Relativistic Hydrodynamics

Luciano Rezzolla
Institute for Theoretical Physics, Frankfurt am Main, Germany

Olindo Zanotti
Laboratory of Applied Mathematics, University of Trento, Italy

1
Errata/Corrige

Last update: 19 April 2018

1
A Brief Review of General Relativity

• Page 28, Eq. (1.115). Change

0 0 0
vx + V W vy W vz
vx = , vy = , vz = .
1 + v x0 V 1 + v x0 V 1 + v x0 V
→
0 0 0
vx + V vy vz
vx = , vy = , vz = .
1 + v x0 V W (1 + v x0 V ) W (1 + v x0 V )

• Page 32, Eq. (1.140). Change

LφV T = φLV T ,

LφV T = φLV T − V LT φ ,

• Page 32, Eq. (1.141). Change

dφ
L V φ = V ν ∂ ν φν = ,
dλ
→
dφ
LV φ = V ν ∂ν φ = ,
dλ
• Page 33, four lines before Eq. (1.147). Change
...not all bases are such that eµ · eν 6= ηµν
→
...not all bases are such that eµ · eν = ηµν
• Page 38, Eq. (1.174). Change

Lη ξ = Lξ η = 0 ,

→
Lη ξ = −Lξ η = 0 ,
 A Brief Review of General Relativity vii

• Page 43, Eq. (1.196). Change

d2 (xµ + ξ µ ) d(xα + ξ µ ) d(xβ + ξ µ )
2
+ Γµαβ (x + ξ) = 0.
dλ dλ dλ
→
d2 (xµ + ξ µ ) d(xα + ξ α ) d(xβ + ξ β )
2
+ Γµαβ (xµ + ξ µ ) = 0.
dλ dλ dλ
• Page 44, Eq. (1.202). Change

Rα[βγδ] = 2(Rαβγδ + Rαδβγ + Rαγδβ ) = 0 .

→
3!Rα[βγδ] = 2(Rαβγδ + Rαδβγ + Rαγδβ ) = 0 .
• Page 45, Eq. (1.207). Change

Γrθθ = − sin θ cos θ , Γθrθ = cot θ .

Γθφφ = − sin θ cos θ , Γφθφ = cot θ .

• Page 45, Eq. (1.208). Change

1 1
Rrθθr = − gθθ = − sin2 θ , Rrθθr = grr = 1 ,
RS2 RS2
→
1 1
Rθφφθ = − gφφ = − sin2 θ , Rφθφθ = gφφ = 1 ,
RS2 RS2
• Page 45, after Eq. (1.208). Change
while the Ricci scalar is simply R = 1/RS2 .
→
while the Ricci tensor is Rij = gij /RS2 and the Ricci scalar is simply given by R =
2/RS2 .
• Page 48, Eq. (1.220). Change
 
1 1
Rµν = 8π Tµν − T gµν + Λgµν ,
2 4π
→  
1 1
Rµν = 8π Tµν − T gµν + Λgµν ,
2 8π
• Page 49, one line after Eq. (1.223). Change
...the coordinate time runs slower than the proper time.
→
...the proper time runs slower than the coordinate time.
viii A Brief Review of General Relativity

• Page 51, Eq. (1.229). Change

" −1 # "   2 #
2
d 2M dr dθ 2 dφ
1− =r + sin θ ,
dτ r dτ dτ dτ

→
" −1 # " 2  2 #
d 2M dr dθ 2 dφ
1− = r + sin θ
dτ r dτ dτ dτ
" 2  −2  2 #
M dt 2M dr
− 2 + 1− ,
r dτ r dτ
2
A Kinetic-Theory Description of
Fluids

• Page 77, Eq. (2.30). Change

Z
S(t) := −kB V H(t) = −kB V ~,u
f (t, x ~ ) ln(f (t, x ~ )) d3 u ,
~,u

→ Z
S(t) := −kB V H(t) = −kB V ~ )) d3 u ,
~ ) ln(f (t, u
f (t, u

• Page 79, after Eq. (2.39). Change

“...through the evolution of the momentum flux ρvj (i.e., the rate of change of linear
momentum per unit time and unit area), ...”
→
“...through the flux of the momentum density tensor ρvi vj + Pij (i.e., the rate of change
per unit time and unit area orthogonal to the i-th direction of the j-th component of the
linear momentum), ...”
• Page 80, after Eq. (2.42). Change
“...unlike the kinetic energy, 12 ρv i vi ,...”
→
“...unlike the kinetic energy density, 12 ρv i vi ,...”
• Page 81, Eq. (2.47). Change

