Lagrangian and Hamiltonian

Formalism of General Relativity

Reefat

Memorial University of Newfoundland

[email protected]

July 4, 2018
 Overview

• Hamiltonian solution for an aymptoticaly flat spacetime

• Asymptotically flat solutions of Einstein equation
• ADM mass for aymptotically flat spacetime
• Angular momentum of a gravitational field
• What does the ”Mass” and ”Angular momentum” means
• An Example for stationary, axially symmetric spacetimes
• Komar Formula
• Bondi-Sachs Mass
Hamiltonian solution for an aymptoticaly flat spacetime

• In a classical system, the gravitational mass is related to the

total energy of the system, hence the Hamiltonian H.
• Therefore, we assume that the same is true for any spacetime.
• The Action A defined in the last class admits two different
solutions for the Hamiltonian.
• For compact manifold, the solution is

HGsolution = 0

• For a non-compact manifold, the solution is

1
I h i√
HGsolution = − N(k − k0 ) − Na (K ab − Khab )rb σd 2 θ
8π St
 Non-compact manifold

• The non-compact solution of the Hamilton’s equation depend

on the asymptotic behavior of lapse and shift.
• Lapse and shift are arbitrary which poses problem when we are
trying to relate it to time invariant quantity.
• We need to fix this.
• We know that spacetime should be aysmptotically flat, i.e. Σt
should coincide with a surface of constant time in
Minkowskian or Lorenzian spacetime, L.
• Coordinates in L are (t, x, y , z)
• We want asymptotic portion of Σt to coincide with a surface
of t = const.
 Non-compact manifold (conti.)
• In the asymptotic portion, the arbitrary coordinates y a of Σt
are related to the spatial L, i.e. we have the asymptotic
relation
y a 7→ y a (x, y , z)
• Similarly, we also have

x α 7→ x α (t, x, y , z)
• t is the proper time for an observer at rest in the asymptotic
region.
α
• The observer moves with a four-velocity u α = ∂x
∂t
.
• Because this vector is normalized and orthogonal it should
coincide with the normal vector nα which fix our asymptotic
relation
∂x α
nα 7→
∂t
 Non-compact manifold (conti.)

• We know

t α = Nnα + N α eaα
 α  α
∂x α ∂x
=N +N
∂t ∂y a

• Once this is fixed, the asymptotic behaviour of Σt is specified

and there is a 1 − 1 correspondence between lapse and shift
and a choice of flow vector(above).
 ADM Mass

The mass, M is defined to be the limit of HGsolution when St is a

2-sphere at the spatial infinity evaluated with the following choice
of lapse and shift which is

N=1 Nα = 0

• Then using the equation

1
I h i√
HGsolution = − N(k − k0 ) − Na (K ab − Khab )rb σd 2 θ
8π St
 ADM Mass(conti.)

• We get the mass function to be

√
I
1
M=− lim (k − k0 ) σd 2 θ
8π st 7→∞ St

• σAB is the metric on St

• k AB is the extrinsic curvature on St and k = σ AB kAB
• k0 is the extrinsic curvature of St embedded in flat spacetime.
• The quantity M is called the ADM mass of the gravitational
field.
 Angular Momentum

• The definition of mass with a particular choice of lapse and

∂x α
shift implies that asymptotically t α 7→ ∂t
.
• The flow vector generates asymptotic time translation.
• The ADM mass is the gravitational Hamiltonian for this
choice of flow vector.
• Since time translation gives us energy or mass.
• Therefore, rotational symmetry should give us angular
momentum, i.e.
∂x α
t α 7→ φα ≡
∂φ
• For this we should have the choice of lapse and shift, N = 0
∂y a
and N α = φα = ∂φ respectively.
 Angular Momentum (conti.)

The angular momentum, J of an asymptotically flat spacetime is

defined to be the limit of the HGsolution when St is a 2-sphere at
spatial infinity which is
√
I
1
J=− lim (Kab − Khab )φa r a σd 2 θ
8π St 7→∞ St

Here,
• Kab is the extrinsic curvature.
• r a is the normal vector
• φa is our specific choice of shift.
 Does this make sense?

We will start with the following spacetime:

4j sin2 θ
   
2m 2m
ds 2 = − 1 − dt 2 + 1 + (dr 2 +r 2 dω 2 )− dtdφ
r r r
(1)
Here:
• j is a generic angular momentum
• m is a generic mass
We want to show that m = M and j = J which were the definition
we derived in the earlier slide.
Asymptotically flat, stationary and axially-symmetric
spacetime

The metric in the last slide is an asymptotically flat spacetime that

is both stationary and axially symmetric.
• We will choose a hypersurface Σt to be surfaces of constant
time, t.
m


• We define the normal vectornα = − 1 − r ∂α t
• The induced metric on 3-hypersurface is
2m
dr 2 + r 2 dΩ2



hab = 1 + r
• The boundary St is the two- sphere at r = R with normal
m


vector r = 1 + r ∂a r
• The induced metric on the 2-hypersurface is
A B 2m

2 2
σAB dθ dθ = 1 + r R dΩ
 Evaluating M

• We evaluate k = r a |a = 2 2m


R 1− R

2
• We the evalute k0 = = R2 1
R0 −m R ,
the extrinsic curvature
of a two-sphere of identical intrinsic geometry but embedded
in a flat spacetime.
0 =σ
• By identical we mean σAB AB
• We calculate k − k0 = − 2m
R2
and
√ 2m
σd 2 θ = R 2 1 + sin2 θdθdφ


