Applied Thermal Engineering 37 (2012) 280e292

Contents lists available at SciVerse ScienceDirect

Applied Thermal Engineering

journal homepage: www.elsevier.com/locate/apthermeng

A numerical study of nanofluid forced convection in ribbed channels

Oronzio Manca*, Sergio Nardini, Daniele Ricci
Dipartimento di Ingegneria Aerospaziale e Meccanica, Seconda Università degli Studi di Napoli, Real Casa dell’Annunziata, via Roma, 29, 81031 Aversa (CE), Italy

a r t i c l e i n f o a b s t r a c t

Article history: In this paper a numerical investigation on forced convection with nanofluids, composed by water and
Received 4 July 2011 Al2O3 nanoparticles, in a two-dimensional channel is carried out. A uniform heat flux is applied on the
Accepted 14 November 2011 external walls. A single-phase approach is employed to model nanofluids and the fluid properties are
Available online 22 November 2011
considered constant with temperature. The particle size is set equal to 38 nm and nanoparticle volume
fractions from 0% to 4% are considered. The flow regime is turbulent and Reynolds numbers are in the
Keywords:
range 20,000e60,000. Furthermore, square and rectangular shapes and different arrangements of ribs
Heat transfer enhancement
are analyzed in terms of different dimensionless heights and pitches of elements. The investigation is
Forced convection
Nanofluids
accomplished by means of Fluent code and its aim consists into find arrangements of ribs such to give
Turbulent regime high heat transfer coefficients and low pressure drops in presence of watereAl2O3 nanofluids. Results are
Ribs presented in terms of temperature and velocity fields, and profiles of average Nusselt number, average
Numerical investigation heat transfer coefficients and required pumping power. Heat transfer enhancement increases with the
particle volume concentration but it is accompanied by increasing required pumping power. The heat
transfer improves, as Reynolds number raises, but also an increase of pumping power is observed.
Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction non-circular channels, longitudinal flows in rod bundles, annuli

Heat transfer enhancement represents an active and important
field of engineering research since it contributes to the effective-
ness increase of heat exchangers. The improvement of the heat
transfer rates in the thermal devices by adopting suitable tech-
niques can result in significant technical advantages and cost
savings. These targets could be achieved by the employment of
different available techniques, divided by Bergles et al. [1] in passive
and active ones. Passive methods use special surface geometries,
like rough surfaces, or fluid additives to achieve the aim of heat
transfer enhancement whereas the active techniques require
external power sources, such as mechanical aids, electric or
acoustic fields and surface vibration [2,3]. Among the available
passive methods, the employment of ribbed surfaces already covers
several commercial applications but other techniques has recently
burst on the scene of engineering research, like the use of nano-
particles as additives in the working fluids in order to improve their
thermal conductivity and convective heat transfer coefficients,
consequently [4].
Considerable data exist for single-phase forced convection flow
over rough surface for geometries like flat plates, circular tubes,
* Corresponding author. Tel.: þ39 081 5010217; fax: þ39 081 5010204.
E-mail addresses: oronzio.manca@unina2.it (O. Manca), sergio.nardini@unina2.it
(S. Nardini), daniele.ricci@unina2.it (D. Ricci).
and flow orthogonal to circular tubes. Artificial roughness is used
in various application like evaporators, condensers, steam con-
densers, gas turbine blade cooling channels, nuclear reactors heat
exchangers, solar air heaters, etc. [5e9]. Several experimental
studies confirmed that this technique consisted of an efficient
passive method to enhance the forced convection heat transfer by
investigating on conventional configurations, like flat plates,
circular tubes, non-circular channels, longitudinal flows in rod
bundles, annuli [2,10e15]. When a fluid flows in a ribbed channel or
duct, ribs tend to break the laminar sub-layer and create local
turbulence, due to flow separation and reattachment between
consecutive ribs. As a result the thermal resistance reduces and
heat transfer mechanisms improve. This effect overcomes the fin-
effect, due to the increased heat transfer area, provided by the
ribs, mounted on the walls [12]. However, the introduction of
artificial roughness leads to increasing friction losses thus it results
to be convenient to enhance that the turbulence only in a region
close to the heat transferring surface, such as in the viscous sub-
layer. This can be done by keeping the height of roughness
elements to be small in comparison with the duct dimensions.
Wang and Sunden [13] underlined that the local heat transfer was
strongly dependent on the rib shape, for example regarding the
region just downstream the rib. Promvonge and Thianpong [14]
suggested the use of triangular-shaped ribs by carrying out exper-
iments in rectangular channels with ribs, arranged in in-line and

