Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
K. M. Udayanandan
Associate Professor
Department of Physics
Nehru Arts and Science College, Kanhangad
1
.
This note is prepared for the Post Graduate Students of Kannur Uni-
versity. The syllabus is
Tensor analysis-Introduction-definition-definition of different rank
tensors-Contraction and direct product-quotient rule-pseudo tensors-
General tensors-Metric tensors
Udayanandan. K. M
Contents
1 Introducing Tensors 5
2.1.1 Scalars . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.1.2 Vectors . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
2.1.3 Tensors . . . . . . . . . . . . . . . . . . . . . . . . . . 18
3 Quotient Rule 21
3
4.1 Spherical Polar Co-ordinate System . . . . . . . . . . . . . . . 26
5.0.2 Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . 32
5.3 Problems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
7 University Questions-Solved 45
Chapter 1
Introducing Tensors
In our daily life we see large number of physical quantities. Tensor is the
mathematical tool used to express these physical quantities. Any physi-
cal property that can be quantified is called a physical quantity. The
important property of a physical quantity is that it can be measured and
expressed in terms of a mathematical quantity like number. For example,
”length” is a physical quantity that can be expressed by stating a number
of some basic measurement unit such as meters, while ”anger” is a property
that is difficult to describe with a number. Hence we will not call ’anger’ or
’happiness’ as a physical quantity. The physical quantities so far identified
in physics are given below. Carefully read them. While reading observe that
some are expressed unbold, some are bold fonted and some are large and
bold. The known physical quantities are absorbed dose rate, acceleration,
angular acceleration, angular speed, angular momentum, area, area
density, capacitance, catalytic activity, chemical potential, molar concentra-
tion, current density, dynamic viscosity, electric charge, electric charge
5
density, electric displacement, electric field strength, electrical con-
ductance, electric potential, electrical resistance, energy, energy density,
entropy, force, frequency, half-life, heat, heat capacity, heat flux density, il-
luminate, impedance, index of refraction, inductance, irradiance, linear
density, luminous flux, magnetic field strength, magnetic flux, magnetic
flux density, magnetization, mass fraction, (mass) Density, mean lifetime,
molar energy, molar entropy, molar heat capacity, moment of iner-
tia, momentum,permeability, permittivity, power, pres-
sure, (radioactive) activity, (radioactive) dose, radiance, radiant intensity,
reaction rate, speed, specific energy, specific heat capacity, specific volume,
spin, stress, surface tension, thermal conductivity, torque, velocity, vol-
ume, wavelength, wave number, weight and work. Every physical quantity
must have a mathematical representation and only then a detailed study of
these will be s possible. Hence we have mathematical tools like theory of
numbers and vectors with which we can handle large number of physical
quantities.
Among the above physical quantities small bold faced quantities are vectors
and un bold are scalars. Generally we say quantities with magnitude only
as scalars and with magnitude and direction as vectors. But there are some
quantities which are given in large font which are not scalars and vectors.
If they are not scalars and vectors what are they? What is special
about these quantities ?. Let us have a look at it. One quality of the
above mentioned odd members is that some like mass, index of refraction,
6
permeability, permittivity sometimes behave as scalars and some times not.
The above mentioned physical quantities like mass, susceptibility. moment
of inertia, permeability and permittivity obey very familiar equations like.
F~ = m~a, P~ = χE,
~ L~ = I~ω , B
~ = µH,
~ D~ = E,
~ F~ = T A,
~ J~ = σ E
~
~v = 6k̂
7
After entering water the velocity is decreased but the direction may not
change. Then the new velocity may be
~v 0 = 3k̂ = 0.5 ~v
and the deviated ball in the water may have different possible velocity like
etc. Consider the first case. The components of the final vector (3,2,5) can
be obtained in different ways. Among them some are given below.
3
3 5
0 0 5
= 1
2 0 3 0 6
5 0 0 58 8
8
This transformation matrix is not diagonal.
