Analytic Transport from Weak to Strong Coupling in the O(N) model

Paul Romatschke1, 2
1
Department of Physics, University of Colorado, Boulder, Colorado 80309, USA
2
Center for Theory of Quantum Matter, University of Colorado, Boulder, Colorado 80309, USA
In this work, a second-order transport coefficient (the curvature-matter coupling κ) is calculated
exactly for the 3+1d O(N) model at large N for any coupling value. Since the theory is “trivial” in
the sense of possessing a Landau pole, the result for κ only is free from cut-off artifacts much below
the Landau pole in the effective field theory sense. Nevertheless, this leaves a large range of coupling
values where this transport coefficient can be determined non-perturbatively and analytically with
little ambiguity. Along with thermodyamic results also calculated in this work, I expect exact large
N results to provide good quantitative predictions for N=1 scalar field theory with φ4 interaction.
arXiv:1905.09290v1 [hep-th] 22 May 2019

I. INTRODUCTION Hooft coupling and larger number of colors [1, 2, 5, 6],

Transport coefficients determine the real-time relax-
ation of a perturbation around a state of equilibrium.
Familiar transport coefficients include conductivities, dif-
fusion coefficients and viscosities. However, these well-
known transport coefficients merely approximate the re-
sponse of a system to a perturbation through a linear
(first order) relationship with the local gradient. In real
systems, there are non-linear corrections (second order,
third order, etc.) which come with their own respec-
tive transport coefficients. For many applications, ig-
noring these higher-order terms constitutes a reasonable
approximation, but for some perturbations, in particular
those where gradients are strong, knowledge of second-
order transport coefficients is important. Also, there are
different types of perturbations (“channels”) which pre-
dominately couple to different combinations of transport
coefficients, for instance the sound channel (longitudi-
nal compression mode coupling to shear and bulk viscos-
ity) and the shear channel (coupling predominantly to
shear viscosity). Sometimes relations of transport coeffi-
cients between different channels exist, such as the well-
known Einstein relation between the diffusion coefficient
and conductivity.
For the purpose of this work, I will consider the some-
what exotic transport coefficient κ, which appears as
second-order correction in the familiar sound and shear
mode channels, and which was introduced in Refs. [1, 2]
in the context of relativistic fluid dynamics. However,
κ enters into the description of relativistic fluid as the
leading order correction when considering the coupling
of matter to perturbations in the curvature of space-
time (e.g. gravitational waves). Because of relations
similar in nature to the Einstein relations for diffusion,
this curvature-matter coupling coefficient κ enters in the
real-time evolution of sound waves in flat space-time
(albeit as a second-order correction). Therefore, even
though κ predominantly governs the interactions between
space-time curvature and matter, this transport coeffi-
cient can be calculated by considering correlation func-
tions in flat space-time (“Kubo formulas”). Results for κ
are currently available for free field theory [3, 4], infinitely
strongly coupled gauge theories in the limit of large ’t
and SU(3) gauge theory from lattice simulations [7].
With the exception of the numerical constraints from
Ref. [7], κ is unknown in any quantum field theory except
near coupling values of λ ' 0, ∞. Given that κ is hardly
of crucial relevance in most transport applications, one
might be tempted to blame this apparent lack of knowl-
edge on an apparent lack of interest.
Unfortunately, the situation is hardly better for other,
more familiar transport coefficients which are of crucial
importance in most transport situations. For instance,
for scalar field theory and QCD the shear viscosity coef-
ficient has been calculated in perturbation theory around
vanishing coupling in Refs. [8–10], and in large N gauge
theories near infinite coupling in Refs. [11–13]. Further-
more, there are constraints from lattice simulations in
SU(3) gauge theory [14, 15], but exact results valid at in-
termediate coupling do not exist for any theory to date.
So why focus on calculating exotic transport coeffi-
cients when there is such need for the shear viscosity?
The answer is that κ is considerably easier to calculate
because it can be extracted from Euclidean (imaginary-
time) rather than retarded (real-time) correlation func-
tions. However, there may be hope to generalize the cal-