1359-4311/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.applthermaleng.2011.11.030
 O. Manca et al. / Applied Thermal Engineering 37 (2012) 280e292
Nomenclature
a thermal diffusivity, m2/s
cp specific heat, J/kg K
d diameter, m
e rib height, m
f friction coefficient Eq. (32)
H channel height, m
h heat transfer coefficient, W/m2 K
L length, m
Nu Nusselt number Eq. (32)
p rib pitch, m
P pressure, Pa
PEC performance evaluation criteria index Eq. (34)
PP pumping power, W
Pr¼n/a Prandtl number
q heat flux, W/m2
Re Reynolds number Eq. (31)
s channel thickness, m
T temperature, K
u velocity component, m/s
W channel width, m
w rib width, m
x, y spatial coordinates, m
staggered arrays. Saha [15] pointed out that Nusselt number and
friction factors increased with increasing values of turbulators
height and Reynolds number in circular, square and rectangular
ducts, characterized by ribs mounted on two opposite surfaces.
The number of paper on the computational side has grown, due
to the development of calculator computational power and
numerical methods and turbulence modeling accuracy. Several
numerical studies have been carried out to simulate and examine
the effect of ribs in channels and ducts [16e19]. Liou et al. [20]
performed both the numerical analysis and experimental ones to
investigate the heat transfer and the fluid flow behaviors in rect-
angular channels, provided with stream wise periodic ribs on one of
the principal walls. They concluded that the flow acceleration and
turbulence intensity are the most significant factors which influ-
ence the heat transfer coefficient. The highest average heat transfer
coefficient was found in correspondence with a pitch-to-rib height
ratio equal to 10. Rau et al. [21] substantially confirmed these
results and they experimentally found that the optimum pitch-to-
rib height ratio value was equal to 9. Hence, the investigations
revealed that not only the rib geometry arrangement but the tur-
bulator shapes as well play a vital role into enhancing the heat
transfer rate. Ryu et al. [22,23] studied both the thermal and fluid-
dynamic behaviors of turbulent flows with two-dimensional ribs
and three-dimensional blocks in the context of surface roughness
effects. They solved the Reynolds-averaged NaviereStokes equa-
tions, coupled with the keu turbulence model with near-wall
treatment, by a finite-volume method. They investigated several
shapes of ribs with a height equal to 0.1, finding that the maximum
resistance occurred at a pitch-to-height ratio below 10. Chaube
et al. [24] confirmed the good matching between the numerical
predictions by means of SST keu turbulent model and experi-
mental results. Manca et al. [25] performed a numerical analysis on
air forced convection in channels with square, rectangular, trian-
gular and trapezoidal, mounted on two opposite heated walls; they
underlined the significant enhancement of Nusselt number as well
as the increase of friction also in the case of turbulent regime. Kim
et al. [26] conducted optimization using the second-order response
surface method to determine the transverse rib geometry required
281
Greek symbols
d Kronecher delta function
l thermal conductivity, W/m K
m dynamic viscosity, Pa s
n kinematic viscosity, m2/s
r density, kg/m3
s turbulent Prandtl number
s wall shear stress, kg/m2
4 nanoparticle volume fraction
u rate of dissipated turbulent kinetic energy
Subscripts
a ambient
av average
bf base fluid
f fluid
in inlet section
m mass
nf nanofluid
out outlet section
p particle
s smooth
t turbulent
w wall
to achieve the highest cooling performance in a circular channel.
They underlined that the rib height was the dominant factor in heat
transfer and friction loss since as the rib height increased, the
Nusselt number ratios and friction losses increased significantly
because of the flow acceleration and the large flow recirculation.
Moreover, as the rib width increased, the Nusselt number ratios and
friction losses decreased because the flow reattachment location
moved closer to the rib turbulators. Liu and Wang [27] carried out
numerical studies by means of a commercial CFD code on semi-
attached internal ribs in order to reduce the large friction factors
and the extension of regions behind the ribs attachment corners,
where low heat transfer rates were evaluated.
Beside the employment of rough surfaces, an innovative way to
enhance the heat transfer rate in thermal equipments is repre-
sented by the introduction of solid nanoparticles in order to
improve the fluid thermal conductivity [28e30]. In fact, several
investigations revealed that nanofluid heat transfer coefficients
could be increased by more than 20% also in the case of low
nanoparticle concentrations [31e33]. Nowadays, a fast growth of
research activity in this heat transfer area has arisen and one can
remark the exponential increase in the number of research articles.
Moreover, the potential market for nanofluids heat transfer appli-
cations was estimated by the CEA in 2007 to be over 2 billion
dollars per year worldwide, with prospects of further growth in the
next 5e10 years, as underlined in [34]. Thus, the impact of this
technology is expected to be relevant in many fields like the
applications in automotive, aerospace, electronic cooling, refriger-