3 6
3 −1 5
5 8
2 8
2 = 5 6 −1 6
6
5 1 1 −8 8
Such a 3 × 3 with all elements non-zero can also be used to transform the
old velocity to new one. Or in general
vx0 v v v v
11 12 13 x
0
vy = v21 v22 v23
vy
0
vz v31 v32 v33 vz
This is the most general matrix which can be used to transform the
incident velocity to the new velocity. This shows that any vector can be
transformed to a new vector generally only by a 3 × 3 matrix in 3D. If the
matrix is diagonal and if the diagonal elements are same it becomes a scalar
multiple. We had seen that all our odd physical quantities always transform
one vector to a new vector. Hence the general form of these transforming
quantities must be a matrix with 9 components. Let us check whether this
is true with a specific example. For this let us find out what is the exact
nature of moment of inertia.
9
1.3 Moment of inertia
In terms of ~r and p~
~ = ~r × p~
L
= ~r × m~v
= ~r × m (~ω × ~r)
We’ve
~ ~ ~ ~ ~ ~ ~
A × B × C = A.C B − A.B C ~ ~
= mr2 ω
~ − m~r (~r.~ω )
h i h i
=m x2 + y 2 + z 2 ωx î + ωy ĵ + ωz k̂ −m xî + y ĵ + z k̂ (xωx + yωy + zωz )
~ are as follows,
then the three components of L
Lx = m y 2 + z 2 ωx − mxyωy − mxzωz
Ly = −myxωx + m x2 + z 2 ωy − myzωz
Lz = −mzxωx − mzyωy + m x2 + y 2 ωz
10
Lx = Ixx ωx + Ixy ωy + Ixz ωz
where
x2 + y 2 + z 2 − x2 = m y 2 + z 2
Ixx = m
x2 + y 2 + z 2 − y 2 = m x 2 + z 2
Iyy = m
x2 + y 2 + z 2 − z 2 = m x2 + y 2
Izz = m
L I I I ω
x xx xy xz x
Ly = Iyx Iyy Iyz
ωy
Lz Izx Izy Izz ωz
11
I11 I12 I13
I = I21 I22 I23
I31 I32 I33
Thus
Lx m (y 2 + z 2 ) −mxy −mxz ωx
Ly = 2 2
−myx −myz
m (x + z ) ω
y
Lz −mzx −mzy m (x2 + y 2 ) ωz
12
Chapter 2
We can see that scalars have one(30 ) component, vectors have 3 (31 ) com-
ponets and tensors we saw had 9 (32 ) components. This shows that all these
mathematical quantities belong to a family in 3 dimensional world. Hence
these three types of physical quantities must have a common type of defi-
nition. Since we classify them based on components we will redefine them
based on componets. Hence we can redefine the scalars and vectors using
coordinate transformation of components.
13
sian to spherical polar or cylindrical we call them as non cartesian tensors.
First we will study cartesian tensors.
2.1.1 Scalars
S0 = S
For if we measure mass of some substance say sugar standing straight and
then slightly tilting we will get the same mass. Thus any invariant quantity
under coordinate transformation is defined as a scalar.
2.1.2 Vectors
14
x1 = r cos φ
x2 = r sin φ
.
0
x1 = r cos (φ − θ) = r (cos φ cos θ + sin φ sin θ)
0
x2 = r sin (φ − θ) = r (sin φ cos θ − cos φ sin θ)
0
x1 = x1 cos θ + x2 sin θ
0
x2 = −x1 sin θ + x2 cos θ
In matrix form
0
x1 cos θ sin θ x1
=
0
x2 − sin θ cos θ x2
generally,
0
X
xi = aij xj
j
0
x1 = a11 x1 + a12 x2
0
x2 = a21 x1 + a22 x2
Comparing
a11 = cos θ; a12 = sin θ
15
Here we had taken position vector and performed transformation. Similarly
any vector can be transformed like this. Hence we can say a physical quantity
can be called as a vector if it obeys the transformation equation similar to
the transformation equation
0
X
xi = aij xj
j
0 0
∂x1 ∂x1
= a11 = a12
∂x1 ∂x2
0 0
∂x2 ∂x2
= a21 = a22
∂x1 ∂x2
Thus the transformation equation can also be written as
0
X ∂x0
i
xi = xj
j
∂x j
This is also the transformation equation for a vector. Let us proceed and find
the transformation from primed to unprimed coordinate system. We know
that for transformation from from X to X’
0
x1 cos θ sin θ x1
=
0
x2 − sin θ cos θ x2
0
x1 x1
= A
0
x2 x2
16
where
cos θ sin θ
A=
− sin θ cos θ
Now
cos θ − sin θ
AT =
sin θ cos θ
which means
A−1 = AT
0
x1 x1
A−1 0
= A−1 A
x2 x2
Rearranging
x1 x01
= A−1
x2 x02
x1 cos θ − sin θ x01
=
x2 sin θ cos θ x02
∂x1 0 ∂x1 0
x1 = 0 x1 + 0 x
∂x1 ∂x2 2
∂x2 0 ∂x2 0
x2 = 0 x1 + 0 x
∂x1 ∂x2 2
Then
0
X
xi = aji xj
j
17
Thus we can conclude and say that a vector is a physical quantity which
transform like
0
X
Ai = aij Aj
j
0
X
Ai = aji Aj
j
Both are definitions of a vector. This is very important and you must note
the indices.
