Syllabus and Books:

Tensor Analysis
Coordinate transformations, scalars, Covariant and Contravariant tensors. Addition,
Subtraction, Outer product, Inner product and Contraction. Symmetric and
antisymmetric tensors. Quotient law. Metric tensor. Conjugate tensor. Length and
angle between vectors. Associated tensors. Raising and lowering of indices. The
Christoffel symbols and their transformation laws. Covariant derivative of tensors.

Books and References:

1) Mathematical Methods for Physicists: T. L. Chow
2) Tensor Calculas; Schaum Outline:David C. Key
3) Mathematical Physics: Arfken, Weber
4) Vector Analysis (Last Chapter): Spigel
Tensors are important in many areas of physics, ranging from topics such as
general relativity and electrodynamics to descriptions of the properties of bulk
matter such as stress (the pattern of force applied to a sample) and strain (its
response to the force), or the moment of inertia (the relation between a torsional
force applied to an object and its resultant angular acceleration). Tensors
constitute a generalization of quantities previously introduced: scalars and vectors.
We identified a scalar as an quantity that remained invariant under rotations of
the coordinate system and which could be specified by the value of a single real
number. Vectors were identified as quantities that had a number of real
components equal to the dimension of the coordinate system, with the
components transforming like the coordinates of a fixed point when a coordinate
system is rotated. Calling scalars tensors of rank 0 and vectors tensors of rank 1.
Before the presentation of the main development of the subject, we begin by introducing the
summation convention, which will prove very useful in writing tensor equations in a more compact
form.
Dummy and Free index

Subscripts that are summed over are called dummy subscripts or dummy index and the
others free subscripts. For example i in aixi is the dummy index and it could be replaced
by any other character (say akxk or anxn). In aijbjk i and k are free index and j is dummy
index. Naturally, free subscripts must not be changed at all, you can not replace a free
index by any other index.

It is worth remarking that when introducing a dummy subscript into an expression, care
should be taken not to use one that is already present, either as a free or as a dummy
subscript. For example, aijbjk cannot, and must not, be replaced by aikbkk or by aiibik, but
could be replaced by aimbmk or by ainbnk.
Einstein’s Summation Convention:
Any expression involving a twice-repeated index (occurring twice as a superscript,
twice as a subscript, or once as a superscript and once as a subscript) shall
automatically stands for its sum over values 1, 2, 3, …., n of the repeated index.
Kronecker Delta and Algebraic Manipulation
Matrix Representation of Kronecker Delta
In matrix language, a ij δ jk can be written as AI = A, where A is the matrix with elements aij
and I is the unit matrix having the same dimensions as A.

( ) ( )
1 0 0 a 11 a 12 a 13
δ ij =δ ij =δ ij= 0 1 0 = I Aij = a 21 a 22 a 23
0 0 1 a 31 a 32 a 33

( )( )( )
a 11 a 12 a 13 1 0 0 a 11 0 0
AI=A= a 21 a 22 a 23 0 1 0 = 0 a 22 0
a 31 a 32 a 33 0 0 1 0 0 a 33
Examples
 Home work
1.

2.

3.

4.

5.

6.

7.
The behavior under coordinate transformation is of central importance for tensor analysis.
 Tensors:

A mathematical object analogous to but more general than a vector,

represented by an array of components that are functions of the
coordinate space.

Stress, Moment of Inertia, Susceptibility, Electro-magnetic field tensor, Einstein tensor

Force (one direction)

Stress = ------------------------------
Unit area (another direction)
Introduction, properties
we identify a tensor of rank n in a d-dimensional space as an object with
the following properties:

• It has components labeled by n indices, with each index assigned values

from 1 through d, and therefore having a total of dn components;
No. of components = (dimension) rank

In 3-D a 1st rank tensor has 31= 3 components. Velocity, momentum ( Pi )

In 3-D a 2nd rank tensor has 32= 9 components. Stress, MoI, Susceptibility ( Sij )
In 4-D a 2nd rank tensor has 42= 16 components. EM field tensor ( F μν ), Richi tensor ( R μν )

• The components transform in a specified manner under coordinate

transformations.