ation, material processing, fuel cells and nuclear reactors [35e37].
Many studies have been carried out but there is a lack of agreement
among results from different research groups [38,39]. In fact, some
issues concerning thermal conductivity evaluation, Brownian
motion of particles, particle migration and variable property
changes with temperature must be carefully considered while
convection is modeled. In particular, the definition of nanofluid
effective thermo-physical properties is very difficult [40e47].
Therefore, several experimental investigations were performed
[40e43] and many attempts have been made to formulate efficient
theoretical models for the predictions but this topic is still
282 O. Manca et al. / Applied Thermal Engineering 37 (2012) 280e292
incomplete [44e47]. However, several configurations have been
analyzed like tubes and channels with different shapes, micro-
channels, impinging jets, micro heat sinks, etc. [30,48e51].
Relatively, few theoretical and experimental investigations have
been reported on convective heat transfer in confined flows
[52e63], especially in the turbulent regime. Experimental results
were obtained on convective heat transfer for laminar [52e55] and
turbulent [56e58] flow of a nanofluid inside a tube. A significant
experimental investigation was accomplish by Pak and Cho [57],
who performed tests in order to explain the turbulent friction and
heat transfer behaviors of nanofluids, composed by Al2O3 and TiO2
particles suspended in water, in a circular pipe. They underlined the
increase of Nusselt number with increasing volume concentration
as well as the Reynolds number and provided a useful correlation
for Nusselt number for fully developed turbulent flows. Experi-
mental results for the convective heat transfer of Al2O3
(27e56 nm)/water based nanofluids flowing through a copper tube
in laminar regime was reported in Ref. [53]. Numerical investiga-
tions on nanofluids are carried out using two approaches
[28,30,48e50,58e63]. The first approach assumes that the
continuum assumption is still valid for fluids with suspended
nanosize particles at low volume fractions, while the other
approach considers a two-phase model for a further description of
both the fluid and the solid phases [59,60]. The single-phase model
with physical and thermal properties, all assumed to be constant
with temperature, was employed in Refs. [28,48e51,58,61e63]. The
hydrodynamic and thermal characteristics of nanofluids, flowing
through a uniformly heated tube, in both laminar and turbulent
regimes with adjusted properties, were investigated in Ref. [28].
The advantages of nanofluids with respect to heat transfer were
discussed in Ref. [58], but it was also found that the inclusion of
nanoparticles introduced drastic effects on the wall shear stress. A
new correlation was proposed in Ref. [61] to describe the thermal
performance of Al2O3ewater nanofluids under turbulent regime
underlining that the addition of nanoparticles in the base fluid
Fig. 1. Sketch of the model: a) geometrical model; b) characteristic rib parameters; c) shapes of the considered ribs e 1. square rib, 2. rectangular rib with w/e ¼ 0.5, 3. rectangular
rib with w/e ¼ 2.0.
increases the heat transfer rates considerably. Izadi et al. [62]
analyzed the developing laminar forced convection in annulus,
underlining that the nanoparticle volume fraction affected signifi-
cantly the temperature profiles but not the dimensionless axial
velocity ones. Rostamani et al. [63] adopted a single-phase model
with temperature-dependent properties of the working fluid, ob-
tained by Buongiorno [45] and Hamilton and Crosser [64], in order
to study numerically the turbulent forced convection in a heated
horizontal tube.
In this paper, a first attempt to employ forced convection of
nanofluids (water/Al2O3) in a ribbed channel is carried out. The
main aim is to investigate the turbulent flow and the heat transfer
characteristics of a two-dimensional ribbed channel with the two
principal walls subjected to uniform heat flux by means of Fluent
code [65], employing the SST keu model. The single-phase model is
considered. Different values of p/e with square and rectangular ribs,
Reynolds numbers in the range between 20,000 and 60,000,
heights and particle volume concentrations were considered. This
work extends to the nanofluids the study presented in Ref. [66],
where only pure water as working fluid was considered.
2. Mathematical description and governing equations
A computational fluid-dynamic analysis of a two-dimensional
channel model, provided with differently shaped ribs, as reported
in Fig. 1, is considered in order to evaluate its thermal and fluid-
dynamic behaviors and study the temperature and velocity fields.
A constant uniform heat flux is applied on the external channel
walls. Different inlet velocities are considered in the ranges of
turbulent regime and the working fluid is water or Al2O3/water
based nanofluid with constant properties.
Governing equations of continuity, momentum and energy are
solved for a steady-state turbulent flow in rectangular coordinates,
under the hypotheses of steady-state, two-dimensional, incom-
pressible, turbulent, constant properties flow conditions [65]:
 O. Manca et al. / Applied Thermal Engineering 37 (2012) 280e292 283