2.1.3 Tensors
Ji = σij Ej
18
0
But the component of J transform like Ji = aip Jp . Substituting
0 0 0
Ji = σij Ej = aip Jp
0 0 0
Ji = σij Ej = aip Jp = aip σpq Eq
0 0 0
σij Ej = aip σpq ajq Ej
Then we get
0
σij = aip ajq σpq
Thus we can see that a tensor of rank 2 will have 2 coefficients or the rank
of a tensor can be obtained from counting the number of coefficients.
0
A =A
0
X
Ai = aij Aj
j
0
XX
Aij = aip ajq Apq
p q
19
is a tensor of rank 2 and if have 3 coefficients we will get tensor of rank 3 etc.
20
Chapter 3
Quotient Rule
KA=B
Here A and B are tensors of known rank and K is an unknown quantity. The
Quotient Rule gives the rank of K. For example
~ = I~ω
L
Here ω
~ and L~ are known vectors, then Quotient Rule shows that I is a second
rank tensor. Similarly,
m~a = F~
~ = J~
σE
~ = P~
χE
all establish the second rank tensor of m, σ, χ. The well known Quotient
Rules are
21
Ki Ai = B
Kij Aj = Bi
Kij Ajk = Bik
Kijkl Aij = Bkl
Kij Ak = Bijk
1.
K Ai = B
Proof
Taking prime on both sides
0 0 0
K Ai = B
Here A has one index and hence it is a vector. Using the transformation
22
equation for a vector 0
0 ∂xi
K Aj = B
∂xj
0
because B = B since it is a scalar. Now using the given rule RHS is
modified and we get
0
0 ∂xi
K Aj = KAj
∂xj
0
0 ∂xi
K − K Aj = 0
∂xj
Now Aj cannot be zero since it is a component and if it vanishes the law
itself does not exist. Hence the quantity within the bracket vanishes.
0
∂xi 0
K= K,
∂xj
23
K is a second rank tensor
4.
KAij = Bkl
Proof
Taking prime
0 0 0
K Aij = Bkl
0 0
! 0 0
0 ∂xi ∂xj ∂xk ∂xl
K Aαβ = Bγν
∂xα ∂xβ ∂xγ ∂xν
0 0
! 0 0
0 ∂xi ∂xj ∂xk ∂xl
K Aαβ = KAαβ
∂xα ∂xβ ∂xγ ∂xν
0 0
!
0 ∂xi ∂xj ∂xγ ∂xν
K=K 0 0
∂xα ∂xβ ∂xk ∂xl
5.
K Ak = Bijk
Proof
24
Taking prime
0 0 0
K Ak = Bijk
0 0 0 0
0 ∂xk ∂xi ∂xj ∂xk
K Al = Bαβl
∂xl ∂xα ∂xβ ∂xl
0 0 0 0
0 ∂xk ∂xi ∂xj ∂xk
K Al = KAl
∂xl ∂xα ∂xβ ∂xl
!
0 ∂xα ∂xβ
K=K 0 0
∂xi ∂xj
K is a second rank tensor. The quotient rule is a substitute for the
illegal division of tensors.
Problems
1. The double summation Kij Ai B j is invariant for any two vectors Ai and
Bj . Prove that Kij is a second-rank tensor.