culation presented here to other transport coefficients.
In this work, I calculate κ for a particular theory
(the O(N) model with quartic interactions) where such
a transport calculation is feasible. Somewhat unfortu-
nately, in the large N limit the O(N) model in 3+1 di-
mensions possesses a positive β-function for all coupling
values. Integrating the β function, the coupling diverges
at a finite energy scale (aka the Landau pole). The the-
ory is thus UV-incomplete or “trivial”. For energy scales
close to the Landau pole, all possible irrelevant opera-
tors contribute, and hence observables will be sensitive to
the particular discretization (the form of the Lagrangian)
chosen for the theory. However, a (non-perturbative)
renormalization program can be carried through for IR-
safe observables such as the pressure, and UV-incomplete
theories may be interpreted as effective low-energy de-
scriptions. Thus, the O(N) model may be considered
phenomenologically viable at energy scales well below the
Landau pole. In practice, sensitivity to the cutoff scale
can be tested for by varying the renormalization scale
 2
     2
parameter, thus providing a quantitative handle on the 1 ~ · ∂µ φ
∂µ φ ~ + λ φ ~ 2 and the energy-momentum
2 N
breakdown of the theory.
tensor component by T xy = ∂ x φ∂ y φ. The Euclidean cor-
relator in (4) thus becomes
II. THE CALCULATION Z
hT xy T xy iE = 2N
P 2 2
kx ky ∆(ωn , k)∆(ωn , k − p) , (5)
K
Hydrodynamics provides the universal low energy/long
wavelength description of matter. As such, hydrodynam- where ∆(ωn , k) is the full two-point function of the scalar
ics can be set up from a gradient expansion and the sym- field,
metries of the system under consideration, and univer- R 4
sally determines the form of the n-point functions of the Dφe− d xL φi (x)φj (0)
R
energy-momentum tensor T µν , cf. Ref. [16]. Using a con- R R
4
= δij ∆(x) . (6)
Dφe− d xL
struction valid up to (including) second order in gradients
such that T µν = T0µν + T1µν + T2µν , the second-order term ~2
for example includes contributions such as [16] Introducing an auxiliary field σ = φN and Lagrange
multiplier ζ and subsequently integrating out σ, the par-
R 4
T2µν = κR<µν> −2(κ−κ∗ )uλ uρ Rλ<µν>ρ +ξ5 ∆µν R+. . . , tition function Z = Dφe− d xL can be rewritten as
R
(1)
where ∆µν = g µν + uµ uν , gµν is the metric tensor, uµ is
~ N ζ2
Z h i
1~
d4 x
R
the fluid velocity , Rµνλρ , Rµν , R are the Riemann tensor, Z= DφDζe
− 2 φ(−+iζ)φ+ 16λ
. (7)
Ricci tensor, and Ricci scalar, and hi denote symmetric
traceless projection. In (1), κ, κ∗ , ξ5 are curvature-matter
In the large N limit, only the zero mode ζ0 contributes,
coupling transport coefficients. Variation of the full T µν
and as a consequence the partition function can be calcu-
with respect to the metric tensor gives the retarded two-
lated exactly from the location of the saddle at iζ0 = z ∗ ,
point function in Minkowski space-time [16]
√
r  2 
Z ζ0
 p2 κ βV N −βV N 16λ +J ( iζ0 )
 κ Z= dζ0 e ,
Gxy,xy
R (ω, pẑ) = P −iηω+ω 2 ητπ − + κ∗ + +. . . , 16λπ
2 2 h √ ∗2
i
(2) −βV N J ( z ∗ )− z16λ
where P is the pressure, η is the shear viscosity coef- =e , (8)
ficient, and τπ , κ, κ∗ are second-order transport coeffi- where V is the volume of R3 , J(m) ≡
cients. Note the dual role of κ, κ∗ in curved space-time R d3−2

T ωn µ2 (2π) 2 2
P
3−2 ln ωn + m in dimensional
Eqns. (1) and flat space-time (2) is similar to the Einstein
relations for diffusion and conductivity. Knowledge of GR regularization [19] and ωn = 2πnT are the bosonic
at vanishing external frequency ω, but finite wavenumber Matsubara frequencies. (Note that this is completely
p is sufficient to determine κ [2, 4, 17]. analogous to the case of 2d and 3d discussed in
I choose to calculate this correlator for the massless Refs. [20–22].) The two-point function thus becomes
O(N) model in 3+1d. In curved space-time, the action 1
for this theory is given by [18] ∆(ωn , k) = , (9)
ωn2 + k2 + z ∗
4 √
Z  
1 µν ~ ~ ~ 2 λ  ~ 2 2
d x −g g ∂µ φ∂ν φ + ξRφ − φ , (3) where the location of the saddle z ∗ is given as the solution
2 N of the non-perturbative “gap-equation”
~ = (φ1 , φ2 , . . . φN ) is an N-component scalar field. √
where φ z ∗ = 4λI( z ∗ ) . (10)
Here ξ is a parameter which takes the value ξ = 61 for a
conformally coupled scalar. Calculating Gxy,xy by vary- Here I(m) = 2 dJ(m)
PR 2 2 2 −1
dm2 = k [ωn +k +m ] is a standard
R
ing the energy-momentum tensor for (3) with respect to thermal integral found in textbooks such as Ref. [19]
the metric, the coefficient proportional to p2 in (2) re-
ceives two contributions that can be expressed in terms m2 m2 µ̄2 e1
∞
mT X K1 nm