v where sk and su are the turbulent Prandtl numbers for k and u,

Continuity : ðrui Þ ¼ 0 (1)
vxi respectively:

" !# 1
v
 vP v vui vuj 2 vui sk ¼ (15)
Momentum : rui uj ¼  þ m þ  dij F1 =sk;1 þ ð1  F1 Þ=sk;2
vxj vxi vxj vxj vxi 3 vxj
v  
1
þ  ru0i u0j ð2Þ su ¼ (16)
vxj F1 =su;1 þ ð1  F1 Þ=su;2
"  # and mt is the turbulent viscosity, computed from:
v v cpmt vT

Energy : ½u ðrE þ PÞ ¼ lþ þ ui sij eff (3)
vxi i vxj Prt vxj rk
mt ¼ a* (17)
u
where E is the total energy, E ¼ cpT  ðP=rÞ þ ðu2 =2Þ and (sij)eff is
The coefficient a* dampens the turbulent viscosity. It is given by
the deviatoric stress tensor, defined as
!



a*0 þ Ret =Rk
vuj vui 2 vu a* ¼ a*N (18)
sij ¼ meff þ  meff i dij (4) ð1 þ Ret =Rk Þ
eff vxi vxj 3 vxj
where Ret ¼ rk/mu, Rk ¼ 6, a*0 ¼ bi/3 and bi ¼ 0.072. In the high-
The transport equations for the SheareStress Transport keu Reynolds number form a* ¼ a*N ¼ 1.0. The blending equation F1
model are reported below in the form, developed by Menter [67]: is given by
!  
v v vk ~ Y þS F1 ¼ tan F41 (19)
ðrkui Þ ¼ Gk þG k k k (5)
vxi vxj vxj
" pffiffiffi ! #
! k 500m 4rk
F1 ¼ min max ; 2 ; (20)
v v vu 0:09uy ry u su;2 Dþ
uy
2
ðrukui Þ ¼ Gu þ Gu  Yu þ Du þ Su (6)
vxi vxj vxj
" #
where Gk is the production of turbulence kinetic energy due to 1 1 vk vu
Dþ
u ¼ max 2r ; 1010 (21)
mean velocity gradients, Gu represents the generation of u: su;2 u vxj vxj
 
~ ¼ min G ; 10b* ku where y is the distance to the next surface and Dþ u is the positive
G k k (7)
portion of the cross-diffusion term.
with Gk ¼ ru0i u0j ðvuj =vxi Þ Yk and Yu represent the dissipation of k and u due to turbulence:

a *
Gu ¼ Gk (8) Yk ¼ rb ku (22)
vt
where nt is the kinematic turbulent viscosity and b* a model Yu ¼ rbu2 (23)
constant; a is given by with b and b* are model constants; moreover:


a*0
þ Ret =Ru bi ¼ F1 bi;1 þ ð1  F1 Þbi;2 (24)
a ¼ aN (9)
ð1 þ Ret =Ru Þ
Du represents the cross-diffusion term while Sk and Su are
where Ru ¼ 2.95. The term a is represented by possible source terms. In particular, Du is given by

aN ¼ F1 aN;1 þ ð1  F1 ÞaN;2 (10) 1 vk vu

Du ¼ 2ð1  F1 Þrsu;2 (25)
u vxj vxj
bi;1 k2
aN;1 ¼ *
 qffiffiffiffiffiffiffi (11) The model constants are summarized in Table 1. A two-
bN su;1 b* dimensional conduction model in heated walls is employed and
N
the equation at the steady-state regime is:
bi;2 k2  
aN;1 ¼  qffiffiffiffiffiffiffi (12) v vT
* l ¼ 0 (26)
bN su;2 b* vxi vxi
N

where k ¼ 0.41 and bi ¼ 0.072. In the high-Reynolds number form The assigned boundary conditions are the following:
a ¼ aN ¼ 1.0.
In Eqs. (5) and (6) the terms Gk and Gu represent the effective - inlet section: uniform velocity and temperature profile;
diffusivity of k and u. They are given by: - outlet section: outflow condition with velocity components
and temperature derivatives equal to zero;
mt
Gk ¼ m þ (13) Table 1
sk Shear stress transport keu model constants.

sk,1 ¼ 1.176 sk,2 ¼ 1.0 su,1 ¼ 2.0 su,2 ¼ 1.168 a1 ¼ 0.31 bi,1 ¼ 0.075
mt bi,2 ¼ 0.0828 a*N ¼ 1.0 aN ¼ 0.52 b*N ¼ 0.09 bi ¼ 0.072 a0 ¼ 1/9
Gu ¼ m þ (14)
Rb ¼ 8.0 Rk ¼ 6.0 Ru ¼ 2.95 z* ¼ 1.5 sk ¼ 2.0 su ¼ 2.0
su
284 O. Manca et al. / Applied Thermal Engineering 37 (2012) 280e292

Table 2 _
qd
Material properties at a temperature of 293 K. Nuav ¼ (32)
ðTw  Tm Þlf
3
Material r [kg/m ] cp [J/kg K] m [Pa s] l [W/m K]
where Tw and Tm represent the average wall temperature and the
Alumina (Al2O3) 3880 773 // 36
Water 998 4182 998  106 0.597 average bulk temperature, respectively.

d 1
f ¼ 2DP (33)
- duct walls: velocity components equal to zero; L ru2av
- external duct walls: uniform heat flux.
The PEC index was adopted in order to compare the thermal and
fluid-dynamic performances of channels with differently shaped
3. Physical properties of nanofluids
ribs [2]:

The working fluid is water or a mixture of water and Al2O3

nanoparticles with a diameter of 38 nm, at different volume frac-
tions equal to 1% and 4%. In Table 2 the values of density, specific
heat, dynamic viscosity and thermal conductivity, given by Roh-
senow et al. [68], are reported for water and Al2O3. The presence of
nanoparticles and their concentrations influence the mixture
properties. A single-phase model was adopted and the following
equations were used to compute the thermal and physical prop-
erties of the considered nanofluids [28,48,57,58,61], given in
Table 3. Density was evaluated by using the classical formula
developed for conventional solideliquid mixtures while the
specific heat values were obtained by assuming the thermal equi-
librium between particles and surrounding fluid [28,48,58,61].
density : rnf ¼ ð1  fÞrbf þ frp
specific heat : ðrcpÞnf ¼ ð1  fÞðrcpÞbf þ fðrcpÞp


Nuav =Nuav;s
PEC ¼ (34)
ðf =fs Þ1=3
The two-dimensional channel model, depicted in Fig. 1, has
a length, L, equal to 250 mm while the height H is set equal to
10 mm and width W is 1000 mm; the hydraulic diameter, d ¼ 2H, is
equal to 20.0 mm. The 3 mm thick wall is made up by aluminum
and a uniform heat flux equal to 20 kW/m2 has been applied. In this
analysis there are two families, characterized by different e/d ratios,
and by changing p/e ones. The range of dimensionless roughness
parameters and Reynolds numbers employed in this investigation
are given below:
 Reynolds number, Re from 20,000 to 60,000;
(27)  Relative roughness pitch, p/e from 4 to 20;
 Dimensionless height, e/d 0.02 and 0.05.
(28)

Nanofluids may be considered as Newtonian fluids for low 5. Numerical procedure

volume fractions, such as up to 10%, and for small temperature
increases. In this way, for the viscosity as well as for thermal A steady-state solution and a segregated method are considered
conductivity, formula given by Refs. [28,48,58,61] was adopted. to solve the governing equations, linearized implicitly with respect
  to dependent variables. A second-order upwind scheme for energy
dynamic viscosity : mnf ¼ mbf 123f2 þ 7:3f þ 1 (29) and momentum equations and the SIMPLE coupling one in order to
couple pressure and velocity are chosen [65]. The residuals for the
  velocity components and energy are set equal to 105 and 108,
thermal conductivity : knf ¼ kbf 4:97f2 þ 2:72f þ 1 (30) respectively. It is assumed that the incoming flow is turbulent with
a turbulent intensity of 1% and at ambient temperature and pres-
sure. Different inlet uniform velocities are applied and they are
4. Geometrical model and data reduction
reported in Table 4. Along the solid walls the no-slip condition is
employed whereas a velocity inlet and outflow conditions are given
The channel has square and rectangular ribs on the walls and
for the permeable walls.
different geometry ratios are considered, as depicted in Fig. 1b and
Four different grid arrangements are tested on the channel with
c. The roughness parameters are determined by rib height (e), pitch
rectangular ribs, characterized by p/e ¼ 10, e/d ¼ 0.05 and w/e ¼ 0.5,
(p), width (w), and shape of turbulators. These parameters have
and water as working fluid in order to perform the grid indepen-
been expressed in form dimensionless roughness parameters:
dence analysis. They have 55,400, 108,100, 189,900 and 385,000
relative pitch, p/e, relative height, e/d, and relative width, w/e. For
nodes, respectively. The grid mesh is structured in each case and the
a specific type, a family of geometrically similar roughness is
grid adoption for yþ ¼ 1 at adjacent wall region has been taking into
possible to identify by changing e/d ratio while maintaining
account in the mesh generation. The third grid case has been
constant p/e and w/e.
adopted because it ensured a good compromise in terms of machine
The dimensionless parameters, here considered, are the Rey-
computational time and accuracy requirements. In fact, comparing
nolds number, the Nusselt number, the friction factor and the PEC
the third and fourth mesh configurations, in terms of wall average
index and they are expressed by the following relations:
temperature, mass-weighted average outlet temperature, average
uav d
Re ¼ (31)
n
Table 4
Inlet velocities [m/s].
Table 3
Properties of considered nanofluids, single-phase model. Re Water Water/alumina 1% Water/alumina 4%
20,000 0.86 0.90 1.14
4 r [kg/m3] cp [J/kg K] m [Pa s] l [W/m K] 30,000 1.28 1.36 1.71
0.00 998.2 4182 998e-6 0.597 40,000 1.71 1.81 2.29
0.01 1027 4053 1083e-6 0.614 50,000 2.14 2.26 2.86
0.04 1113 3707 1486e-6 0.667 60,000 2.57 2.71 3.43
 O. Manca et al. / Applied Thermal Engineering 37 (2012) 280e292 285

outlet velocity and pressure drops, the corresponding percentage
relative errors are 0.9%, 0.3% and 1.5%, respectively.
Results were validated by comparing the obtained numerical
data with the literature ones in terms of average Nusselt numbers
and friction coefficients, in the case of smooth channel and ribbed
one with water, as reported in Ref. [66] or nanofluid. In the case of
watereAl2O3 nanofluid, the validation has been performed only on
a smooth channel, considering a fully developed flow. The
following correlations are adopted for water [69]:
0:8
DittuseBoelter correlation : Nus ¼ 0:024Re Pr
for 0.5 < Pr < 120, 6.0  103 < Re < 1.0  107; (L/d) > 60;
RePrf =8
Petukhov correlation : Nus ¼