2. The equation Kij Ajk = Bik holds for all orientations of the coordinate
~ and B
system. If A ~ are arbitrary second rank tensors show that K is a
second rank tensor.
25
Chapter 4
Non-Cartesian Tensors-Metric
Tensors
The metric tensor gij is a function which tells how to compute the distance
between any two points in a given space. Its components can be viewed
as multiplication factors which must be placed in front of the differential
displacements dxi in a generalized Pythagorean theorem. We can find the
metric tensor in spherical polar coordinates and cylindrical coordinates.
26
x = r sin θ cos φ
y = r sin θ sin φ
z = r cos θ
Then,
dx = sin θ cos φdr + r cos φ cos θdθ − r sin θ sin φdφ
dy = sin θ sin φdr + r cos θ sin φdθ + r sin θ cos φdφ
dz = cos θdr − r sin θdθ
In Cartesian Coordinate system
ds2 = dx2 + dy 2 + dz 2
ds2 = (sin θ cos φdr+r cos φ cos θdθ−r sin θ sin φdφ)(sin θ cos φdr+r cos φ cos θdθ−r sin θ sin φdφ)+
(sin θ sin φdr+r cos θ sin φdθ+r sin θ cos φdφ)(sin θ sin φdr+r cos θ sin φdθ+r sin θ cos φdφ)+
(cos θdr − r sin θdθ)(cos θdr − r sin θdθ)
ds2 = sin2 θ cos2 φdr2 + r sin θ cos φ cos2 θdr dθ − r sin2 θ sin φ cos φdr dφ+
r sin θ cos θ cos2 φdr dθ + r2 cos2 θ cos2 φ dθ2 − r2 sin θ cos θ sin φ cos φdθ dφ−
r sin2 θ sin φ cos φdφdr − r2 sin θ cos θ sin φ cos φdθ dφ + r2 sin2 θ sin2 φdφ2 +
sin2 θ sin2 φdr2 + r sin θ cos θ sin2 φdr dθ + r sin2 θ sin φ cos φdr dφ+
27
r sin2 θ sin φ cos φdφ dr + r2 sin θ cos θ sin φ cos φdφ dθ + r2 sin2 θ cos2 θdφ2 +
cos2 θdr2 − r sin θ cos θdr dθ − r sin θ cos θdr dθ + r2 sin2 θdθ2
= dr2 + r2 dθ2 + r2 sin2 θdφ2
2 2
dx 1 0 0 dr
dy 2 = 0 r2 0 dθ2
dz 2 0 0 r2 sin2 θ dφ2
ie,
1 0 0 g11 g12 g13
gij = 0 r2 0 = g21 g22 g23
2 2
0 0 r sin θ g31 g32 g33
= g11 drdr+g12 drdθ+g13 drdφ+g21 dθdr+g22 dθdθ+g23 dθdφ+g31 dφdr+g32 dφdθ+g33 dφdφ
X
ds2 = gij dqi dqj
i,j
K Ai = B
K Ai = Bj
K is a second rank tensor and hence gij is a tensor of rank two. Thus gij is
called the metric tensor of rank two.
28
4.2 Cylindrical coordinate system
x = ρ cos φ
y = ρ sin φ
z=z
Differentiating and substituting we get
dx2
2
1 0 0 dρ
dy 2 = 0 ρ2 0 dφ2
dz 2 0 0 1 dz 2
Thus gij for cylindrical coordinate system can be found out.
Problem
In Minkowiski space we define x1 = x , x2 = y , x3 = z , and x0 = ct. This is
done so that the space time interval ds2 = dx20 − dx21 − dx22 − dx23 (c =velocity
of light). Show that the metric in Minkowiski space is
1 0 0 0
0 −1 0 0
(gij ) =
0 0 −1 0
0 0 0 −1
29
Solution
x1 = x ⇒ dx1 = dx
x2 = y ⇒ dx2 = dy
x3 = z ⇒ dx3 = dz
x0 = ct ⇒ dx0 = c dt
= c2 dt2 − dx2 − dy 2 − dz 2
Then
dx20 1 0 0 0 c2 dt2
dx21 0 −1 0 0 dx2
=
dx22 0 0 −1 0 dy 2
dx23 0 0 0 −1 dz 2
1 0 0 0
0 −1 0 0
gij =
0 0 −1 0
0 0 0 −1
30
Chapter 5
Contravariant vector
We’ve X
x0i = aij xj
j
or in general
X ∂xi0
i0
A = Aj
j
∂xj
31
5.0.2 Exercises
32
0
where V i and V j are the components of velocity.
