4ŷ + 10ẑ and ⃗B = 2x̂ + 3ŷ, find the projection of ⃗A on ⃗B.
18 Chapter 1. Mathematical Preliminaries
√
8. Given ⃗A = (10/ 2)(x̂ + ẑ) and ⃗B = 3(ŷ + ẑ), find the projection of ⃗B on ⃗A as a vector
in the direction of ⃗A.
9. Find the relationship which the cartesian components ⃗A and ⃗B must satisfy if the
vector fields are everywhere parallel.
10. Express the unit vector directed toward the origin from an arbitrary point on the line
described by x=0, y=3.
11. Given the plane 4x + 3y + 2z = 12, find the unit vector normal to the surface in the
direction away from the origin.
12. Express the unit vector directed toward the point (0, 0, h) from an arbitrary point in
the plane z = −2.
13. Find the unit vector directed from (2, −5, −2) toward (14, −5, 3).
14. Show that the absolute value of ⃗A · ⃗B × C ⃗ is the volume of the parallelopiped with
edges ⃗A, ⃗B and C.
⃗
15. Given ⃗A = 2x̂ − ẑ, ⃗B = 3x̂ + ŷ, and C ⃗ = −2x̂ + 6ŷ − 4ẑ, show that C
⃗ is ⊥ to both ⃗A
⃗
and B.
16. Convert the following rectangular Cartesian coordinate (1, 3, 5) into an equivalent
(a) cylindrical and (b) spherical coordinates. ´
17. Let ⃗F = ⟨xy, −xy⟩ and 0 let D be given by 0 ≤ x ≤ 1, 0 ≤ y ≤ 1. Compute ∂ D ⃗F · d⃗r
´
and ∂ D ⃗F · n̂ ds.
18. If ∇ · ⃗F = 0, ⃗F is said to be incompressible. Show that any vector field of the form
⃗F(x, y, z) = ⟨ f (y, z), g(x, z), h(x, y)⟩ is in-compressible. Give a non-trivial example.
19. Compute ∇ · ⃗F and ∇ × ⃗F
⃗F = (2y − cos (x)) x̂ − z2 e3x ŷ + x2 − 7z ẑ
(u · v)2 ≤ (u · u)(v · v)
with equality holding if and only if u = λ v for some λ ∈ R. Prove the Cauchy-
Schwarz inequality. [Hint: Use angles.]
23. Show that the equation a·r = a2 defines a two dimensional plane in three dimensional
space, where a is the minimal length vector from the origin to the plane. [Hint:
A plane is the translate of the linear span of two vectors. The Cauchy-Schwarz
inequality may come in handy.]
24. The volume of a tetrahedron is V = bh/3, where b is the area of a base and h is the
height (distance from base to apex). Consider a tetrahedron with one vertex at the
origin and the other three vertices at positions ⃗A, ⃗B and C.
⃗ Show that we can write
1
V = ⃗A · (C
⃗ × C)
⃗
6
1.4 Dirac Delta Function 19
This demonstrates that the volume of such a tetrahedron is one sixth of the volume
of the parallelepiped defined by the vectors ⃗A, ⃗B and C.
⃗
25. Using index notation, prove Lagrange’s identity:
(⃗A × ⃗B) · (C
⃗ × ⃗D) = (⃗A · C)(
⃗ ⃗B · ⃗D) − (⃗A · ⃗D)(⃗B · C)
⃗
26. Either directly or using index methods (Levi-Cevita symbols), show that, for any
scalar field φ and vector field ⃗A,
(a) ∇ · (φ ⃗A) = ∇φ · ⃗A + φ ∇ · ⃗A,
(b) ∇ × (φ ⃗A) = ∇φ × ⃗A + φ ∇ × ⃗A, and
(c) ∇2 (φ ψ) = (∇2 φ )ψ + 2∇φ · ∇ψ + φ ∇2 ψ
27. Show that, for r ̸= 0,
(a) ∇ · r̂ = 2r , and
(b) ∇ × r̂ = 0
28. A function φ satisfying ∇2 φ = 0 is called harmonic
(a) Using Cartesian coordinates, show that φ = 1/r is harmonic, where r = (x2 +
y2 + z2 )1 /2 ̸= 0.
