Добавил:
Опубликованный материал нарушает ваши авторские права? Сообщите нам.
Вуз: Предмет: Файл:

Supersymmetry. Theory, Experiment, and Cosmology

.pdf
Скачиваний:
79
Добавлен:
01.05.2014
Размер:
12.6 Mб
Скачать

Exercise 10 Show that (C.81) and (C.84) are invariant under:

δS φi = 2 η ψi

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

µ

α˙

Dµφi

 

 

 

 

 

 

 

 

 

δS ψ= 2 ηαFi i 2 σαα˙

η¯

 

 

 

 

 

µ

Dµψi

 

¯a aj

 

 

 

 

 

 

δS Fi = −i 2 η¯σ¯

 

2¯λ ti φj

a

 

 

 

¯a

− η¯σ¯µλ

a

 

 

 

 

δS Aµ

= ησµλ

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

i

 

 

 

ν )αβ ηβ Fµνa

δS λαa = −ηαDa

 

 

(σµσ¯

 

2

δS D

a

µ

 

¯a

+ ¯σ¯

µ

 

a

.

 

= iησ

Dµλ

 

 

Dµλ

Exercises 449

(C.99)

Hints: If one compares with the individual transformations of the chiral and vector supermultiplets (i.e. (C.29) and (C.70) with ordinary derivatives replaced by covariant derivatives), one identifies a single extra term (the second term in δS Fi). To check its

form, cancel all terms proportional to F i in the supersymmetry transformation of the

i

δS Fi + g

 

¯a

¯

aj

φj .

 

Lagrangian. They originate from only two terms: F

 

2λ

δS ψi ti

Problem 1 One of the problems of supersymmetry is the absence of conventions to which all authors would adhere. In this exercise, we will reformulate some of the expressions in the appendix to make explicit the conventions chosen here and in other textbooks. You may choose to prove part or all of the following equations, or just use them when you try to compare with other textbooks.

The first convention is one of metric signature. We introduce here a sign which is +1 for a metric signature (+, −, −, −) (our convention) and 1 for a metric signature (−, +, +, +). In the computations necessary for what follows, the main consequence of the choice of metric signature is the formula:

µ, γν } = 2 ηµν

 

which may be written

 

 

 

(σµσ¯ν + σν σ¯µ)

β

= 2 ηµν δβ

,

 

α

α

 

σµσν + σ¯ν σµ)α˙ ˙

= 2 ηµν δα˙˙ .

 

β

β

 

Thus, one has

Tr (σµσ¯ν ) = 2 ηµν , and, using (B.61) of Appendix B,

(θσ

µ ¯

ν ¯

1

θ

2 ¯2

.

θ)(θσ

θ) =

2

θ

We start by constructing supersymmetry charges which satisfy:

¯

} = 2 Q

µ

i∂µ

 

{Qα, Qα˙

σαα˙

 

 

 

¯

¯

}

{Qα, Qβ } = 0 = {Qα˙

, Qβ˙

where Q is ±1.

(C.100)

(C.101)

(C.102)

(C.103)

(C.104)

(C.105)

450 Superfields

 

 

 

 

The translation generator has the general form:

 

Pµ =

i

µ,

(C.106)

γ

P

 

 

 

 

 

 

 

with γP a normalization factor8.

Following the method of Section C.1, we define the superspace translation by:

µ µ ¯

S (y , η, η¯) = exp i γP y Pµ + γQ ηQ + γ ¯ η¯Q . (C.107)

Q

Then, one obtains, besides (C.106),

 

 

 

 

 

 

i

 

 

Qα =

 

 

 

 

 

 

+ Q σααµ ˙ θ¯α˙ µ

γQ

∂θα

Q¯α˙ =

 

i

 

 

+ Q θασααµ ˙ µ

 

 

 

γQ¯

∂θ¯α˙

where we have defined

ρQ Q γQ γ ¯ ,

Q

which we will take to be real.

F ¯

The supersymmetry transformation on a superfield (x, θ, θ) reads

& '

F ¯ α ¯α˙ F ¯

δS (x, θ, θ) = Q η Qα ¯ η¯α˙ Q (x, θ, θ).