It is a contravariant vector.
0
X ∂xj
Ai = Aj
j
∂x0i
If A & B are tensors of same rank and both expressed in a space of the
same number of dimensions, then
Aij + B ij = C ij
Aij − B ij = Dij
Amn = Anm
33
we call it as a symmetric tensor and if on other hand
Amn = −Anm
B nm = B mn
which is symmetric
C nm = −C mn
which is antisymmetric. So Amn can be represented as a combination of sym-
metric and anti symmetric parts.
5.3 Problems
1. If a physical quantity has no component in one coordinate system, then
show that it does not have a component in other coordinate systems
Answer
We have transformation equations like
ij 0 ∂x0i ∂x0j kl
A = A
∂xk ∂xl
X dxi0
i0
V = Vj
j
dxj
dθ
Vθ = Vy
dy
These equations show that if a physical quantity exist in one coordinate
system it exists in other coordinate system also. Hence if it vanishes in
a system it vanishes in other systems also.
34
2. The components of a tensor A is equal to the corresponding compo-
nents of tensor B in one particular coordinate system. Show that A=B
Answer
Consider transformation equation from X to X’
0
x1 = x1 cos θ + x2 sin θ
0
x2 = −x1 sin θ + x2 cos θ
The product of two tensors is a tensor whose rank is the sum of the ranks
of the given tensors. This product which involves ordinary multiplication of
the components of the tensor is called outer product or direct product.
∂xk
ai 0 = ak
∂xi0
0 ∂xk ∂xj 0
ai b j = ak bl
∂xi0 ∂xl
35
gives a second rank tensor. In general
0 0 0
0 0 0 ∂xi ∂xn ∂xk ∂xl mpq
Aij B kl = Cjikl = C
∂xm ∂x0j ∂xp ∂xq n
0 0 0
where Aij is of rank 2 , B kl is of rank 2 and Cjikl is of rank 4.
Contraction of Tensors
When dealing with vectors, we formed a scalar product by summing products
of corresponding components
~B
A. ~ = A1 B1 + A2 B2 + A3 B3
X
= Ai Bi
i
~B
A. ~ = Ai Bi
∂xl
=
∂xk
= δkl Blk
= Bkk
It is of rank 0. Here,Bkk is scalar. So contraction reduces the rank by two.
Exercises
36
1. Show that Kronecker delta is a mixed tensor of rank 2?
Soln: 0 0 0
0 ∂xi ∂xi ∂xj
δji = 0 = ∂x l ∂xm
∂xj ∂xl ∂xm
0
∂xi ∂xm ∂xl
=
∂xl ∂xj 0 ∂xm
0
∂xi ∂xm l
= δ
∂xl ∂xj 0 m
Thus Kronecker delta is a mixed tensor of rank 2
∂Ai ∂Ai
0 = Ti & = Tj
∂xi ∂xj
then the equation becomes
∂xj
Ti = 0 Tj
∂xi
∂Ai...n
so the rank of ∂xi
0 is n+1. Differentiation increases the rank of a
tensor.
37
Chapter 6
Let us again look into some properties of the vectors and scalars. So far
our coordinate transformations have been restricted to pure passive rota-
tions. We now consider the effect of reflections or inversions. We have seen
in branches of physics like nuclear physics that mirror symmetry is very im-
portant. Now let perform reflection on scalars and vectors. Then we observe
interesting properties . An ordinary vector ~r under reflection remains as
such. Thus
~r = xî + y ĵ + z k̂
under reflection becomes
~r0 = ~r
38
Then ~r is said to be a polar vector which is the name given to an unchanged
vector. If we are considering the components of polar vector, under reflection
there is flip in their sign.