1:07 þ 12:7ðf =8Þ1=2 Pr2=3  1
with
McAdams correlation fs ¼ 0:184Re0:2 (38)
for 2.0  104 < Re < 3.0  105.
The considered correlations for the Al2O3ewater mixture
flowing in the smooth channel are [48,57,58,61]:
PakeChoi correlation : Nus ¼ 0:021Re0:8 Pr0:5 (39)
Maiga et al: correlation : Nus ¼ 0:085Re0:71 Pr0:35 (40)
0:4
(35)
There is a very good correspondence with the given correlations
in terms of average Nusselt number, as reported in Fig. 2a and
b while Fig. 2c shows the small percentage error in terms of friction
factor and a good fitting with the experimental curves. A difference
of 5% at most has been calculated by comparing results with ones
for Petukhov correlation.
(36)
6. Results and discussion

Results are reported in terms of average Nusselt number,

fs ¼ ð1:84 log10 Re  1:64Þ2 (37)
convective heat transfer coefficient, PEC index, required pumping
for 0.5 < Pr < 200, 1.0  104 < Re < 5.0  106; power profiles, as a function of Reynolds number, ranging from

a 450
b 450
water (present model) water (present model)
Dittus-Boelter Dittus-Boelter
400 Petukhov 400 Petukhov
water-Al2O3 1% (present model) water-Al2O 3 4% (present model)
Pak-Choi, water-Al2O3 1% Pak-Choi, water-Al2O 3 4%
Maiga et al., water-Al2O3 1% Maiga et al., water-Al2O 3 4%
350 350

300 300

250 250

200 200

150 150

20000 30000 40000 50000 60000 20000 30000 40000 50000 60000
Re Re

c
0.026

water (present model)

water-Al2O 3 1% (present model)
water-Al2O 3 4% (present model)
0.024 McAdams
Petukhov
f

0.022

0.02

20000 30000 40000 50000 60000

Re
Fig. 2. Validation of results: a) average Nusselt number, 4 ¼ 1%; b) average Nusselt number, 4 ¼ 4%; c) friction factor.
286 O. Manca et al. / Applied Thermal Engineering 37 (2012) 280e292

a 2.4
b 2.8
ϕ = 0.0
e/d = 0.02 ϕ = 0.0
2.3 2.6 ϕ = 0.01
ϕ = 0.01
2.2 e/d = 0.05 ϕ = 0.04
2.4
ϕ = 0.04
Square ribs, Re = 40000
2.1 2.2
Nuav /Nuav,s

Nuav/Nuav,s
2 2

1.9 1.8

ϕ = 0.01
1.8 1.6
ϕ = 0.04 Re = 30000
ϕ = 0.0
1.7 ϕ = 0.01 1.4
Re = 60000 Square ribs
ϕ = 0.04 e/d = 0.05
ϕ = 0.0
1.6 1.2
5 10 15 20 5 10 15 20
p/e p/e

Fig. 3. Average Nusselt number profiles, referred to the smooth channel results in water, as a function of p/e, for square ribs and 4 ¼ 0.00, 0.01 and 0.04: a) e/d ¼ 0.05, Re ¼ 30,000
and 60,000; b) e/d ¼ 0.02 and 0.05, Re ¼ 40,000.

w/e = 0.5
2.5
2.4 Rectangular ribs (w/e = 0.5)
Re = 40000
2.4 Rectangular ribs (w/e = 0.5)
e/d = 0.05
2.3 2.2

2.2
2
Nuav /Nuav,s
Nuav/Nuav,s

2.1

1.8
2

1.9 1.6 ϕ = 0.0

ϕ = 0.0
ϕ = 0.0 e/d = 0.02
1.8 ϕ = 0.01 Re = 30000
ϕ = 0.01
ϕ = 0.04
1.4 ϕ = 0.01
ϕ = 0.0
1.7 ϕ = 0.04 e/d = 0.05
ϕ = 0.01 Re = 60000
ϕ = 0.04
ϕ = 0.04
1.6 1.2
5 10 15 20 5 10 15 20
p/e p/e

w/e = 2.0

2.4 Rectangular ribs (w/e = 2.0) Rectangular ribs (w/e = 2.0)

2.4
e/d = 0.05 Re = 40000
2.3
2.2
2.2

2.1 2
Nuav/Nuav,s

Nuav/Nuav,s

2 1.8

1.9
1.6
1.8 ϕ = 0.0 ϕ = 0.0
ϕ = 0.01 Re = 30000 ϕ = 0.0
1.4 e/d = 0.02
ϕ = 0.04 ϕ = 0.01
1.7
ϕ = 0.0 ϕ = 0.01
ϕ = 0.01 Re = 60000 ϕ = 0.04 e/d = 0.05
1.6 1.2 ϕ = 0.04
ϕ = 0.04
5 10 15 20 5 10 15 20
p/e p/e