Q

The next step is to find covariant derivatives which satisfy:

(C.108)

(C.109)

(C.110)

¯

} = 2 D

µ

i∂µ

 

 

¯

¯

}

{Dα, Dα˙

σαα˙

{Dα, Dβ } = 0 = {Dα˙

, Dβ˙

 

 

¯

¯

¯

¯

} = 0,

 

{Dα, Qβ } = {Dα, Qβ˙

} = {Dα˙

, Qβ } = {Dα˙

, Qβ˙

 

with D = ±1. One finds:

 

 

 

Dα =

λD

 

 

 

 

µ

θ¯α˙ µ

 

 

 

 

 

 

 

 

 

− iρQ σαα˙

,

 

 

 

γQ

∂θα

 

 

 

D¯α˙ =

λ ¯

 

 

 

 

 

µ

 

 

 

 

 

 

D

 

 

 

 

 

 

 

 

 

 

 

γQ¯

 

∂θ¯α˙

− iρQ θασαα˙

µ

,

where λ

D

and λ ¯

are two normalization constants which satisfy:

 

D

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

λ

 

λ

¯ =

D

.

 

 

 

 

 

 

 

 

 

 

 

D

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

D

 

Q

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

We now turn to the chiral superfield. Defining the variable

 

 

 

 

 

y

µ

= x

µ

− iρQ θσ

µ ¯

 

 

 

 

 

 

 

 

 

θ,

 

 

(C.111)

(C.112)

(C.113)

(C.114)

8Note that, depending on the respective values of Q and γP , one sometimes does not recover from (C.105) the standard anticommutation relations (C.7), i.e. if Q γP = 1. This is indeed the case of the most common convention, the one of Wess and Bagger [362].

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Exercises 451

we have

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Qα =

i

 

 

 

,

 

 

 

 

 

 

 

 

 

 

 

γ

Q

 

∂θα

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

i

 

 

 

 

 

 

 

 

 

 

 

 

Q¯α˙ =

 

 

 

 

 

+ 2Q σµθ)α˙

 

 

 

,

 

(C.115)

γQ

∂θ¯α˙

∂yµ

 

 

 

 

λD

 

 

 

 

 

 

 

 

 

Dα =

 

 

 

 

 

 

2Q σµθ¯ α

 

,

 

 

γQ

 

∂θα

∂yµ

 

 

¯

α˙

=

λ ¯

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(C.116)

 

D

 

 

 

 

 

.

 

 

 

 

 

 

 

 

 

 

D

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

¯

 

 

 

 

 

 

 

 

 

 

 

 

 

γQ¯ ∂θα˙

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

¯

α˙

Φ = 0 as

Thus, we can write a chiral superfield Φ which satisfies the condition D

 

Φ = φ(y) + α

 

 

θψ(y) + α

θ2 F (y),

 

(C.117)

ψ

2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

F

 

 

 

 

 

 

 

where αψ and αF are real normalization constants.

The supersymmetry transformations (C.110) read in terms of component fields:

δS φ = αψ 2 η ψ

√ √

αψ δS ψα = 2 αF F ηα − i 2 ρQ (σµη¯)α µφ

¯α˙

 

α˙

 

 

 

µ

η)

α˙

µφ

 

 

 

 

 

αψ δS ψ

= 2 αF F η¯

− i 2 ρQ σ

 

 

 

 

 

 

µ

 

 

 

 

 

 

 

µ

 

 

 

 

 

 

 

 

2ρQ αψ η¯σ¯

µψ

(C.118)

αF δS F = i 2ρQ αψ µψσ

η¯ = −i

 

or in four-component notation (see Exercise 2 for notation and Appendix B for our conventions on gamma matrices):

 

 

δS A = αψ ¯Ψ, δS B = αψ ¯5Ψ,

 

 

 

αψ δS Ψ = αF (F1 − iγ5F2) −iρQ γµµ (A + 5B) ,

 

 

(C.119)

αF δS F1 = −ρQ αψ ¯γµi∂µψ, αF δS F2 = −ρQ αψ ¯γ5γµµΨ.

As for the Lagrangian, we have

 

 

 

 

 

 

 

 

 

 

 

 

Φ(x)Φ(x) θ2θ¯2 = αF2 F (x)F (x) + ρQ2

21

µφ (x)µφ(x) 41

φ(x) φ (x)

 

1

 

 

i

 

2

ψσ

µ

¯

 

µ

¯

4 φ (x) φ(x) +

2

 

ρQ

αψ

 

µψ − ∂µψσ

ψ ,

and

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

∂W

1 2W

 

 

 

 

 

 

W (Φ))|θ2 = αF F

 

 

(φ(y)) − αψ2

 

 

 

 

(φ(y)) ψα(y)ψα(y).