R
• Finally, by substituting in our mass-function

√
I
1
M=− lim (k − k0 ) σd 2 θ
8π st 7→∞ St

we see that
m=M
 Evaluating J
• At first, we calculate Kab φa r a = Kφr (1 − m
r ) where
Kab = nα;β eaα ebβ
• The non-vanishing components of the metric are
 
tt 2m 2j
g =− 1+ g tφ = − 3
r r
which gives
3j sin2 θ
Γtφr = −
r2
2
• Thus we get Kφr = − 3j sin
R2
θ
and finally by substituting this in
√
I
1
J=− lim (Kab − Khab )φa r a σd 2 θ
8π St 7→∞ St
we get
j =J
 Komar Formula
There is an alternative method of calculating the mass, M and
angular momentum, J. The alternative definition is given by
Komar Formula. The definitions are:
1
M=− lim ∇α ξbβ dSαβ (2)
8π St 7→∞
1
J=− lim ∇α ξφβ dSαβ (3)
16π St 7→∞
where:

ξbβ is the timelike Killing vector

ξφβ is the rotational Killing vector

The surface element is given by

√
dSαβ = −2n[α rβ] σd 2 θ

where nα and rα are the timelike and spacelike normal to St .

Komar formula is equivalent to Hamilton’s definition

• To prove that they are equivalent we need to prove the

following relations:
2m
− 2∇α ξtβ nα rβ = − = k − k0
r2
∇α ξφβ nα rβ = Kab φa r b

• The relations are proved in page-150 of the book

• Thus, we see that for stationary and axially symmetric
spacetimes, the Komar formula is equivalent to our
Hamiltonian definitions for the mass and angular momentum.
Komar formula is equivalent to Hamilton’s definition
(conti.)

• We can Stoke’s Theorem on Komar formula to show that

Z  √

1
M=2 Tαβ − Tgαβ nα ξtβ hd 3 y
2
ZΣ  
√
1
J = −2 Tαβ − Tgαβ nα ξφβ hd 3 y
Σ 2

• If Σ is composed of more than one boundary there will be

more terms in RHS of the above equation.
• Only in the case of stationary, axially symmetric spacetimes
that M and J can be defined as hypersurface integrals
because that is when Komar integrals can be used to calculate
the mass ad angular momentum.
 Bondi-Sachs mass

• The ADM mass was defined keeping t fixed and taking the
limit at spatial infinity.
• There exists another way of reaching infinity, which is by
taking null infinity instead of spatial infinity.
• We can then define a new mass called called Bondi-Sachs
mass.
 Bondi-Sachs mass (conti.)
• We introduce the null coordinates u = t − r (retarded time)
and v = t + r (advance time).
• In this coordinates, a two-surface of constant t and r becomes
a surface of constant u and v .
• Null infinity corresponds to the limit v 7→ ∞ keeping u fixed
• Thus, the Bondi-Sachs mass is
√
I
1
MBS (u) = − (k − k0 ) σd 2 θ (4)
8π S(u,v 7→∞)
• This definition is important because of its physical importance
in gravitating body that radiates.
When an isolated body emits radiation, the rate of change of
MBS (u) is directly related to the outward flux of radiated energy. I
F denotes this flux, then the Bondi-Sachs mass satify:
√
I
dMBS
=− F σd 2 θ (5)
du S(u,v 7→∞)
 Distinction between ADM and Bondi-Sachs

• ADM mass and Bondi-Sachs mass are the same for stationary
spacetimes.
• For dynamical spacetimes, these are different.
• Bondi-Sachs mass decrease according to the formula defined
earlier.
• ADM mass does not change even for dynamic spacetimes.
Distinction between ADM and Bondi-Sachs (conti.)
• We will consider Schwarzschild metric but with
r
u = t − r − 2M ln( 2M − 1) as the coordinates.
• The mass parameter M 7→ m(u)
• The metric takes the form

ds 2 = −fdu 2 − 2dudr + r 2 dΩ2

where f = 1 − 2m(u)/r
• The energy-momentum tensor takes the form:
dm/du
Tαβ = − lα lβ
4πr 2
where lα = −∂α u is tangent to the radial, outgoing null
geodesics.
• When the strong energy conditions are satisfied, m
decreases with increasing retarded time.
• This spacetme is known as the outgoing Vaidya metric.
 Calculating ADM for Vaidya spacetime
• We choose a spacelike hypersurface,asymptotically flat, and
coincide with t = const. of Minkowskian metric.
• A suitable choice is to let Σ be a surface of constant
t = u + r for which the unit normal

nα = −(2 − f )−1/2 ∂α (u + r )

• The induced metric is

hab dy a dy b = (2 − f )dr 2 + r 2 dΩ2

• We then choose 2-sphere ar r = R with the unit normal

vector ra = (2 − f )1/2 dα r
• We then follow all the calculations earlier and end up with:

MADM (t) = m(−∞)

which shows us that ADM mass is constant.

 ADM and Bondi-Sachs conclusion

• ADM mass of a dynamical system is constant.

• Bondi-Sachs mass represents all the mass present on a null
hypersurface of constant u.
• The hypersurface fails to intersect any of the radiation that
was emitted prior to the retarded time u.