a b
Fig. 4. Average Nusselt number profiles, referred to the smooth channel results in water, as a function of p/e, for rectangular ribs and 4 ¼ 0.00, 0.01 and 0.04: a) e/d ¼ 0.05,
Re ¼ 30,000 and 60,000; b) e/d ¼ 0.02 and 0.05, Re ¼ 40,000.
 O. Manca et al. / Applied Thermal Engineering 37 (2012) 280e292 287

a 3.2
ϕ = 0.0
b 3.2
ϕ = 0.0
e/d = 0.02 ϕ = 0.0
3 ϕ = 0.0 e/d = 0.02 3
ϕ = 0.01
ϕ = 0.01
ϕ = 0.01
ϕ = 0.01 2.8 e/d = 0.05
2.8 ϕ = 0.04
ϕ = 0.04 e/d = 0.05
2.6 ϕ = 0.04
ϕ = 0.04
2.6
Square ribs, Re = 40000
2.4
2.4

hav/hav,s
hav/hav,s

2.2
2.2
2
2
1.8
1.8 1.6

1.6 1.4
Rectangular ribs (w/e = 0.5)
1.4 1.2 Re = 40000

1.2 1
5 10 15 20 5 10 15 20
p/e p/e

c 3.2
Rectangular ribs (w/e = 2.0)
3 Re = 40000

2.8

2.6

2.4
hav/hav,s

2.2

1.8

1.6 ϕ = 0.0
ϕ = 0.0 e/d = 0.02
1.4 ϕ = 0.01
ϕ = 0.01
1.2 ϕ = 0.04 e/d = 0.05
ϕ = 0.04
1
5 10 15 20
p/e

Fig. 5. Average convective heat transfer coefficient profiles, referred to the smooth channel results in water, as a function of p/e at Re ¼ 40,000, 4 ¼ 0.00, 0.01 and 0.04, e/d ¼ 0.05: a),
square ribs; b) rectangular ribs, w/e ¼ 0.5; c) rectangular ribs, w/e ¼ 2.0.

20,000 to 60,000, roughness pitch values, from 4 to 20, relative
height of 0.02 and 0.05, and particle volume concentrations of 0%,
1% and 4%. The ribbed channel presents square and rectangular ribs,
with w/e ¼ 0.5 and 2.0, mounted on both the principal walls, heated
by a uniform heat flux.
The average Nusselt number profiles, referred to the results
obtained in the cases with water in smooth channels, as a function
of p/e, for different Re, e/d and 4 values, are plotted in Figs. 3 and 4
for both the considered shapes of turbulators. In particular, Fig. 4
shows the results for square elements: it is shown that the

a 35
b 30

30
25

25 ϕ = 0.0
20 ϕ = 0.0
e/d = 0.02
ϕ = 0.01
ϕ = 0.0
PP/PPs

PP/PPs

20 ϕ = 0.01 e/d = 0.05

ϕ = 0.01 Re = 30000 ϕ = 0.04
ϕ = 0.04 15
ϕ = 0.04
ϕ = 0.0 Square ribs
15 ϕ = 0.01 Re = 60000
Re = 40000
ϕ = 0.04
10

10

5
Square ribs
5 e/d = 0.05

5 10 15 20 5 10 15 20
p/e p/e

Fig. 6. Required pumping power profiles, referred to the smooth channel results in water, as a function of p/e, for square ribs and 4 ¼ 0.00, 0.01 and 0.04: a) e/d ¼ 0.05, Re ¼ 30,000
and 60,000; b) e/d ¼ 0.02 and 0.05, Re ¼ 40,000.
288 O. Manca et al. / Applied Thermal Engineering 37 (2012) 280e292

w/e = 0.5
35 30

30 25
ϕ = 0.0
ϕ = 0.0 e/d = 0.02
25 ϕ = 0.01
20 ϕ = 0.01
ϕ = 0.0 ϕ = 0.04 e/d = 0.05
ϕ = 0.01
PP/PPs

ϕ = 0.04

PP/PPs
20
Re = 30000
ϕ = 0.04 Rectangular ribs (w/e = 0.5)
ϕ = 0.0 15
Re = 40000
ϕ = 0.01 Re = 60000
15 ϕ = 0.04
10

10

Rectangular ribs (w/e = 0.5) 5

5 e/d = 0.05

5 10 15 20 5 10 15 20
p/e p/e

w/e = 2.0
35 30
ϕ = 0.0 ϕ = 0.0
ϕ = 0.01 Re = 30000 ϕ = 0.0 e/d = 0.02
ϕ = 0.04 ϕ = 0.01
30
ϕ = 0.0 25 ϕ = 0.01
ϕ = 0.01 Re = 60000 ϕ = 0.04 e/d = 0.05
ϕ = 0.04 ϕ = 0.04
25 Rectangular ribs (w/e = 2.0) 20
e/d = 0.05
PP/PPs