The behavior under coordinate transformation is of central importance for tensor analysis and
conforms both with the way in which mathematicians define linear spaces and with the physicist’s
notion that physical observables must not depend on the choice of coordinate frames.
Coordinate transformation relations:
Tensors are defined by means of their properties of transformation under coordinate
transformation.

Lets start with a prototype and then give a general definition.

Consider two neighboring points in the manifold P and Q with
Coordinates xi and xi+dxi, respectively.
⃗
PQ
The two points define an infinitesimal displacement vector .

The components of the vector in the xi-coordinate system are dxi.

i i
x̄ d x̄ .
The components of the vector in the -coordinate system are

When we transform the coordinates, their differentials transform according to the

relation i ∂ x̄ i j
d x̄ = dx
∂ xj
First Order/ Rank Tensor
Consider a vector field V=(Vi) defined on some curve of region Rn. For each i, the
component Vi=Vi(x) is a scalar field (real value function) as x varies over the curve .

Let the n components V1, V2, V3, …, Vn of V be expressible as n real-valued functions in each
admissible coordinate system; as

and .
Contravariant Tensor

i
∂ x̄
T̄ = r T r
i
∂x
r i
This relation can easily be inverted to express T in terms of T̄
r
∂x
T = i T̄ i
r
∂ x̄
They are called components of a contravariant vector or contravariant tensor of
the first rank or first order. The components of the contravariant tensors
transform like the coordinate differentials.

One can show easily that velocity and acceleration are contravariant vectors.
Similarly, force, momentum, angular momentum are represented by the
components of first-order tensors.
A parametric graph
The tangent vector field T=(Ti) is defined by the usual
differentiation formula
Covariant Tensor

r
∂x
T̄ i= i T r
∂ x̄
This relation can easily be inverted to express T r in terms of T̄ i

i
∂ x̄
T r = r T̄ i
∂x

They are called components of a covariant vector or covariant tensor of the

first rank or first order.

The gradient of a scalar field is a covariant vector.

 r
∂x
T̄ i= i T r
∂ x̄
Electric field in bar ( Ē ) and unbar ( E) frame

∂ϕ ∂ϕ φ is a scalar quantity
E= and Ē= and is independent of
∂x ∂ x̄
coordinate system.
 r
∂x
T̄ i= i T r
∂ x̄
Electric field in bar ( Ē ) and unbarred (E) frame

i i
∂ϕ ∂ϕ ∂ϕ ∂ x ∂x
E= i
and Ē= i
= i i
Ē = i
E
∂x ∂ x̄ ∂ x ∂ x̄ ∂ x̄
The contravariant
components of a vector
obtained by projecting onto
the coordinate axes.

The covariant components are

obtained by projecting onto the
normal line to the coordinate
hyperplanes.
Contravariant and covariant vectors

The contravariant components of a vector obtained by projecting onto the

coordinate axes.

The covariant components are obtained by projecting onto the normal line to the
coordinate hyperplanes.