T
of Euclidean two-point correlation functions [17], I(m) = − − ln 2 + 2 ,
16π 2  16π 2 m 2π n=1 n
ξp2 ~ 2
 
∂ xy xy (11)
κ = lim 2 hT T iE (p) + hφ iE (0) . (4) where µ̄2 = 4πµ2 e−γE is the renormalization scale pa-
p→0 ∂p 2
rameter in the MS scheme. Inspecting (10), one can
Here h·iE denotes Euclidean correlation functions, e.g. non-perturbatively renormalize the theory by introduc-
those calculated in a spacetime S 1 × R3 where one direc- ing a renormalized coupling constant λR as
tion has been compactified on a circle of radius β = T −1 ,
as in standard thermal quantum field theory [19]. The 1 1 1
= + 2 . (12)
corresponding Euclidean Lagrangian is given by L = λR λ 4π 
 3

Curvature-Matter transport coefficient for O(N) model in 3+1d

0

-0.0001

-0.0002

-0.0003 Cut-off scale sensitivity

χ=1
κ/(N T2)

-0.0004

-0.0005

-0.0006 λR=∞ λR=0

-0.0007

-0.0008

-0.0009
0 0.2 0.4 0.6 0.8 1
1/(1+λR)

FIG. 1. Transport coefficient κ for the O(N) model in 3+1 dimensions in the large N limit as a function of coupling. Results
1
are shown using a compactified interval 1+λ R
∈ [0, 1] in order to show all coupling values. Arrows indicate free theory and
strong coupling limits, respectively. The band was generated by varying the scale through choosing χ ∈ [ 12 , 2] in (15), in turn
quantifying the sensitivity of κ to the cut-off scale. See text for details.

This renormalization condition implies a positive β- with fixed λR and varying χ ∈ [ 12 , 2].
function for all couplings. Integrating up the renormal- In practice, the renormalized gap equation (14) pos-
ization group equation gives sesses two solutions for mB . Only the smaller one of these
1 1 Λ2 corresponds to a local minimum of the exponent, thus the
= 2
ln LP , (13) larger one will be discarded in the following. The solu-
λR (µ̄) 4π µ̄2
tion mB then fixes the form of the two-point function (9)
where ΛLP is the Landau pole of the theory (defined as non-perturbatively, and in turn allows calculation of the
the scale where λR (ΛLP ) = ∞). transport coefficient κ from (4). Specifically, performing
Expressing the thermal mass in (9) as z ∗ = m2B T 2 , the the angular averages in (5) leads to
dimensionless parameter mB is determined from (10) as
8 n K1 (nm B)
8 n K1 (nm B)
Z
κ 2 P
ξ
P P
mB = n
= n
, (14) ≡ 4k 6 ∆4 (ωn , k) − 7k 4 ∆3 (ωn , k) + I(mB T ) .
Λ 2 e1 4π 2 µ̄ e1
2 2N 105 K 2
ln LP2
mB T2 λR + ln m2 T 2 B (16)
either in terms of the ratio ΛLP /T or in terms of the Inspecting this equation, one notices that the last term
renormalized running coupling. (Note that mB is in- is divergent for  → 0, cf. (11). Therefore, unless this
dependent from the choice of the renormalization scale divergence is exactly canceled by the other contributions,
∂ 2
the result for κ is meaningless. Using ∂m 2 ∆ = −∆
parameter µ̄ as it should be for a physical observable.)
Note that while the gap equation (14) formally is well- repeatedly and performing a standard thermal sum, one
defined for all temperature scales T ∈ [0, ΛLP finds
2 ], close to
the Landau pole there will be modifications arising from
radiative corrections to the effective theory Lagrangian. ∂3 k6
Z Z
P 6 4
(This may be verified explicitly by adding a term such A= k ∆ (ωn , k) = − (1 + 2nB (Ek )) ,
 3 K (∂z ∗ )3 k 12Ek
as λ22 φ ~ 2 to the Lagrangian (3), which is allowed for Z
P 4 3 ∂2
Z
k4
ΛLP
B= k ∆ (ωn , k) = (1 + 2nB (Ek )) ,(17)
an UV-incomplete theory.) It is possible to test for the K (∂z ∗ )2 k 4Ek
sensitivity to the cut-off scale by e.g. choosing units as
2 √
ΛLP e
− 2π
λ R where Ek = k 2 + z ∗ and nB (x) = ex/T1 −1 . Expanding
2πT = , (15) P∞
nB (x) = n=1 e−nβx both of the above integrals can be
χ
 4

