SP 5
SP 5
Solution:
2 5
Aij :
1 3
′ ′
cos θ sin θ
p11 = cos θ p12 = sin θ
rotation matrix: ⇒
− sin θ cos θ ′ ′
p21 = − sin θ p22 = cos θ
2 X
X 2
′ ′
i′ j ′
pjj Aij pii pjj
′ ′
ij
A =A pii =
i=1 j=1
′ ′ ′ ′ ′ ′ ′ ′ ′ ′
A1 1 = A11 p11 p11 + A12 p11 p12 + A21 p12 p11 + A22 p12 p12
′ ′ ′ ′ ′ ′ ′ ′ ′ ′
A1 2 = A11 p11 p21 + A12 p11 p22 + A21 p12 p21 + A22 p12 p22
1
′ ′ ′ ′ ′ ′ ′ ′ ′ ′
A2 1 = A11 p21 p11 + A12 p21 p12 + A21 p22 p11 + A22 p22 p12
′ ′ ′ ′ ′ ′ ′ ′ ′ ′
A2 2 = A11 p21 p21 + A12 p21 p22 + A21 p22 p21 + A22 p22 p22
5.2. Evaluate Aij Bij , where Aij is the tensor from problem 1 above and the
components of Bij are
4 0
.
2 7
Solution:
2 X
X 2
ij
A Bij = Aij Bij
i=1 j=1
= 8 + 0 + 2 + 21
= 31
2
Solution:
and
X N
N X N
X
Aij δij = Aij δij = Aii
i=1 j=1 i=1
In the last step, we made use of the fact that δij = 1 when i = j and is zero
when j 6= i. Thus, all but one of the terms in the sum over j vanish.
Solution:
Since Aij
klm is a tensor,
′ ′ ′
Aik′jl′ m′ = Aij i j k l m ′
klm pi pj pk ′ pl′ pm′
′ ′ ′ ′ ′
Bkj ′ m′ = Ahk′ hj ′ m′ = Aik′jl′ m′ δil′
′
′
= Aij i j k l m l ′ ′
klm pi pj pk ′ pl′ pm′ δi′
′
= Aij i j k l m ′
klm pi pj pk ′ pi′ pm′
′
= Aij j k m ′
l i
klm (pi′ pi ) pj pk ′ pm′
′
= Aij l j k m
klm δi pj pk ′ pm′
′
= Alj j k m
klm pj pk ′ pm′
′
j
= Bkm pjj pkk′ pm
m′
3
5.5. Consider Euclidean 2-space and a rotation of the axes through angle θ.
Suppose Aij is a rank 2 tensor defined for this space and transformation and
that its components are given by Aij = i + j. Also define B = Aii . Verify
explicitly that B is a scalar.
Solution:
′ ′
p21 = − sin θ = p12′ , p22 = cos θ = p22′
′ ′ ′
B ′ = Aii′ = A11′ + A22′
2 X
X 2
Aij p1i pj1′
′ ′
A11′ =
i=1 j=1
′ ′ ′ ′
= A11 p11 p11′ + A12 p11 p21′ + A21 p12 p11′ + A22 p12 p21′
′ ′ ′ ′ ′
A22′ = A11 p21 p12′ + A12 p21 p22′ + A21 p22 p12′ + A22 p22 p22′
= 2(− sin θ)2 + 3(− sin θ) cos θ + 3 cos θ(− sin θ) + 4 cos2 θ
′ ′
B ′ = A11′ + A22′ = 2 + 0 + 0 + 4 = 6 = B
4
5.6. What is the metric tensor for Euclidean 3-space?
Solution:
(
1 , i=j
d~s 2 = dx2 + dy 2 + dz 2 ⇒ gij = δij =
0 , i 6= j
Solution:
g i j = g ik gkj = δ i j
In the first equality, g ik raises the index k on gkj (or, alternatively, gkj lowers
the index k on g ik ). The second equality follows from the definition of g ik .