(C.120)

 

Φ

2

Φ2

452 Superfields

Thus the action for the scalar field reads (up to total derivatives which do not contribute)

S =

d4xd2θd2θ¯ Φ(x, θ, θ¯)Φ(x, θ, θ¯) + d4y d2θ W (Φ(y, θ)) + h.c.

 

 

d4x ( ρQ2 µφ∂µφ −

 

dW

 

2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

2

 

 

 

 

2

 

 

 

=

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

(C.121)

 

 

 

 

 

 

µψ −

2 ψ

 

+

2

 

 

+ 2

αψ

Q

ψσ

 

µψ + ψσ¯

 

 

ψ

,

 

 

1

2

 

 

 

µ

 

¯ ¯

µ

 

 

d

W

2

 

d

W

¯2

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

where we have solved for the auxiliary field:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

1

 

dW

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

F =

 

 

 

.

 

 

 

 

 

 

 

 

 

(C.122)

 

 

 

 

 

 

αF

 

 

 

 

 

 

 

 

 

One may write the preceding action using four-component spinors (see Exercise 2 for notation). For example, in the case where

 

 

 

 

 

 

W (Φ) = 1 mΦ2

+ 1

λΦ3

,

 

(C.123)

 

 

 

 

 

 

 

 

2

3

 

 

 

 

 

 

we have9

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

S =

 

 

 

 

 

 

 

 

 

 

 

 

 

d4x $ ρQ2 µφ∂µφ − mφ + λφ2

 

2

 

− λΨL φ ΨR

 

+ αψ

 

2 Q

Ψγ

 

µΨ

2 mΨΨ − λΨR φΨL

. (C.124)

 

2

 

1

 

¯

µ

 

1

¯

 

¯

 

¯

 

This can be shown directly to be invariant under the supersymmetry transformations (C.119).

Finally, we write the abelian vector superfield decomposition as:

 

 

 

 

 

 

 

 

¯

 

 

i

 

2

[M + iN ]

 

 

i

¯2

[M

− iN ]

 

 

 

 

V = C + iθχ − iθχ¯ +

 

2

θ

 

 

 

 

2

θ

 

 

 

 

+ α

 

ρ

 

θσµθA¯

µ

 

 

θ2θ¯

 

 

α

λ¯α˙

+

 

i

σµ∂ χ)α˙

 

 

 

(C.125)

 

 

 

 

 

 

2

 

 

 

 

A

 

Q

 

 

 

 

 

Q

α˙

 

λ

 

 

 

 

 

 

 

µ

 

 

 

 

 

.

+ Q θ¯2θα αλ λα +

 

i

(σµµχ¯)α

 

1

 

ρQ2

θ2θ¯2 D +

1

C

 

 

 

 

 

 

 

2

 

 

2

2

The corresponding gauge invariant superfield reads

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

λ ¯

 

 

 

 

 

 

 

 

 

 

 

 

 

β

 

 

 

 

 

 

 

 

 

 

µν

 

β

 

 

 

 

 

 

 

 

 

 

D

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Wα = D

 

λ

λα

ρQ µ δαD + A (σ

 

)α

 

Fµν

θβ

 

γQ¯

 

 

(C.126)

 

 

 

 

 

 

+ ρ

 

α

θ2σ

αα˙

µ

λ¯α˙

 

,

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Q

 

 

λ

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

9Note that, for ρQ = 1 (as in Wess and Bagger [362]), we obtain an unusual form for the Dirac equation: µµΨ = −mΨ. It is to recover the standard Dirac equation that we have chosen ρQ = +1,

and thus real γQ , γQ¯ (see (C.109)).

Exercises 453

whereas the complete supersymmetry transformations are

 

δS C = i (ηχ − η¯χ¯) ,

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

δS χα = ηα(M + iN ) − iρQ (σµη¯)α (αA Aµ − i∂µC) ,

 

α

δ

A

 

= i α λσµη¯

 

 

α

ησµλ¯

+ η∂

χ + η∂¯

 

χ,¯

 

 

 

 

A S

 

 

 

µ

 

 

 

 

λ

 

 

 

λ

 

¯

 

 

µ µ

 

 

 

 

µ

 

µ

η¯) ,

(C.127)

 

δ

 

M =

 

ρ

 

 

 