Speed of Sound for O(N) model in 3+1d The entropy density s ≡ ∂P

∂T may be obtained most easily
0.014 from (8) and (10), so that contributions proportional to
∂z ∗
0.012 ∂T cancel, leading to the result

∞
0.01
Cut-off scale sensitivity
N T 3 m3B X K3 (nmB )
s= . (20)
0.008
χ=1 2π 2 n=1 n
cs2-1/3

0.006 3 2
For weak coupling where mB → 0, s → sfree = 4N 90 T π
,
0.004 the well-known Stefan-Boltzmann result for a free theory.
0.002
From the thermodynamic relation  + 3P = sT and
this result, the trace anomaly can be evaluated to be
0
0 0.2 0.4
1/(1+λR)
0.6 0.8 1
N T 4 m4B
 − 3P = − . (21)
32π 2
FIG. 2. Speed of sound squared for the O(N) model in Note that the result is negative and that most contri-
3+1 dimensions in the large N limit as a function of coupling. butions have canceled because of the gap equation (14).
1
Results are shown using a compactified interval 1+λ R
∈ [0, 1] Finally, the speed of sound squared can be calculated as
in order to show all coupling values. The band was generated c2s = s/T
∂s , and evaluated numerically, see Fig. 2. Note
by varying the scale through choosing χ ∈ [ 21 , 2] in (15), in ∂T
that the speed of sound is very close to (and above)
turn quantifying the sensitivity of c2s to the cut-off scale.
the conformal result c2s = 13 , which indicates that the
O(N) model, though not a conformal theory (CFT), is
evaluated analytically, finding numerically very close to a CFT for most coupling val-
ues. Indeed, it has not escaped my attention that the
∞ √ !
35z ∗ 1 37 z∗ X K1 (nβ z ∗ ) ratio s/sfree calculated from (20) seems to go to a con-
4A = − + + ln 2 − 8 , stant value of approximately 85 percent for λR → ∞ and
64π 2  105 µ̄ n=1
nβ
χ = 1, very much in line with the universal strong-weak
∞ √ !
105z ∗ 1 1 z∗ X K1 (nβ z ∗ ) thermodynamic behavior found in 2+1d CFTs [22, 24].
7B = − + + ln 2 − 8 .
128π 2  15 µ̄ n=1
nβ