(b) Let α = (α1 , α2 , α3 ) be a vector of non-negative integers, and define |α| =
α1 + α2 + α3 . Let ∂α be the differential operator ∂xα1 ∂yα2 ∂zα3 . Prove that any
function of the form φ = r2|α| +1 ∂α (1/r) is harmonic. [Hint: Use vector calculus
identities to expand out ∇2 (rn f ) where f := ∂α (1/r) and use Euler’s theorem and
the fact that mixed partials commute.]
2. Electrostatics
E = E0 ẑ
∂V
=⇒ − = E0
∂z
=⇒ V = E0 z = −E0 r cos θ
∇2V = 0 (2.1)
−→ The second equation on the previous slide is known as the Legendre Equation
−→ The solution to the first one is
B
Z(r) = Arℓ + (2.7)
rℓ+1
where the constants A and B are to be determined from Boundary conditions.
P0 (cos θ ) = 1
P1 (cos θ ) = cos θ
1
3 cos2 θ − 1
P2 (cos θ ) =
2
.....
i.e. there is ONE AND ONLY ONE WAY TO GET P(θ ) = cos θ and that is when ℓ = 1
24 Chapter 2. Electrostatics
R3
V (r, θ ) = A1 r − 2 cos θ (2.12)
r
Here,
−→ ∇2 ψ0 (r) = 0 −→ The source term
−→ So how does it work?
Consider eq.(2.19); multiply both sides with f (r ′ ) and integrate over r ′ :
ˆ ˆ
∇ G(r, r ) f (r ) d r = −4π f (r ′ )δ (r − r ′ )d 3 r′
′ ′
2
3 ′
(2.21)
where we can take f (r ′ ) inside ∇2 since the operator is only a function of r and not r ′ .
We then get
ˆ
2 ′ ′ 3 ′
∇ G(r, r ) f (r )d r = −4π f (r) = ∇2 ψ(r) (2.22)
Therefore, in the previous equation, ψ(r) must be equal to the expression within the square
brackets, PLUS another term, the Laplacian of which is zero...
...so that
ˆ
ψ(r) = ψ0 (r) + G(r, r ′ ) f (r ′ )d 3 r′ (2.23)
So now, we must find G(r, r ′ ) through the equation that defines it:
At first glance, it seems like this is going to be equally difficult as the Poisson equation
solution itself without this method!
−→ However, the reason it is simplified is that we can write the Green’s function in terms
of its fourier transform:
ˆ
′ 1 ′
3
G(r, r ) = G̃(k) exp ik · (r − r ) d k (2.25)
(2π)3
−→ By writing the Green’s function in terms of its fourier transform, we can change
the equation we wish to solve from a differential equation to an algebraic equation in the
26 Chapter 2. Electrostatics
following way
First, substitute the expression for δ (r − r ′ ):
ˆ
′ 1 ′
3
δ (r − r ) = exp ik · (r − r ) d k (2.26)
(2π)3
and then substitute for G(r, r ′ ) in eq.(2.24) to get
ˆ
∇ G̃(k) exp ik · (r − r ) d k = −4π exp ik · (r − r ′ ) d 3 k
2 ′
3
(2.27)
Now, remember that the operator ∇2 is only a function of r and not r ′ , so that it ONLY
operates on the exponential inside the integral on the LHS of eq.(2.27)
2
−→ Also, when it acts on the exponential inside the integral, it is effectively ∂∂r2
−→ So, performing the differentiation twice on the exponential inside the integral on the
LHS, we get
ˆ ˆ
k G̃(k) exp ik · (r − r ) d k = 4π exp ik · (r − r ′ ) d 3 k
2 ′
3
(2.28)
We have seen in §2 on Electrostatics that static charges exert a force on each other, which
is proportional to the inverse squared distance between them, and to the multiple of the two
charges. While it is reasonable to expect that such a force would also exist between moving
charges, applying Coulomb’s Law to moving charges is not possible. Therefore, for charges
moving with a constant speed; in other words, for steady currents, we need to describe
a different force, which is referred to as the magnetic force. Thus, while electrostatics
deals with static electric charges, magnetostatics deals with stationary electric currents, i.e.,
electric charges moving with constant speeds, and the interaction between these currents.