α

ηλ + α

η¯λ + i (ησ ∂ χ¯

 

 

χσ

 

S

 

N =

 

Q

λ

 

 

 

 

 

λ

 

 

¯

 

 

µ µ

 

µ

 

µ

 

 

 

δ

 

 

Q

α

 

ηλ α

η¯λ + i (ησ ∂

χ¯ +

 

χσ

 

η¯) ,

 

 

S

 

 

 

 

 

λ

 

ρα

λ

 

 

 

 

 

 

 

µ

 

 

 

µ

 

 

 

 

α δ

 

λ

α

= i ρ η D

A

(σµν η)

α

F

µν

,

 

 

 

 

 

 

 

 

 

λ S

 

 

 

 

Q

 

α

 

λ¯

Q

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

δ

S

D = α

ησµ

α

 

η¯σ¯µ

λ .

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

λ

 

 

µ

 

 

 

λ

 

µ

 

 

 

 

 

 

 

 

 

 

 

 

 

 

The conventions followed by various authors are summarized in the following table.

Authors:

HERE Wess-Bagger

Sohnius

Derendinger

YOUR

 

 

[362]

[342]

[101]

CONVENTIONS

 

 

 

 

 

 

 

 

 

+1

1

+1

 

+1

 

5

+1

i

i

 

 

Q

+1

+1

+1

 

+1

 

γP

+1

1

+1

 

1

 

γQ

+1

i

+1

 

1

 

γQ¯

+1

i

+1

 

1

 

ρQ

+1

1

+1

 

+1

 

D

1

1

+1

 

1

 

λD

+1

i

+1

 

+1

 

λD¯

1

i

+1

 

1

 

αψ

+1

+1

2

 

+1

 

αF

+1

+1

1

1

 

αA

+1

+1

1

 

 

αλ

−i

+1

+1

 

 

 

Appendix D

An introduction to cosmology

Since the metric signature often chosen in general relativity is di erent from the one adopted in this book, we introduce in this appendix a sign which is +1 for a metric signature (+, −, −, −) (our convention) and −1 for a metric signature (−, +, +, +).1

D.1 Elements of general relativity

In general relativity, matter curves spacetime and the central dynamical object is the metric gµν (x) which depends on the spacetime coordinates x0, x1, . . . , xD−1 (to be general, we consider in this section D-dimensional spacetimes). Physics should not depend on the way we define the spacetime coordinates; in other words, it should be invariant under general coordinate transformations xµ → x µ(x). Since the elementary line element

ds2 = gµν dxµdxν

(D.1)

should be invariant, the infinitesimal transformation

 

dx µ =

∂x µ

dxν

(D.2)

 

 

∂xν

 

implies the following transformation law for the metric tensor gµν and its inverse gµν

g

(x ) =

∂xρ

 

∂xσ

g

 

(x), g µν (x ) =

∂x µ

 

∂x ν

gρσ(x),

(D.3)

 

 

ρσ

 

 

µν

∂x µ ∂x ν

 

∂xρ ∂xσ

 

 

 

 

 

which are the respective transformation laws of a covariant and a contravariant tensor. One may define similarly the transformation laws of a contravariant vector V ν or

covariant vector Vν :

V ν =

∂x ν

V ρ, V =

∂xρ

V

 

,

(D.4)

 

 

ρ

 

∂xρ

ν

∂x ν

 

 

 

 

 

 

 

However, the spacetime derivatives of vectors do not transform covariantly. For example,

∂V ν

=

∂x ν ∂xσ ∂V ρ

+

2x ν ∂xσ

V ρ.

 

 

 

 

 

 

 

 

 

 

∂x µ

 

∂xρ ∂x µ ∂xσ

∂xρ∂xσ ∂x µ

 

 

 

 

1In the notation of Misner, Thorne and Wheeler [287], our conventions are as follows: for g sign, +1 for Riemann sign, and +1 for Einstein sign.

Elements of general relativity 455

The situation is reminiscent of the one encountered in gauge theories. An ordinary derivative µΨ of a gauge nonsinglet field Ψ does not transform covariantly under a gauge transformation. One has to introduce the covariant derivative DµΨ = µΨ − igAaµtaΨ which has the same gauge transformation as Ψ (see Section A.1.4 of Appendix Appendix A).