3 kB T 2 
= , p= = nkB T ,
2 m 3 nm
→
3 kB T 2
= , p= n m  = nkB T ,
2 m 3
• Page 82. Change Eq. (2.51)

3kB T 3k T
u2 i =
h~ ui2 = B − ~v 2 ,
− h~
m m
→
3kB T 3k T
u2 i =
h~ ui2 = B + ~v 2 ,
+ h~
m m
x A Kinetic-Theory Description of Fluids

• Page 83, Eq. (2.54). Change

3/2
mu2
  
m
f0 (u) = 4πnu2 exp − .
2πkB T 2kB T
→ 3/2
mu2
  
m
4πu2 f0 (u) = 4πnu2 exp − .
2πkB T 2kB T
• Page 84, Fig. 2.5. Change the labels on the axes: v → u.
• Page 84, Eq. (2.59). Change
Z
mn m
T = u − ~v )2 f0 d3 u =
(~ u − ~v )2 i .
h(~
3kB 3kB
→ Z
m m
T = u − ~v )2 f0 d3 u =
(~ u − ~v )2 i .
h(~
3nkB 3kB
• Page 84, last paragraph of Sect. 2.2.4. Change
“(Problem 1 is dedicated to showing...)”
→
“(Problem 4 is dedicated to showing...)”
• Page 86, after Eq. (2.70). Change
“...represents the flux of energy per unit surface and unit time, i.e., the Newtonian energy
flux density vector.”
→
“...represents the flux of energy per per unit time and unit area, i.e., the Newtonian energy-
density flux vector.”
• Page 90, after Eq. (2.82). Change
“...and recalling that p0x = 0 in the local Lorentz rest frame...”
→
“...and recalling that px0 = 0 in the local Lorentz rest frame...”
• Page 90, last line. Change
“The relativistic Maxwell–Boltzmann equation can then be obtained...”
→
“The relativistic Boltzmann equation can then be obtained...”
• Page 91, Eq. (2.88). Change
p
K := (p1 )α (p2 )α − m4 c4 .

→ p
K := (p1 )α (p2 )α − m2 c2 .
• Page 91, after Eq. (2.88). Change
“Note that the collisionless Maxwell-Boltzmann equation, namely (2.86) with...”
→
“Note that the relativistic collisionless Boltzmann equation, namely (2.86) with...”
 A Kinetic-Theory Description of Fluids xi

• Page 95, second line of Sec. 2.3.4. Change

“...we multiply the relativistic Maxwell–Boltzmann equation (2.86) by...”

→
“...we multiply the relativistic Boltzmann equation (2.86) by...”
• Page 95, after Eq. (2.110). Change

“...can be transformed into a volume integral in momentum space...”

→
“...can be transformed into a surface integral in momentum space...”
• Page 97, second line of Sec. 2.3.6. Change

“...is a solution of the relativistic Maxwell–Boltzmann equation (2.86)...”

→
“...is a solution of the relativistic Boltzmann equation (2.86)...”
• Page 107, Eq. (2.168). Change
 
∂p
c2s := ,
∂e s
→
 
∂p
c2s := c 2
,
∂e s

• Page 108, Eq. (2.171). Change

1 2
c2s = (c ) .
h s N
→
c2 2
c2s = (c ) .
h s N
• Page 108, Eq. (2.175). Change
 
3 3
G > c2s , G < c2s .
2 2
→
3 c2s 3 c2s
 
G > , G < .
2 c2 2 c2
• Page 108, Eq. (2.172) and (2.173). Change
  
1
dp d ln h
c2s = = , (2.1)
h
dρ s d ln ρ s
"    # "    #
1 ∂p d ∂p 1 ∂p p ∂p
= + = + . (2.2)
h ∂ρ  dρ ∂ ρ h ∂ρ  ρ2 ∂ ρ
→
xii A Kinetic-Theory Description of Fluids

c2
   
dp d ln h
c2s = =c 2
, (2.3)
h dρ s d ln ρ s
"  # "    #
c2 c2
 
∂p d ∂p ∂p p ∂p
= + = + . (2.4)
h ∂ρ  dρ ∂ ρ h ∂ρ  ρ2 ∂ ρ

• Page 116, Eq. (2.232). Change

p(5ρh − 8p)
c2s = ,
3ρh(ρh − p)
→

p(5ρh − 8p)
c2s = c2 ,
3ρh(ρh − p)
• Page 117, Eq. (2.234). Change
h − c2
 
γ(γ − 1) γp
c2s = 2 = (γ − 1) = .
c + γ h ρh
→
c2 γ(γ − 1) c2 (h − c2 )(γ − 1) c2 γp
c2s = = = .
c2 + γ h ρh
• Page 118, footnote 34. Change