5.8. Given that gµν = diag(-1, -1, -1, 1), show that g µν = gµν .
Solution:
α = 1 ⇒ −g 11 = 1 ⇒ g 11 = −1
α = 2 ⇒ −g 12 = 0 ⇒ g 12 = 0
α = 3 ⇒ −g 13 = 0 ⇒ g 13 = 0
α = 4 ⇒ g 14 = 0
5
Similarly when µ = 2, 3.
α = 1 ⇒ −g 41 = 0 ⇒ g 41 = 0
α = 2 ⇒ −g 42 = 0 ⇒ g 42 = 0
α = 3 ⇒ −g 43 = 0 ⇒ g 43 = 0
α = 4 ⇒ g 44 = 1
We have specified all the components of g µν and found diag(-1, -1, -1, 1).
If the problem had said “verify” rather than “show”, then we could just treat
g µν gνα as a matrix multiplication and verify that we get the identity matrix
δij :
−1 0 0 0 −1 0 0 0 1 0 0 0
0 −1 0 0 0 −1 0 0 0 1 0 0
0 0 −1 0 = 0 X
0 0 −1 0 0 1 0
0 0 0 1 0 0 0 1 0 0 0 1
Solution:
6
The other transformation coefficients are all zero.
4
X
= − g3ν A1ν
ν=1
= −(−A13 ) = A13
′ ′ ′ ′
(c) A4 1 = Aµν p4µ p1ν
4
X ′ ′ ′ ′ ′
= p1ν [A1ν p41 + A2ν p42 + A3ν p43 + A4ν p44 ]
ν=1
4
X ′
= p1ν [−γβ A1ν + γ A4ν ]
ν=1
′ ′
= p11 [−γβ A11 + γ A41 ] + p14 [−γβ A14 + γ A44 ]
|{z} |{z}
γ −γβ
7
5.10. If T µν is a 4-tensor, show that under a spatial rotation T 44 remains
unchanged, T 4i and T i4 transform as 3-vectors, and T ij transforms as a 3-
tensor.
Solution:
Note: It’s not correct to multiply the transformation matrix times a tensor
to get the tensor in the new coordinate system. That only works for vectors.
We need 2 transformation matrices for a rank 2 tensor and it’s not just simple
multiplication.
′ ′ ′ ′
T µ ν = pµµ pνν T µν
′ ′ ′ ′
T 4 4 = p4µ p4ν T µν = (1) (1) T 44 = T 44 X
8
If it helps to see it written out in detail, here it is:
′ ′ ′ ′ ′ ′ ′ ′ ′ ′
T4 4 = p41 p41 T 11 + p41 p42 T 12 + p41 p43 T 13 + p41 p44 T 14
′ ′ ′ ′ ′ ′ ′ ′
+p42 p41 T 21 + p42 p42 T 22 + p42 p43 T 23 + p42 p44 T 24
′ ′ ′ ′ ′ ′ ′ ′
+p43 p41 T 31 + p43 p42 T 32 + p43 p43 T 33 + p43 p44 T 34
′ ′ ′ ′ ′ ′ ′ ′
+p44 p41 T 41 + p44 p42 T 42 + p44 p43 T 43 + p44 p44 T 44
The are 16 terms; all are zero except for the last one.
′
T 4 j = p4µ pjν T µν = pjν T 4ν = pji T 4i = Rj i T 4i
′ ′ ′ ′ ′ ′
X
A B
′ ′
In step A, we made use of p4i = 0 and p44 = 1. In step B, we made use
′
of pj4 = 0. The final result is the rotation matrix acting on a 3-component
object since i = 1, 2, 3.
′
T j 4 = pjµ p4ν T µν = pjµ T µ4 = pji T i4 = Rj i T i4
′ ′ ′ ′ ′ ′
X
′
T i j = piµ pjν T µν = pii pjj T ij = Ri i Rj j T ij
′ ′ ′ ′ ′ ′ ′
X