Similarly here one defines the covariant derivatives

DµV ν = µV ν + Γν µλV λ, DµVν = µVν Γρµν Vρ,

(D.5)

where Γρµν , the analog of the gauge field, is called a Christo el symbol or spin connection (it does not transform as a tensor under general coordinate transformations) and is defined in terms of the metric as:

Γρµν = 21 gρσ [µgνσ + ν gµσ − ∂σ gµν ] .

(D.6)

One can then check that DµV ν transforms as a (mixed) tensor:

 

(DµV ν ) =

∂xρ

 

∂x ν

DρV σ .

(D.7)

 

 

 

∂x µ ∂xσ

 

In the same way that one defines the field strength by di erentiating the gauge field (Fµν = µAν − ∂ν Aµ − ig [Aµ, Aν ]), one introduces the Riemann curvature tensor:

Rµναβ = αΓµνβ − ∂β Γµνα + ΓµασΓσνβ ΓµβσΓσνα.

(D.8)

By contracting indices, one also defines the Ricci tensor Rµν and the curvature scalar R

Rµν ≡ Rαµαν , R ≡ gµν Rµν .

(D.9)

It is easy to show that

 

[Dµ, Dν ] V ρ = Rρσµν V σ ,

(D.10)

which is reminiscent of a similar equation in gauge theories (see (A.51) of Appendix Appendix A).

An important notion connected with covariant di erentiation is the notion of parallel transport. A vector is said to be parallel-transported along a curve parametrized by τ if its covariant derivative projected along the curve is zero:

DV µ

 

dxλ

 

 

≡ DλV µ

 

= 0.

(D.11)

For example, the velocity U µ = dxµ/dτ (τ proper time: 2 = gµν dxµdxν ) of a freely falling object is parallel-transported along the free fall trajectory (this generalizes the notion of zero acceleration in flat space). Such a path is called a geodesic. In other words, a geodesic is a curve that parallel transports its tangent vector along itself. On a sphere the geodesics are the great circles.

456 An introduction to cosmology

Let us now restrict to a 2n-dimensional compact manifold K with metric gij and consider a small loop γ in the (xi, xj ) plane at a point P . Take a vector V k at P and parallel transport it along γ. If the space is flat, when one reaches P again the vector recovers its previous value. This is not true in curved space and V k is changed by the amount (cf. (D.10)):

δV k Rklij V l.

(D.12)

Let us restrict our attention to vectors tangent to K and note TP the tangent space at point P . The holonomy group at point P is the set of transformations hγ that associate to V TP the tangent vector V TP obtained by parallel transporting V along the loop γ through P (each loop determines one element of HP ).

Using the equation of parallel transport (D.11) together with (D.5), one can write (δV k = −dxiΓkij V j ), using Stokes’ theorem,

V = hγ V = exp γ Γdx V = exp S Rds V,

(D.13)

where S is a surface delimited by the loop γ. It is straightforward to check the group structure of HP (take two loops γ1 and γ2 through P : hγ1 hγ2 = hγ1γ2 ). Let us note that (D.13) shows that the spin connection can be interpreted as the gauge field associated with the holonomy group (Γkij = Ai kj ).

Generally, for a complex manifold, HP SO(2n), independently of P . Take the example of the sphere S2 and consider first the loops γ which consist of a triangle made of (parts of) three great circles. Since these are geodesics it is easy to parallel transport a tangent vector V . By changing the angles of this triangle, one may generate any rotation in the tangent plane, hence any transformation of SO(2).

The action of general relativity in D dimensions has the form

S

=

dDx

 

 

1

(R

2 λ) +

S

matter (gµν , Ψ

· · ·

)

(D.14)

 

g

 

2κD2

 

 

"| |

 

 

 

 

 

where κ2D = 8πG(D) (κ2 = 8πGN in the four-dimensional case), g is the determinant of gµν , Ψ stands for the matter fields and λ is the cosmological constant. Using

2

 

δSmatter

= T µν ,

(D.15)

 

 

 

 

"|g|

δgµν

 

one derives (see Exercise 1) from the Lagrangian (D.14) the equations of motion of the metric components, i.e. Einstein’s equations:

Gµν ≡ Rµν 21 gµν R = κD2 Tµν + λgµν ,

(D.16)

where the tensor Gµν thus defined is called the Einstein tensor. In four dimensions,

Gµν = 8πGN Tµν + λgµν .