“A fluid obeying the ideal-fluid equation of state with  = 0 would also have a zero
temperature and could provide a reasonable model for a cold and old neutron star.”
→
“A fluid obeying a general polytropic equation of state can have, at least mathematically,
 = 0, although such a choice would be difficult to justify from a physical point of
view. However, if the polytropic transformation is isentropic, then the specific internal
energy is fully determined and is proportional to the rest-mass density [cf. Eq. (2.248)
and discussion around it]. A polytropic and isentropic equation of state is often used to
obtain a reasonable approximation of the description of matter of a cold and old neutron
star.
• Page 119, after Eq. (2.247)
Put differently, a polytropic equation of state is equivalent to an ideal-fluid equation of
state only under those isentropic transformations for which the adiabatic index of the
fluid γ is the same as the adiabatic index of the polytrope Γ.”
→
“Put differently, if a fluid obeys the ideal-fluid equation of state and is isentropic, then
its equation of state can also be written in a polytropic form [cf., Eq. (2.242)], with
polytropic exponent Γ = γ; in this case, the polytropic exponent is also the adiabatic
index. On the other hand, if a fluid obeys the polytropic equation of state and is isentropic,
then it is at least formally possible to express the pressure as p = ρ(Γ − 1) [cf., Eq.
(2.228)]. However, this does not necessarily mean that such a fluid obeys an ideal-fluid
equation of state. For this to be the case, Γ must be the ratio of the specific heats cp /cV
and the specific internal energy must be a function of the temperature only.”
 A Kinetic-Theory Description of Fluids xiii

• Page 119, Eq. (2.249). Change

 −1
Γp Γ(Γ − 1)p 1 1
c2s = = = + .
ρh ρ(Γ − 1) + Γp ΓKρΓ−1 Γ−1
→
 −1
2 Γp Γ(Γ − 1)p 1 1
c2s =c =c 2
= c2 + .
ρh ρ(Γ − 1) + Γp ΓKρΓ−1 Γ−1
3
Relativistic Perfect Fluids

• Page 139, before Eq. (3.28). Change The simplest quantity to determine is the rest-mass
density current, namely → The simplest quantity to determine is the rest-mass current,
namely
• Page 139, before Eq. (3.28). Change

J µ̂ : flux of rest-mass current density in the µ̂-direction,

→
J µ̂ : flux of rest-mass in the µ̂-direction,
• Page 139, before Eq. (3.29). Change

T 0̂0̂ : total energy density,

T 0̂î : flux of energy density in î-th direction,
T î0̂ : flux of î-momentum in 0̂-th direction (î-momentum density),
T ĵ î : flux of ĵ-th component of momentum density in î-th direction .
→
T 0̂0̂ : total energy density,
T 0̂î : flux of energy in î-th direction,
T î0̂ : flux of î-momentum in 0̂-th direction ,
T ĵ î : flux of ĵ-momentum in î-th direction .
• Page 139, (3.29). Change the second line as follows:
T 0̂î = T 0̂î = 0 ,
→
T 0̂î = T î0̂ = 0 ,
• Page 146, Eq. (3.69). Change
1
Lu (huµ ) = − ∇µ p − ∇µ h .
ρ
→
1
Lu (huµ ) = − ∇µ p = −∇µ h .
ρ
 Relativistic Perfect Fluids xv

• Page 146, Eq. (3.71). Change

ξ µ ∇µ p 1
Lu (huµ ξ µ ) = − − ξ µ ∇µ h = − Lξ p − Lξ h ,
ρ ρ
→

ξ µ ∇µ p 1
Lu (huµ ξ µ ) = − = − Lξ p = −Lξ h ,
ρ ρ
• Page 146, before Eq. (3.72). Change
“and thus use the condition (3.65) with Lu p = 0 and Lu h = 0, to finally obtain”
→
“and thus use the condition (3.65) with Lξ p = 0 to finally obtain”
• Page 146, after Eq. (3.72). Change
Note the similarity between expression (3.72) and the corresponding equation (1.185)
along geodesic trajectories, i.e., Lu (uµ ξ µ ).
→
Note the similarity between expression (3.72) and the corresponding equation (1.185)
along geodesic trajectories, i.e., Lu (uµ ξ µ ) = 0.
• Page 155, caption of Fig. 3.4. Change
... Show with blue solid lines
→
... Shown with blue solid lines

• Page 177, Eq. (3.253). Change

Z
PRαβ := Iν N α N β dνdΩ .

→
Z
PRαβ := hαγ hβδ TRγδ = hαγ hβδ Iν N γ N δ dνdΩ .