~ =A
C ~×B
~
(C1 , C2 , C3 ) → (C1 , C2 , C3 )
~ 0 = −C
C ~
1
The vector change its sign, components do not. In physics we have such
physical quantities which always appear in the cross product form like
~ = ~r × p~
L
~τ = ~r × F~
~ × ~r
~v = ω
∂B~
=∇×E ~
∂t
39
~ ,~τ , ω, B
Here L ~ are pseudo vectors.
Generally we say
S 0 = J S → P seudoscalar
~ 0 = Jaij C
C ~ j → P seudovector
i
40
The Levi-Civita symbol is not a physical quantity but it is always associated
with some physical quantity. To establish this consider cross product of two
~ and B
vectors A ~
~ =A
C ~×B ~
î ĵ k̂
= A1 A2 A3
B1 B2 B3
C1 = A2 B3 − A3 B2
C2 = A3 B1 − A1 B3
C3 = A1 B2 − A2 B1
The three equation can be generalized as
Ci = εijk Aj Bk
= A2 B3 − A3 B2
Similarly putting i=2, i=3 we can deduce C2 and C3 .
Now we will show that Levi-Civita[LC] symbol εijk is a pseudo tensor?
Proof:If Levi Civita symbol is a pseudo tensor then its transformation equa-
tion is
ε0ijk = |a| aip ajq akr εpqr
For an ordinary vector
x0i = aij xj
For rotation this is
0
x1 cos θ sin θ x1
0 =
x2 − sin θ cos θ x2
41
Which can be written as
x0i = |A|aij xj
where determinant is 1, but in the case of pseudo vectors it will be −1 Then
for i = 1, j = 1, k = 1
1 = |a| a1p a2q a3r εpqr
3
X
= |a| a1p a2q a3r εpqr
pqr=1
Then
X X
a1p a2q a3r εpqr = [a11 a2q a3r ε1qr + a12 a2q a3r ε2qr + a13 a2q a3r ε3qr ]
pqr qr
= a12 a23 a31 ε231 +a13 a22 a31 ε321 +a11 a23 a32 ε132 +a13 a21 a32 ε312 +a11 a22 a33 ε123 +a12 a21 a33 ε213
= a12 a23 a31 − a13 a22 a31 − a11 a23 a32 + a13 a21 a32 + a11 a22 a33 − a12 a21 a33
= a11 (a22 a33 − a23 a32 ) + a12 (a23 a31 − a21 a33 ) + a13 (a21 a32 − a22 a31 )
a11 a12 a13
= a21 a22 a23
a
31 a32 a33
so,
1 = |a| |a| = 1
since by definition of pseudo tensors
|a| = −1
ε0ijk = RHS
which shows that LC satisfies the definition of a pseudo tensor.
Solved Problems
1. Use the antisymmetry of εijk ~ ~ ~
to show that A. A × B = 0
42
Solution:
~ and B
Take cross product of A ~
î ĵ k̂
~×B
A ~ = A1 A2 A3
B1 B2 B3
= −A3 A2 B1 + A2 A3 B1 + A3 A1 B2 − A1 A3 B2 − A2 A1 B3 + A1 A2 B3
εijk Ai Aj Bk = 0
~ A
ie, A. ~×B ~ =0
∂ ∂ ∂
∇ = î + ĵ + k̂
∂x1 ∂x2 ∂x3
∂φ ∂φ ∂φ
∇φ = î + ĵ + k̂
∂x1 ∂x2 ∂x3
î ĵ k̂
∇ × ∇φ = ∂x1 ∂x∂ 2 ∂x∂ 3
∂
∂φ ∂φ ∂φ
∂x ∂x ∂x
1 2 3
∂ ∂φ ∂ ∂φ ∂ ∂φ ∂ ∂φ ∂ ∂φ ∂ ∂φ
= î − −ĵ − +k̂ −
∂x2 ∂x3 ∂x3 ∂x2 ∂x1 ∂x3 ∂x3 ∂x1 ∂x1 ∂x2 ∂x2 ∂x2
∂ ∂
= εijk φ
∂xj ∂xk
43
3
X ∂ ∂ X ∂ ∂ ∂ ∂ ∂ ∂
εijk φ= ε1jk φ + ε2jk φ + ε3jk φ
ijk=1
∂xj ∂xk jk
∂x j ∂xk ∂x j ∂xk ∂x j ∂xk
Expanding we get
=0
44
Chapter 7
University Questions-Solved
3. What are contravarient, co-variant and mixed tensors? Show that ve-
locity and acceleration are contravarient and the gradient of a field is
a covariant tensor.