The study of magnetic forces or magnetism has historically involved the use of what are
called magnetic materials. We shall deal with these in the next chapter; for the discussion
in this chapter, we will concentrate on wires carrying steady currents. As we shall see
later, this simplifies our treatment and understanding of magnetic forces, i.e. magnetism
considerably.
we get
µ0 (r − r ′ ) ′
dB = j(r ′ ) × dV (4.3)
4π |r − r ′ |3
−→ Now,
∇ × ( f A) = f (∇ × A) + A × ∇ f (4.4)
since the other term reduces to zero (because the ∇ operator is a function of r only whereas
j is a function of r ′ only)
∇ · (A × B) = B · (∇ × A) − A · (∇ × B) (4.7)
so that
ˆ " #
µ0 (r − r ′ )
∇·B = ∇ · j(r ′ ) × dV ′ (4.8)
4π V′ |r − r ′ |3
And so
ˆ " #
µ0 (r − r ′ )
∇·B = ∇ · j(r ′ ) × dV ′
4π V′ |r − r ′ |3
ˆ
µ0 ′ 1
= − j(r ) · ∇ × ∇
4π V′ |r − r ′ |
≡ 0 (4.9)
−→ “No magnetic monopoles” is built–in to Biot–Savart Law!
This also makes sense if we define
B ≡ ∇ × A where
ˆ
µ0 j(r ′ )
A = dV ′ (4.10)
4π |r − r ′ |
4.2 Divergence and Curl of the Magnetic Field 31
∇ × B = µ0 j(r) (4.11)
F = qv × B (4.13)
F = q (E + v × B) (4.14)
5. Magnetic Materials and Inductance
6. Time Varying Fields and Maxwell’s Equa
∇ · ∇ × B ≡ 0 = µ0 ∇ · j (6.1)
but Ampere’s Law MUST hold for all current densities, not only divergenceless ones!
Turns out, conservation of charge saves us here, in the following way
For any volume V , the sum total of charges moving in and out must be exactly the same:
˛ ˆ
∂ ∂ρ
− j · dS = ρdV =⇒ ∇ · j + =0 (6.2)
∂t ∂t
So we can now substitute for the divergence of the current density thus:
∂ ∂E
∇ · j = −ε0 [∇ · E] =⇒ ∇ · j + ε0 (6.3)
∂t ∂t
∂E
∇ × B = µ0 j + µ0 ε0 (6.4)
∂t
˛ ˆ
∂
E · dl = − B · dS (6.5)
∂t
or, equivalently, using the definition of the curl:
∂B ∂A
∇×E = − = ∇× − (6.6)
∂t ∂t
This CLEARLY suggests that we can write the electric field consists of 2 parts and can
therefore be written as
∂A
E = −∇ϕ − (6.7)
∂t
and the magnetic field is
B = ∇×A (6.8)
Let us now use the four Maxwell equations in their dynamic form
ρ
∇·E = (6.9)
ε0
∂B
∇×E = − (6.10)
∂t
∇·B = 0 (6.11)
∂E
∇ × B = µ0 j + µ0 ε0 (6.12)
∂t
and substitute in the above to get, after some algebra,
1 ∂ 2ϕ
2 ρ ∂ 1 ∂ϕ
−∇ ϕ = + ∇·A+ 2 (6.13)
c2 ∂t 2 ε0 ∂t c ∂t
1 ∂ 2A
2 1 ∂ϕ
− ∇ A = µ0 j − ∇ ∇ · A + 2 (6.14)
c2 ∂t 2 c ∂t
The above equations are the Maxwell equations in terms of the potentials. Note that this is
a set of four equations, since A is the vector potential.