• Page 179, footnote 26. Change

“Multifluids of this type as sometimes also referred to as...”
→
“Multifluids of this type are sometimes also referred to as...”
4
Linear and Nonlinear Hydrodynamic
Waves

• Page 200, Eq. (4.54). Change

det(At − λAx ) = 0 ,
→
det(Ax − λAt ) = 0 ,
• Page 213, Fig. 4.5. The tangents to the fluidlines on either side of the rarefaction tail
should be exactly the same and not as shown.
• Page 216, below Eq. (4.114). Change
“ It is also convenient to rewrite the continuity equation (4.112) and the conservation of
energy (4.113)”
→
“ It is also convenient to rewrite the continuity equation (4.112) and the conservation of
momentum (4.113)”
• Page 216. Change Eq. (4.117) as follows
 
2 2 ha hb
(hb Wb vb ) − (ha Wa va ) = − + JpK .
ρa ρb
→  
ha hb
(hb Wb vb )2 − (ha Wa va )2 = + JpK .
ρa ρb
• Page 217, first paragraph. Change
“The classical Hugoniot adiabat is readily obtained from (4.118) after recalling that in
the Newtonian limit hN = 1 +  + p/ρ ≈ 1”
→
“The classical Hugoniot adiabat is readily obtained from (4.118) after recalling that in
the Newtonian limit h = 1 +  + p/ρ ≈ 1”
• Page 221, Eq. (4.138). Change

2 (3ea + eb )(3eb + ea ) 4
Wab = = Wa2 Wb2 ,
16e1 e2 9
→

2 (3ea + eb )(3eb + ea ) 4
Wab = = Wa2 Wb2 ,
16ea eb 9
 Linear and Nonlinear Hydrodynamic Waves xvii

• Page 222, after Eq. (4.141), the expression for the shock velocity should be modified as
follows:

VS± = ρb Wb vb /(ρb Wb ± ρa )
→
VS± = ρb Wb vb /(ρb Wb ∓ ρa )
• Page 225, Fig. 4.11, panel on bottom right. The labels in the spacetime diagram should
be corrected as follows:
R← → R→ and S→ → S← .
• Page 228, last sentence in the first paragraph should be modified as follows:

These values mark the transition from one wave pattern to another one, and that are
directly computed from the initial conditions (Rezzolla and Zanotti, 2001).
→
These values mark the transition from one wave pattern to another one, and are directly
computed from the initial conditions (Rezzolla and Zanotti, 2001).
5
Reaction Fronts: Detonations and
Deflagrations

• Page 284, problem 2. Change

“Derive the inequalities (5.4)–(5.6) across a reaction front [Hint: start from the laws of
conservation of momentum and energy (4.114)–(4.113)].
→
“Derive the inequalities (5.4)–(5.6) across a reaction front [Hint: start from the laws of
conservation of momentum and energy (4.113)–(4.114)].
6
Relativistic Non-Perfect Fluids

• Page 297, Eq. (6.73). Change: χ → χt

• Page 301, Eq. (6.85). Change:

   
1 β1 µ
qν = −κ Dν ln T + aν + β1 q̇ν + T ∇µ u qν ,
2 T
→

   
1 β1 µ
qν = −κT Dν ln T + aν + β1 q̇ν + T ∇µ u qν ,
2 T

• Page 306: invert the inequalities in Eqs. (6.115) and (6.116).

7
Formulations of the Einstein–Euler
Equations

• Page 337, Eq. (7.100). Change

1
Γ̃ijk = Γijk − (δji Γm i m il m
km + δk Γjm − γjk γ Γlm ) = Γijk + 2(δji ∂k ln φ + δki ∂j ln φ − γjk γ il ∂l ln φ) ,
3
→
1
Γ̃ijk = Γijk − (δji Γm i m il m
km + δk Γjm − γjk γ Γlm ) = Γijk + δji ∂k ln φ + δki ∂j ln φ − γjk γ il ∂l ln φ ,
3
• Page 339, Eq. (7.108). Change
(3) 1
R + K 2 = K ij Kij + 4πE = Ãij Ãij + K 2 + 4πE ,
3
→
(3) 1
R + K 2 = K ij Kij + 4πE = Ãij Ãij + K 2 + 16πE ,
3
• Page 339, last two lines:
(note that γ̃ij and Ãij , have only five independent components each since they are trace-
less)
→
(note that γ̃ij and Ãij , have only five independent components each since they have traces
that are equal to three or zero, respectively)
• Page 345, first term in Eq (7.131). Change

