Professional Documents
Culture Documents
Jake Bian
September 30, 2015
Contents
1 Lorentz Group Orbits
3
3
5
5
5
6
Consider vector space V ' R4 , let V denote its dual vector space. Let ( )
denote a basis for V , and (dx ) the dual basis in V . Define a symmetric
2-tensor V V as
= dt dt +
3
X
dxi dxi
(1.1)
i=1
(1.2)
(v, v) = (v, v)
(1.3)
(1.4)
(1.5)
3
X
(1.6)
i=1
(1.7)
(1.8)
Now assume v 0 > 0 - that is, v is some point on Bv above the t = 0 plane.
We also restrict ourselves to the component of G connected to the identity. The
following observations follow from the pictures above
When (v, v) > 0, there exist smooth 1-dimensional curves on Bv which
connect v, a point above the t = 0 plane, to a point v 0 , below the t = 0
plane. Since Bv is the orbit of the Lorentz group, these 1-dimensional
curves are the orbits of smooth 1-parameter groups of Lorentz transformations starting from the identity taking v 7 v 0 with v 00 < 0.
1 These
When (v, v) < 0, there are no curves connecting the upper half of the
hyperbloid to the bottom half. Therefore there cannot be smooth 1parameter families of Lorentz transformation which takes v below the
t = 0 plane. There are no other transformations that can take v below
the plane because were restricted to the component of the Lorentz group
connected to the identity, in particular reflection across a plane is not a
smooth transformation, hence lies on another component of the Lie group
G.
This is the desired result.
(2.1)
= [ (v v)] = ( )[v v]
(2.2)
= ( )(v, v)
(2.3)
(2.4)
(2.5)
as desired.
3.1
Skew-Symmetry
(3.1)
= 1 + (1 + 1) + O(2 )
(3.2)
compute
Plug this into result of the last problem 2.5, dropping the O(2 ) term we
find
(1 + (1 + 1)) =
(3.3)
Therefore
(1 + 1) = 0
(3.4)
(3.5)
= (1 )[u v] + (1 )[u v]
(3.6)
= (u, v) + (u, v)
(3.7)
Let denote the metric transpose (or metric pullback) of . That is, is
a linear operator on V such that
( u, v) = (u, v)
(3.8)
(3.9)
(3.10)
(3.11)
where weve labelled the dummy indices so that we can factor out the vector components easily. The definition of the metric pullback 3.8 says the two
quantities above are equal, hence
b ba = ab b
(3.12)
In the usual physicists notation the metric lowers an index, the equation
becomes
a = a
(3.13)
(3.14)
This is the desired result in component form. One can apply the metric
inverse twice to convert this last equation into an identical result about tensors
with pure-vector (upper) indices.
3.2
4.1
a
b
V V , for wich a basis is {dx dx }, and in these coordinates we label the
components of the tensor as
= ab dxa dxb
(4.1)
1 : V V V V
(4.2)
(4.3)
1 [
] = ab dxa [dxb ] = ab b dxa = a dxa (4.4)
where we took the usual definition of the wedge product between a V and
b V as
a b a b b a
(4.5)
(4.6)
= dx [x ] b dx [x ] a
a
= x b x a
(4.7)
(4.8)
(4.9)
(4.10)
Observe the metric inner product in the first component is invariant under
the following shifting of indices
ab (xa b xb a ) 7 ab (xa b xb a )
(4.11)
This is desired result. To get the exact form of the result as suggested in
the problem statement, recall is related to an element of G by Lie algbera
exponentiation, infinitesmally we can consider a Lorentz group element infinitesmally close to the identity
= 1 +
(4.12)
(4.13)
(4.14)
4.2
(4.15)
(4.16)
= (x a x a )b + (x b x b )a
(4.17)
(4.18)
(4.21)
[Ji , Jj ] =
(5.1)
(5.2)
Rewrite the first term using an identity for the levi-civita symbol
1
1
iab jcd ac M bd = icb jcd M bd = (cbi cdj )M bd
4
4
1
1
1
= (bd ij bj id )M bd = Mbb ij M ji = M ij
4
4
4
(5.3)
(5.4)
where in the last equality we used the fact that M ab is skew symmetric, and
therefore has vanishing diagonals. Repeating this for each term in the original
expression and taking the sum, we find
[Ji , Jj ] = M ij
(5.5)
Now observe
1
kab kcd M cd
2
1
= (ac bd ad bc )M cd
2
1
= (M ab M ba ) = M ab
2
abk Jk =
(5.6)
(5.7)
(5.8)
(5.9)
Consider commutator
1
1
iab [M ab , M j0 ] = iab (( aj M b0 a b) j 0)
2
2
the first term looks like
[Ji , Kj ] =
(5.10)
1
1
1
iab ( aj M b0 ) = ijb M b0 = ijb Kb
(5.11)
2
2
2
the j i part vanishes because each term is multiplied by a factor like a0 ,
which vanishes because a > 0. The b a part yields an identical term. Hence
[Ji , Kj ] = ijb Kk
(5.12)
0i
0i
= (g M ) (g M
00
=g M
ij
= M
ij
0j
00
ij
+g M )
= ijk Jk
(5.13)
(5.14)
(5.15)
where terms like gi0 vanish because i, j > 0 and gab is diagonal. We also
used the fact g 00 = 1 for flat Minkovski metric.
(6.1)
where
Z
H0
d3 x a Aa|x
(6.2)
h
2 + U (x) and the interaction part is
where A(x) 2m
Z
H1 d3 xd3 y V (x y)a (x)a (y)a(y)a(x)
(6.3)
We first verify the usual commutators of the number operator with the ladder
operators
Z
[N, a(x)] =
Z
=
Z
=
(6.4)
(6.5)
d3 y (x y)a(y) = a(x)
(6.6)
identically
[N, a (x)] =
Z
Z
=
Z
=
(6.7)
d3 ya (y)[a(y), a (x)] + 0
(6.8)
a (y)(y x) = a (x)
(6.9)
where we used the known commutation relations between the ladder operators. Also observe
[N, A] = 0
(6.10)
where A was defined as above, where the term involving the derivative vanishes because N has no x dependence. Now compute
Z
[N, H0 ] =
d x [N, a Aa] =
(6.11)
d3 x a A(a) + a Aa = 0
=
=
(6.13)
To compute the remaining term, we use a trick. Make the following rearrangement using commutation relations
a (x)a (y)a(y)a(x) = a (x)a (y)a(x)a(y)
(6.14)
(6.15)
(6.16)
(6.17)
(6.18)
d x n(x) and
Z
[N, n(x)] =
d3 y [n(y), n(x)] = 0
(6.19)
Z
H1 =
(6.20)
(6.21)
(6.22)
(6.23)
(6.24)
Finally
Z
[N, H1 ] = V (0)[N, N ] +
Hence
[N, H] = [N, H0 + H1 ] = [N, H0 ] + [N, H1 ] = 0
10
(6.26)