(D.17)

Friedmann–Robertson–Walker Universes 457

D.2 Friedmann–Robertson–Walker Universes

Under the assumption that the Universe is homogeneous and isotropic on scales of order 100 Mpc (1 pc = 3.262 light-year = 3.086 × 1016 m)2, one may try to find an homogeneous and isotropic metric as a solution of Einstein equations. The most general ansatz is, up to coordinate redefinitions, the Robertson–Walker metric:

ds2 = c2dt2 − a2(t) γij dxidxj ,

 

(D.18)

 

dr2

2 + sin2 θdφ2

,

 

γij dxidxj =

 

+ r2

(D.19)

1 − kr2

where a(t) is the cosmic scale factor, which is time-dependent in an expanding or contracting Universe. The constant k which appears in the spatial metric γij can be chosen to be equal to ±1 or 0: the value 0 corresponds to flat space, i.e. usual Minkowski spacetime; the value +1 to closed space (r2 < 1) and the value 1 to open space. Note that r is dimensionless whereas a has the dimension of a length. Physical distances (we will say proper distances) may be measured at time t along rays of constant θ and φ as

d(t) =

0r

"

|grr|dr = a(t) 0r 1 drkr 2 .

(D.20)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

The components of the Einstein tensor now read (see Exercise 2):

 

 

G00 = 3

 

a˙ 2

+

k

,

 

 

 

 

(D.21)

 

a2

a2

 

 

 

 

 

Gij = γij

a˙ 2

+ 2aa¨ + k ,

 

 

(D.22)

where we use standard notation: a˙ is the first

time derivative

of the cosmic scale factor,

a¨ the second time derivative.

For the energy–momentum tensor, we follow our assumption of homogeneity and

isotropy and assimilate the content of the Universe to a perfect fluid:

 

Tµν = − pgµν + (p + ρ)UµUν ,

(D.23)

where Uµ is the velocity 4-vector (U 0 = 1, U i = 0). It follows from (D.23) that T00 = ρ and Tij = a2ij . The pressure p and energy density ρ usually satisfy the equation of state:

p = wρ.

(D.24)

The constant w takes the value w 0 for nonrelativistic matter (negligible pressure) and w = 1/3 for relativistic matter (radiation). In all generality, the perfect fluid consists of several components with di erent values of w.

2To keep in mind orders of magnitude, the visible disk of a typical spiral galaxy has radius 10 kpc, a typical halo has radius larger than 50 kpc, and a typical intergalactic distance is 6 Mpc.

458 An introduction to cosmology

One now obtains from the (0, 0) and (i, j) components of the Einstein equations (D.17):

3

 

a˙ 2

+

k

= 8πGN ρ + λ,

(D.25)

a2

a2

a˙ 2 + 2aa¨ + k = 8πGN a2p + a2λ.

(D.26)

D.2.1 Friedmann equation

The first of the preceding equations can be written as the Friedmann equation, which gives an expression for the Hubble parameter H ≡ a/a˙ measuring the rate of the expansion of the Universe:

H2

a˙ 2

=

1

(λ + 8πGN ρ)

k

(D.27)

 

 

 

 

.

a2

3

a2

Note that the cosmological constant appears as a constant contribution to the Hubble parameter.

This equation should be supplemented by the conservation of the energy–momentum tensor which simply yields:

ρ˙ = 3H(p + ρ).

(D.28)

Hence a component with equation of state (D.24) has its energy density scaling as ρ a(t)3(1+w). Thus nonrelativistic matter (often referred to as matter) energy density scales as a3. In other words, the energy density of matter evolves in such a way that ρa3 remains constant. Radiation scales as a4 and a component with equation of state p = −ρ (w = 1) has constant energy density. The latter case corresponds to a cosmological constant as can be seen from (D.26)–(D.27) where the cosmological constant can be replaced by a component with ρΛ = −pΛ = λ/(8πGN ).

The Friedmann equation allows us to define the Hubble constant H0, i.e. the present value of the Hubble parameter, which sets the scale of our Universe at present time. Because of the troubled history of the measurement of the Hubble constant, it has become customary to express it in units of 100 km s1 Mpc1 which gives its order of magnitude. Present measurements give

H0

h0 100 km s1 Mpc1 = 0.7 ± 0.1.

The corresponding length and time scales are:

H0

c

= 3000 h01 Mpc = 9.25 × 1025 h01 m,

(D.29)

 

H0

tH0

1

= 3.1 × 1017 h01 s = 9.8 h01 Gyr.

(D.30)

 

H0