Consider a set of n quantities A1 , A2 , A3 , ...An in a system of variables
xµ and these quantities have values Ā1 , Ā2 , Ā3 , ...Ān in another system
45
of variablesx̄µ . If these quantities obey the transformation relation
∂ x̄µ α
Āµ = A
∂xα
then the quantities Aα are said to be the components of a contravariant
vector or a contravariant tensor of rank one.
Consider a set of n quantities A1 , A2 , A3 , ...An in a system of variables
xµ and these quantities have values Ā1 , Ā2 , Ā3 , ...Ān in another system
of variablesx̄µ . If these quantities obey the transformation equation
∂xα
Āµ = Aα
∂ x̄µ
then the quantities Aα are said to be components of a covariant tensor
of rank one.
We have
X ∂x0
dx0i = i
dxj
j
∂xj
d2 x0i X ∂x0i ∂ 2 xj
=
dt2 j
∂xj ∂t2
46
by contraction of tensors ,put k = i.
∂ x̄j ∂xd ab
Āij
ij = A
∂xb ∂ x̄j ad
∂xd ab
Āij
ij = A
∂xb ad
Āij d ab
ij = δb Aad
Āij ab
ij = Aab ⇒ Ā = A
is a scalar.
47
Now applying inverse transformation law of dxl and dxm , we get
∂xl i ∂xm j
ḡij dx̄i dx̄j = glm dx̄ . j dx̄
∂ x̄i ∂ x̄
∂xl ∂xm i j
= glm dx̄ dx̄
∂ x̄i ∂ x̄j
∂xl ∂xm
ḡij − glm i dx̄i dx̄j = 0
∂ x̄ ∂ x̄j
As dx̄i and dx̄j are arbitrary contravarient vectors, we must have
∂xl ∂xm
ḡij − glm =0
∂ x̄i ∂ x̄j
ie,
∂xl ∂xm
ḡij = glm
∂ x̄i ∂ x̄j
Which is the transformation law for the second order covariant tensor.
Hence gij is a covariant tensor of rank two.
gij can be expressed as
1 1
gij = (gij + gji ) + (gij − gji ) = Aij + Bij
2 2
48
As dxi and dxj are arbitrary vectors, we have
Bij = 0
ie,
1
(gij − gji ) = 0
2
gij = gji
which shows that gij is symmetric.
ds2 = dx2 + dy 2 + dz 2
where the coefficients gµν are the functions of the coordinates xµ , sub-
ject to the restriction g = determinant of gµν i.e. |gµν | =
6 0.
The quadratic differential form gµν dqµ dqν is independent of the coor-
dinate system and is called the Riemannian metric for n-dimensional
space. The space which is characterised by Riemannian metric is called
Riemannian space. Here the quantities gµν are components of a covari-
ant symmetric tensor of rank two, called the metric tensor or funda-
mental tensor.
49
Transformation law for the tensor Tα is
∂xl
T̄α = Tl
∂ x̄α
∂ x̄α ∂ x̄β mn
S̄ αβ = S
∂xm ∂xn
∂ x̄α ∂xb a
Āαβ = A
∂xa ∂ x̄β b
50
by contraction of tensors ,put β = α.
∂ x̄α ∂xb a
Āαα = A
∂xa ∂ x̄α b
∂xb a
Āαα = A
∂xa b
Āαα = δab Aab
Āαα = Aaa
which is a tensor of rank(2-2) zero.
9. The double summation Kij Ai Bj is invariant for any two vectors Ai and
Bj . Prove that Kij is a second order tensor.