Inserting these results into (16), I find that the  → 0 III. DISCUSSION AND CONCLUSIONS
divergence as well as the sums over Bessel functions both
cancel for ξ = 16 , giving rise to the finite and simple result The transport coefficient κ given in (18) may be eval-
uated for any value of the renormalized coupling λR by
solving (14) numerically. The sensitivity to cut-off scale
13N T 2 m2B effects may be tested by the choice (15) through varying
κ=− , (18)
2520π 2 χ. Results for κ for all couplings are shown in Fig. 1.
From this figure, it can be seen that cut-off scale sensi-
with mB given by the solution of (14). This is the main
tivity is minor (smaller than 10 percent) for λR < ∼ 2.45
result of this work. A quick cross-check reveals that in the
and less than a factor of two even for λR → ∞. This
free theory limit λR → 0 Eq. (14) gives mB → 0, so that
compares favorable with the situation found for the QCD
limλR →0 κ = 0 (matching the result found in Ref. [17] for
shear viscosity calculated to NLO in perturbation theory
a conformally coupled scalar).
[10]. The weak sensitivity to cut-off scale effects suggests
Of course, also thermodynamic properties of the
that the result (18) constitutes an example of a transport
O(N) model in 3+1 dimensions may be evaluated non-
coefficient that is known non-perturbatively for a large
perturbatively along the same lines. For instance, the
range of coupling values. As such, this example may be
pressure (minus the free energy) is found from (8) as
useful for instance for testing approximation techniques
P = ln Z
βV . It is worth pointing out that – using the
h 2 3/2
i (either at weak or at strong coupling), or conceivably in
m4
explicit result [19] for J(m) = − 64π2  + ln µ̄ me 2
1
− early-time cosmology where curvature matter couplings
m T2 2 P K 2 (nβm) will play an important role in the dynamics.
2π 2 n n2 – the non-perturbative coupling renor-
The results found in this work are exact only in
malization (12) is sufficient to remove all divergences in
the strict large N limit. However, based on the non-
the pressure (cf. Ref. [23]), so that no counterterm for
perturbative results for scalar theory in 1+1d in Ref. [21],
the cosmological constant is required. This leads to
I conjecture that the large N results in this work consti-
" ∞
# tute quantitatively good approximations for finite N at
NT4 Λ2
e 3/2 X K 2 (nm B ) arbitrary coupling, including N=1 scalar φ4 theory.
P = m4B ln LP2 2 + 32m2B .
64π 2 mB T n=1
n2 Is it possible to non-perturbatively evaluate other
(19) transport coefficients in a similar manner? The answer