Contravariant 1 ∂ x̄ 1 r Covariant
T̄ = r T
∂x r
i
∂ x̄ ∂x
T̄ i= r T r 1 ∂ x̄
1
1 ∂ x̄
1
2 ∂ x̄
1 T̄ i= i T r
T̄ = 1 T + 2 T + 3 T 3 ∂ x̄
∂x ∂x ∂x ∂x i
2 ∂ x̄
2
1 ∂ x̄
2
2 ∂ x̄
2 ∂ϕ ∂ϕ ∂x
T̄ = 1 T + 2 T + 3 T 3 Ē= i = i i
∂x ∂x ∂x ∂ x̄ ∂ x ∂ x̄
(Perpendicular projection)
(Parallel projection)
When the coordinate axes are orthonormal (unit length and are orthogonal) there is
no difference between contravariant and covariant components.
Properties
1. If a tensor is identically equal to zero in one coordinate system it is equal to zero in any coordinate
system.
2. All scalars are not tensors, although all tensors of rank 0 are scalars (e.g., Temperatue vs frequency).
3. All vectors are not tensors, although all tensors of rank 1 are vectors.
(e.g., Momentum and position, the position vector itself is not a tensor as the positioning will change with respect to
other coordinate system, the difference between any two position vectors is a tensor of rank 1!)
4. All dyads or matrices are not tensors, although all tensors of rank 2 are dyads or matrices.
Second Rank Tensor
The need for second rank tensors comes when we need to consider more than one
direction to describe one of the physical properties. A good example of this is if we
need to describe the electrical conductivity of a general, anisotropic crystal. We know
that in general for isotropic conductors that obey Ohm's law:

J=σE

Which means that the current density j is parallel to the applied electric field, E and
that each component of j is linearly proportional to each component of E. (e.g. j 1= σE1).

However in an anisotropic material, the current density induced will not necessarily be
parallel to the applied electric field due to preferred directions of current flow within
the crystal. This means that in general each component of the current density vector can
depend on all the components of the electric field:
So in general, electrical
conductivity is a second
rank tensor and can be
specified by 9 independent
coefficients, which can be
represented in a 3×3 matrix
as shown below:
R μνσρ Reimann curvature tensor
 Ti j
k

right
k i k The δlk are indeed the components of a mixed
T ij T T second-rank tensor, independent of the
wrong j
number of dimensions of our space.
 i s m 2
ij ∂ x̄ ∂ x r ∂ x ∂ x̄
T̄ = r T s
T̄ 12 = 1 T n
∂ x ∂ x̄ j
∂ x̄ ∂ x n m

m 1 2 3 m
∂x ∂x ∂x ∂x ∂ x̄
T̄ 1= 1 T m= 1 T 1 + 1 T 2 + 1 T 3 T 1= T̄
∂ x̄ ∂ x̄ ∂ x̄ ∂ x̄ ∂ x1 m

First rank Contravariant and covariant vectors

Contravariant Covariant
r
∂ x̄
i ∂x
T̄ i= r T r T̄ i= i T r
∂x ∂ x̄
i
∂ϕ ∂ϕ ∂x
Ē= i = i i
∂ x̄ ∂ x ∂ x̄
Q1)

( T i ) =( T 1 ,T 2 ) =( x 2 ,x 1 )
T 1 =x 2 and T 2 =x 1

1 1 2 2
1 ∂ x̄ 1 ∂ x̄ 2 2 ∂ x̄ 1 ∂ x̄ 2
T̄ = 1
T + 2
T and T̄ = 1
T + 2
T
∂x ∂x ∂x ∂x
∂ x̄ 1 ∂ x̄ 1 2 ∂ x̄ 2 2 ∂ x̄ 2 1
=0 , =2 x , =x , =x
∂ x1 ∂ x2 ∂ x1 ∂ x2
Q2)

Answer:

1 1 2 1 2
∂x −x ∂x 1 ∂x 1 ∂x
1
= 2
, 1
= 2
, 2
= 2
, 2
=0
∂ x̄ 2(x )
2 ∂ x̄ x ∂ x̄ 2x ∂ x̄
Second Rank Tensor

essential condition is that the components transform according to the

coordinate transformation laws.
Basic operations with tensors
(3) (Outer or tensor product )

ναβ
=C μλ

The direct product is a technique for creating new, higher-rank tensors.