Dj Θij = 0 ,

→
Dj Θij = 0 ,
• Page 345, Eq (7.132). Change
1 1
Σij := Θij − γij Θkl Θkl = γ 1/3 Lt γ̄ij ,
3 2
→
1 1
Σij := Θij − γij Θkl γ kl = γ 1/3 Lt γ̃ij ,
3 2
 Formulations of the Einstein–Euler Equations xxi

• Page 345, after Eq (7.132). Change

where γ̄ij := γ −1/3 γij is the conformal metric.
→
where γ̃ij := γ −1/3 γij is the conformal metric.
• Page 345, first term in Eq (7.133). Change
Dj Σij = 0 ,
→
Dj Σij = 0 ,
• Page 354, first line. Change
“constraint decouples from the Hamiltonian constraint and it is possible to solve the latter
to obtain the three vectors V̄ i ”
→
“constraint decouples from the Hamiltonian constraint and it is possible to solve the
former to obtain the three vectors V̄ i ”
• Page 354, fourth line. Change
“The calculation of initial data via the solution of the constrains simplifies considerably
if”
→
“The calculation of initial data via the solution of the constraints simplifies considerably
if”
• Page 356, first sentence before Eq. (7.180). Change
“we further introduce the conformal metric γ̄ [cf., Eq. (7.152)], such that”
→
“we further introduce the conformal metric γ̃ [cf., Eq. (7.152), although we here use a
tilde rather than a bar to be closer to the notation of Bonazzola et al. (2004)], such that”
• Page 356, Eq. (7.180). Change
f := det(fij ) = γ̄ := det(γ̄ij ) ,
→
f := det(fij ) = γ̃ := det(γ̃ij ) ,
• Page 356, Eq. (7.181). Change
ψ = (γ/γ̄)1/12 = (γ/f )1/12 .
→
ψ = (γ/γ̃)1/12 = (γ/f )1/12 .
• Page 384, second equation in Exercise 7. Change
1 1
Di Dj φ = − D̃i D̃j + 2 ∂i φ ∂j φ .
2φ 2φ
→
2 1
Di Dj φ = D̃i D̃j φ + ∂i φ ∂j φ − γij ∂ k φ ∂k φ .
φ φ
8
Numerical Relativistic
Hydrodynamics: Finite-Difference
Methods

• Page 393, Eq. (8.16). Change

(h)
j = C̃hp̃j + O(hp̃j +1 ) ,

→
(h)
j = Chpj + O(hpj +1 ) ,

• Page 393, below Eq. (8.16). Change “with C̃ a constant” to “with C a constant”.
• Page 393, Eq. (8.30). Change
log R(h, k)
p̃ := ,
log(h/k)
→
log |R(h, k)|
p̃ := ,
log(h/k)
• Page 395, the 7th line before Eq. (8.37). Change

“...its application across a time interval ∆t introduces an associate truncation error j (h).”
→
“...its application across a time interval ∆t introduces an associated truncation error
(h).”
• Page 406, Eq. (8.87). Change
2 2
ũ(x, t) = e−εk t eik[x−(v+βk )t] ,

→ 2 2
ũ(x, t) = e−εk t eik[x−(λ+βk )t] ,
10
Numerical Relativistic
Hydrodynamics: High-Order
Methods

• Page 464, Eq. (10.14). Change

Z 1
Aik := Ψk (ξ)dξ , ∀ Ii ∈ Sjl .
0
→
Z
Aik := Ψk (ξ)dξ , ∀ Ii ∈ Sjl .
Ii
11
Relativistic Hydrodynamics of
Non-Selfgravitating Fluids

• Page 518, Eq. (11.84). Change

dW M u
= 2 dr + du .
W r W W
→
dW M u
= 2 2 dr + 2 du .
W r W W
12
Relativistic Hydrodynamics of
Selfgravitating Fluids

• Page 596, 12th line after Eq. (12.13)

“M = 2.01 ± 0.4 M ” → “M = 2.01 ± 0.04 M ”
• Page 597, caption of Fig. 12.1
“M = 2.01 ± 0.4 M ” → “M = 2.01 ± 0.04 M ”
• Page 601, second term in Eq. (12.31). Change

ε(1 + wR )2
 
2 1
H0 = 2 1 − .
Ri wi
→
ε(1 + wi )2
 
2 1
H0 = 2 1 − .
Ri wi
• Page 601, second paragraph, change:
... the energy density ρ(r) and the pressure p(r) ...
→
... the energy density e(r) and the pressure p(r) ...
• Page 606, caption of Fig. 12.4:
“K = 100” → “K = 164”
• Page 606, second but last line:
“K = 100” → “K = 164”