Notice that the terms in the square brackets in both the equations are EXACTLY the
same. This is not a coincidence. These represent what is known as a Gauge Freedom, i.e.
the freedom to choose the divergence of the vector potential and the time rate of change of
the scalar potential.
B = ∇×A (6.15)
6.3 Non–Uniqueness of Scalar & Vector Potential 37
due to which we can only define A up to the gradient of a scalar function, say χ, so that
the transformation A → A′ = A + ∇χ produces no difference in the magnetic field B.
d2y dy
1 − x2 2
− 2x + n(n + 1)y = 0 (A.1)
dx dx
is known as Legendre’s equation1 . The parameter n is a real number, and in most
applications, only integral values are needed.
To obtain a solution, assume
∞
y= ∑ ak xm+k (A.2)
k=0
(A.3)
or
∑(m + k)(m + k − 1)ak xm+k−2 − ∑{(m + k)(m + k + 1) − n(n + 1)}ak xm+k = 0 (A.4)
Now, the ONLY way that the entire LHS can equal zero is if each power of x equals zero.
Let us look at the coefficient of the lowest power of x, i.e. coefficient of xm−2 , which must
be = 0
Now let us examine the next higher power of x, whose coefficients must also be = 0
m(m + 1)a1 = 0 =⇒ a1 = 0; (A.6)
and, combined with the next such relations, we will get a1 = a3 = a5 . . . = 0. Thus,
alternate terms are absent in eq.(A.2).
Now consider eq.(A.4). In order to make the power of x m + k, we can make the
substitution k → k + 2 WLOG (since all ak s are assumed to be zero). This makes the
equation
∑(m+k +2)(m+k +1)ak+2xm+k − ∑{(m+k)(m+k +1)−n(n+1)}ak xm+k = 0 (A.7)
such that we may write
(m + k + 2)(m + k + 1)ak+2 − {(m + k)(m + k + 1) − n(n + 1)}ak = 0
(m + k − n)(m + k + n + 1)
=⇒ ak+2 = ak
(m + k + 1)(m + k + 2)
For m = 0, this becomes
(n − k)(n + k + 1)
ak+2 = ak
(k + 1)(k + 2)
Therefore,
n(n + 1)
a2 = a0 ,
2!
(n − 2)(n + 3) n(n + 1)(n − 2)(n + 3)
a4 = a2 = a0 (A.8)
3·4 4!
etc. The coefficients for the case m = 1 can be similarly calculated.
Hence, the two solutions are
n(n + 1) 2 n(n + 1)(n − 2)(n + 3) 4
y1 = a0 1 − x + x +...
2! 4!
(n − 1)(n + 2) 3 (n − 1)(n + 2)(n − 3)(n + 4) 5
y2 = a0 x − x + x +... (A.9)
3! 5!
These polynomials, with the value of a0 normalized such that their value is 1 at x = 1, are
known as Legendre Polynomials are are denoted by Pn (x).
Note the following properties of Legendre Polynomials:
1. These series, when non–terminating, are convergent for |x| < 1
2. When n is an integer, one or the other of the two series in eqs.(A.9) terminates
3. Some of the first few Legendre Polynomials are
P0 (x) = 1
P1 (x) = x
1
3x2 − 1
P2 (x) =
2
1
5x3 − 3x
P3 (x) =
2
1
35x4 − 30x2 + 3
P4 (x) =
8
1 5
P5 (x) = 63x − 70x3 + 15x
8
... (A.10)
Bibliography
Books
Articles
Index
C
Conducting Sphere in Uniform Electric Field
21