Given that Kij Ai Bj is invariant.then
∂xl ∂xm
K̄ij Āi B̄j = Klm Āi . B̄j
∂ x̄i ∂ x̄j
∂xl ∂xm
= Klm Āi B̄j
∂ x̄i ∂ x̄j
∂xl ∂xm
K̄ij − Klm Āi B̄j = 0
∂ x̄i ∂ x̄j
As Āi and B̄j are arbitrary covariant vectors, we must have
∂xl ∂xm
K̄ij − Klm =0
∂ x̄i ∂ x̄j
51
ie,
∂xl ∂xm
K̄ij = Klm
∂ x̄i ∂ x̄j
Which is the transformation law for the second order covariant tensor.
Hence Kij is a covariant tensor of rank two.
11. If Aµ and Bµ are any two vectors, then prove that Aµ Bµ is invariant.
The transformation law for the vectors are,
µ ∂ x̄µ l
Ā = A
∂xl
and
∂xm
B̄µ = Bm
∂ x̄µ
Then transformation law for Aµ Bµ is
∂ x̄µ l ∂xm
Āµ B̄µ = A Bm
∂xl ∂ x̄µ
∂ x̄µ ∂xm l
Āµ B̄µ = A Bm
∂xl ∂ x̄µ
∂xm l
Āµ B̄µ = A Bm
∂xl
Āµ B̄µ = δlm Al Bm
Āµ B̄µ = Al Bl
52
Hence Aµ Bµ is invariant tensor.
12. Show that any tensor of rank 2 can be expressed as the sum of a
symmetric] and an antisymmetric tensors of rank 2.
Any tensor Aµν of rank 2, may be expressed as
1 µν 1
Aµν = (A + Aνµ ) + (Aµν − Aνµ )
2 2
= B µν + C µν
where B µν = 12 (Aµν + Aνµ ) and C µν = 12 (Aµν − Aνµ )
From addition and subtraction laws of tensor it follows that B µν and
C µν are tensors of rank 2.
Interchanging indices in B µν and C µν , we get
1 νµ 1
B µν = (A + Aµν ) = (Aµν + Aνµ ) = B µν
2 2
and
1 νµ 1
C νµ = (A − Aµν ) = − (Aµν + Aνµ ) = C µν
2 2
Which shows that B is symmetric, while C µν is antisymmetric, both
µν
13. Write down the transformation rule for 2 rank contravarient and co-
variant tensors. Show that contraction of a 2 rank tensor result in an
invariant.
For the 2 rank contravarient tensor Aij , transformation rule is
∂ x̄i ∂ x̄j mn
Āij = A
∂m ∂xn
For the 2 rank covariant tensor Bµν , transformation rule is
∂xl ∂xm
B̄µν = Bab
∂ x̄µ ∂ x̄ν
∂ x̄µ ∂xm l
Āµν = A
∂xl ∂ x̄ν m
53
by contraction of tensors ,put ν = µ.
∂ x̄µ ∂xm l
Āµµ = A
∂xl ∂ x̄µ m
∂xm l
Āµµ = A
∂xl m
Āµµ = δlm Alm
Āµµ = All
hence which is invarient.
54
then the contravariant tensor Aµν or covariant tensor Aµν of second
rank is antisymmetric.
16. ω
~ is any arbitrary contravarient vector. It is known that Aij ω ~ is a
covariant vector (summation convention is used). Show that Aij is a
covariant tensor of rank 2.
Given that Aij ω
~ is a covariant vector and ω
~ is any arbitrary contravari-
ent vector. then
ω
~ = ωj
Aij ω
~ j = Ci
Ci is a covariant vector.
A~ωj = Ci −→ (1)
17. Using the inner product of a tensor and applying contraction principle
obtain the length L of a tensor Ai .
Ai is a vector.
consider Aj , and taking direct product of Ai and Aj
Ai Aj
Ai Aj
Then p
L= Ai Aj
18. Obtain the metric tensor for two dimensional plane in polar coordi-
nates.
55
The metric tensor gµν is given by
In polar coordinates
ds2 = dr2 + r2 dθ2
If x1 = r , x2 = θ then comparing above equations, we get
g11 = 1
g12 = g21 = 0 = 0
g22 = r2
The metric tensor gµν in matrix form is written as
g11 g12 1 0
gµν = =
g21 g22 0 r2
19. Obtain the law of transformation for the elements of a third rank tensor
corresponding to a coordinate transformation.
56