We may continue this process and multiply more than two vectors together, taking
care that their indexes are all different. In this way we can construct tensors of higher
rank. The total number of free indexes of a tensor is its rank (or order).
 (Scalar or dot product)

In fact any scalar product of two first-order tensors (vectors) is a zero-order tensor (scalar),
as might be expected since it can be written in a coordinate-free way as u · v.
μ
P P μ =constat ( scalar )
 μδ Contraction is another
T νρ

order reducing operation,

μδ μρ μν like the inner product but
for δ=ρ T νρ
=T νρ
=T applying to single tensor.

An operation that produces the opposite effect – namely, generates a tensor of

smaller rather than larger order – is known as contraction and consists of making
two of the subscripts equal and summing over all values of the equalised subscripts.
E is invariant with respect to
coordinate changes.
The product of covariant and contravariant tensors is an invariant in all coordinate systems:

For example product of two vectors in unbarred coordinate system

α μ
∂x μ ∂ x̄
Now Ā μ= μ
A α B̄ = β
Bβ
∂ x̄ ∂x

μ ∂ x α ∂ x̄ μ
Ā μ B̄ = μ β
Aα Bβ
∂ x̄ ∂ x
∂ xα β
= A α B
∂ xβ
α
=δ A α B β
β

α
=A α B

Thus
Energy-Momentum Vector

P μ =( p0 ,p1 ,p2 ,p 3 )= ( E
,p ,p ,p
c x y z )
P μ =( p 0 ,p 1 ,p 2 ,p 3 ) =( E
c
,− p x ,− p y ,−p z )
0 1 E2
2
c
3
(
P P μ =( p p0 +P P 1 +p p 2 +p p 3 )= 2 − p2x −p 2y − p2z
μ
)
E2
P P μ = 2 −p 2
μ
c ( ) 2 2 2
E =p c +m 0 c
2 4

( E2
c 2 )
− p2 =m 20 c 2 =invariant

( Ē 2
c 2
− p̄ 2
)=m 2 2
0 c =invariant
Electro-magnetic field tensor

F μν

μν
F μν F

E2
(
trace ( F μν F )=2 2 −B2
μν
c ) The trace remains same in all inertial frame.
(6) Symmetric and antisymmetric tensors:
The order in which the indices appear in our description of a tensor is important. In
general, Amn is independent of Anm , but there are some cases of special interest.

A tensor is called symmetric with respect to two contravariant or two covariant indices
if its components remain unchanged upon interchange of the indices:

If, for all m and n, Amn = Anm or Amn = Anm , A is symmetric.

A tensor is called anti-symmetric with respect to two contravariant or two covariant

indices if its components change sign upon interchange of the indices:

If, for all m and n, Amn = -Anm or Amn = -Anm, A is antisymmetric.

Symmetry and anti-symmetry can be defined only for similar indices,

not when one index is up and the other is down (for mixed tensor).
Independent components of Symmetric and antisymmetric tensors
A 2nd rank tensor A mn has 32 components in 3-dimensions.

If Amn is a symmetric tensor than A mn = Anm

i.e., A12 = A21, A13=A31, and A23=A32
Out of the 9 components only 6 components are independent, namely A11, A22, A33, A12, A13, and A23

If Amn is a antisymmetric tensor than A mn = -Anm

i.e., A12 = -A21, A13= -A31, A23= -A32
Diagonal components are zero i.e., A 11=A22=A33=0
Out of the 9 components only 3 components are independent, namely A12, A13, and A23
Rules to calculate components:

Symmetric tensor = (Total no. of components + no. of diagonals) / 2

antisymmetric tensor = (Total no. of components - no. of diagonal) / 2

Q) Find the number of independent components of a symmetric and an antisymmetric fourth

rank tensor in 4-dimensions.
Electro-magnetic field tensor

A μ= ( ϕ
,A ,A ,A
c x y z )

∂ A 1 ∂ A 0 ∂ A x ∂ ϕ −1 ∂ ϕ E x
F 01= − = − = =
∂ x0 ∂ x1 c ∂ t c ∂ x c ∂x c
Quotent Law
Arfken-P213