to this question likely is affirmative since other channels
of the energy-momentum tensor two-point function cou-
ple to transport coefficients such as κ∗ , ξ5 , ξ6 in a similar
manner, cf. Refs. [4, 16, 17] for details. Interestingly,
given the relation between κ, κ∗ and the shear viscosity
coefficient discussed in Ref. [25], or the Haack-Yarom re-
lation [26], this could open a possible “backdoor” avenue
towards an exact non-perturbative determination of the
shear viscosity coefficient in the 3+1d O(N) model.
It is my hope that this work could instill interest and
[1] Sayantani Bhattacharyya, Veronika E Hubeny, Shi-
raz Minwalla, and Mukund Rangamani, “Nonlinear
Fluid Dynamics from Gravity,” JHEP 02, 045 (2008),
arXiv:0712.2456 [hep-th].
[2] Rudolf Baier, Paul Romatschke, Dam Thanh Son, An-
drei O. Starinets, and Mikhail A. Stephanov, “Rel-
ativistic viscous hydrodynamics, conformal invariance,
and holography,” JHEP 04, 100 (2008), arXiv:0712.2451
[hep-th].
[3] Paul Romatschke and Dam Thanh Son, “Spectral sum
rules for the quark-gluon plasma,” Phys. Rev. D80,
065021 (2009), arXiv:0903.3946 [hep-ph].
[4] Guy D. Moore and Kiyoumars A. Sohrabi, “Thermody-
namical second-order hydrodynamic coefficients,” JHEP
11, 148 (2012), arXiv:1210.3340 [hep-ph].
[5] Stefano I. Finazzo, Romulo Rougemont, Hugo Marro-
chio, and Jorge Noronha, “Hydrodynamic transport
coefficients for the non-conformal quark-gluon plasma
from holography,” JHEP 02, 051 (2015), arXiv:1412.2968
[hep-ph].
[6] Sašo Grozdanov and Andrei O. Starinets, “Second-order
transport, quasinormal modes and zero-viscosity limit
in the Gauss-Bonnet holographic fluid,” JHEP 03, 166
(2017), arXiv:1611.07053 [hep-th].
[7] Owe Philipsen and Christian Schäfer, “The second
order hydrodynamic transport coefficient κ for the
gluon plasma from the lattice,” JHEP 02, 003 (2014),
arXiv:1311.6618 [hep-lat].
[8] Sangyong Jeon, “Hydrodynamic transport coefficients in
relativistic scalar field theory,” Phys. Rev. D52, 3591–
3642 (1995), arXiv:hep-ph/9409250 [hep-ph].
[9] Peter Brockway Arnold, Guy D Moore, and Laurence G.
Yaffe, “Transport coefficients in high temperature gauge
theories. 2. Beyond leading log,” JHEP 05, 051 (2003),
arXiv:hep-ph/0302165 [hep-ph].
[10] Jacopo Ghiglieri, Guy D. Moore, and Derek Teaney,
“QCD Shear Viscosity at (almost) NLO,” JHEP 03, 179
(2018), arXiv:1802.09535 [hep-ph].
[11] G. Policastro, Dan T. Son, and Andrei O. Starinets,
“The Shear viscosity of strongly coupled N=4 supersym-
metric Yang-Mills plasma,” Phys. Rev. Lett. 87, 081601
(2001), arXiv:hep-th/0104066 [hep-th].
[12] Yevgeny Kats and Pavel Petrov, “Effect of curvature
squared corrections in AdS on the viscosity of the dual
gauge theory,” JHEP 01, 044 (2009), arXiv:0712.0743
[hep-th].
[13] Alex Buchel, “Resolving disagreement for eta/s in a CFT
plasma at finite coupling,” Nucl. Phys. B803, 166–170
(2008), arXiv:0805.2683 [hep-th].
5
further progress in the field of non-perturbative transport
calculations.
IV. ACKNOWLEDGMENTS
This work was supported by the Department of Energy,
DOE award No de-sc0017905. I would like to thank T.
DeGrand and S.P. de Alwis for helpful discussions.
[14] Harvey B. Meyer, “A Calculation of the shear viscosity
in SU(3) gluodynamics,” Phys. Rev. D76, 101701 (2007),
arXiv:0704.1801 [hep-lat].
[15] Sz. Borsányi, Zoltan Fodor, Sandor D. Giordano, Mat-
teo andKatz, Attila Pasztor, Claudia Ratti, Andreas
Schäfer, Kalman K. Szabo, and Balint C. Toth, “High
statistics lattice study of stress tensor correlators in pure
SU (3) gauge theory,” Phys. Rev. D98, 014512 (2018),
arXiv:1802.07718 [hep-lat].
[16] Paul Romatschke and Ulrike Romatschke, “Relativistic
Fluid Dynamics In and Out of Equilibrium – Ten Years of
Progress in Theory and Numerical Simulations of Nuclear
Collisions,” (2017), arXiv:1712.05815 [nucl-th].
[17] Pavel Kovtun and Ashish Shukla, “Kubo formulas for
thermodynamic transport coefficients,” JHEP 10, 007
(2018), arXiv:1806.05774 [hep-th].
[18] Leonard E. Parker and D. Toms, Quantum Field Theory
in Curved Spacetime, Cambridge Monographs on Math-
ematical Physics (Cambridge University Press, 2009).
[19] Mikko Laine and Aleksi Vuorinen, “Basics of Thermal
Field Theory,” Lect. Notes Phys. 925, pp.1–281 (2016),
arXiv:1701.01554.
[20] Paul Romatschke, “Simple non-perturbative resumma-
tion schemes beyond mean-field: case study for scalar
φ4 theory in 1+1 dimensions,” JHEP 03, 149 (2019),
arXiv:1901.05483 [hep-th].
[21] Paul Romatschke, “Simple non-perturbative resumma-
tion schemes beyond mean-field II: thermodynamics of
scalar φ4 theory in 1+1 dimensions at arbitrary cou-
pling,” (2019), arXiv:1903.09661 [hep-th].
[22] Paul Romatschke, “Finite temperature CFT results for
all couplings: O(N) model in 2+1 dimensions,” (2019),
arXiv:1904.09995 [hep-th].
[23] J. P. Blaizot, Edmond Iancu, and A. Rebhan, “Ap-
proximately selfconsistent resummations for the thermo-
dynamics of the quark gluon plasma. 1. Entropy and
density,” Phys. Rev. D63, 065003 (2001), arXiv:hep-
ph/0005003 [hep-ph].
[24] Oliver DeWolfe and Paul Romatschke, “Strong Coupling
Universality at Large N for Pure CFT Thermodynamics
in 2+1 dimensions,” (2019), arXiv:1905.06355 [hep-th].
[25] Philipp Kleinert and Jonas Probst, “Second-Order Hy-
drodynamics and Universality in Non-Conformal Holo-
graphic Fluids,” JHEP 12, 091 (2016), arXiv:1610.01081
[hep-th].
[26] Michael Haack and Amos Yarom, “Universality of second
order transport coefficients from the gauge-string dual-
ity,” Nucl. Phys. B813, 140–155 (2009), arXiv:0811.1794
[hep-th].