PP/PPs

20
15
Rectangular ribs (w/e = 2.0)
Re = 40000
15
10

10

5
5

5 10 15 20 5 10 15 20
p/e p/e

a b
Fig. 7. Required pumping power profiles, referred to the smooth channel results in water, as a function of p/e, for rectangular ribs and 4 ¼ 0.00, 0.01 and 0.04: a) e/d ¼ 0.05,
Re ¼ 30,000 and 60,000; b) e/d ¼ 0.02 and 0.05, Re ¼ 40,000.

a 1.4 b 1.5

1.4
1.3
1.3

1.2 1.2

1.1
1.1
PEC

PEC

1
1
0.9 ϕ = 0.0, e/d = 0.02
ϕ = 0.01, e/d = 0.02
ϕ = 0.04, e/d = 0.02
0.9 ϕ = 0.0 0.8 ϕ = 0.0, e/d = 0.05 w/e = 0.5
ϕ = 0.01, e/d = 0.05
ϕ = 0.01 e/d = 0.02 Square ribs ϕ = 0.04, e/d = 0.05
ϕ = 0.04 Re = 40000 0.7 ϕ = 0.0, e/d = 0.02
ϕ = 0.01, e/d = 0.02
0.8 ϕ = 0.0 ϕ = 0.04, e/d = 0.02 w/e = 2.0
ϕ = 0.01 e/d = 0.05 0.6 ϕ = 0.0, e/d = 0.05 Re = 40000
ϕ = 0.01, e/d = 0.05
ϕ = 0.04 ϕ = 0.04, e/d = 0.05 Rectangular ribs
0.7 0.5
5 10 15 20 5 10 15 20
p/e p/e

Fig. 8. PEC index profiles, referred to the smooth channel results in water, as a function of p/e at Re ¼ 40,000, 4 ¼ 0.00, 0.01 and 0.04, e/d ¼ 0.05: a), square ribs; b) rectangular ribs.
 O. Manca et al. / Applied Thermal Engineering 37 (2012) 280e292 289

maximum values of Nuav/Nuav,s ratio are observed at the same
pitch, revealed in the case of ribbed channels in water, such as at p/
e ¼ 8. This result is confirmed for all the considered Reynolds
number and e/d values, as reported in Fig. 3a and b. The introduc-
tion of nanoparticles provides higher Nusselt number values, due to
the larger thermal conductivity of nanofluids, and this behavior is
more significant as volume concentration increases, as shown in
Fig. 3a. In fact, for Re ¼ 30,000 Nuav/Nuav,s is equal to 1.99, 2.03 and
2.16 at most for 4 ¼ 0.0, 0.01 and 0.04. The presence of smaller
turbulators leads to a decrease of 15e20% in terms of Nusselt
number, as depicted in Fig. 3b.
The results for the rectangular turbulators, with w/e ¼ 0.5 and
2.0, are presented in Fig. 4. The highest values of average Nusselt
number is evaluated at p/e ¼ 10, for both the geometric ratios.
However, ribs with w/e ¼ 0.5 result to have more significant
thermal performance since they provide Nuav/Nuav,s values equal to
2.39 at most for 4 ¼ 0.04 at Re ¼ 60,000 and e/d ¼ 0.05, as reported
in Fig. 4a. Smaller elements lead to lower Nusselt number values.
The increase of thermal performances is clearer if Fig. 5 is
observed. It depicts the results in terms of average convective heat
transfer coefficients, referred to the smooth channel in water, for all
the considered shapes at e/d ¼ 0.02 and 0.05, for Re ¼ 40,000 and
4 ¼ 0.0, 0.01 and 0.04. In fact, hav/hav,s ¼ 2.00, 2.14 and 2.50 at e/
d ¼ 0.05 and Re ¼ 40,000 in the configuration with square ribs for
4 ¼ 0.00, 0.01 and 0.04, respectively while hav/hav,s ¼ 1.94, 2.10 and
2.44 in the case of rectangular ribs with w/e ¼ 2.0, as shown in
Fig. 5c. However, if the thermal performances result to be enhanced
by the employment of rough surfaces and nanofluids, higher
pressure drops are expected. Figs. 6 and 7 depict :
the required
pumping power profiles, PP, defined by PP ¼ V DP and referred to
the smooth channel in water. The maximum values of PP/PPs are

a 1000 smooth
a 250
smooth smooth
square, p/e = 8 smooth
square, p/e = 8 square, p/e = 12
rectangular w/e = 0.5, p/e = 10 200 square, p/e = 12
800 rectangular w/e = 0.5, p/e = 10 rectangular w/e = 0.5, p/e = 10
rectangular w/e = 2.0, p/e = 10 rectangular w/e = 0.5, p/e = 10
rectangular w/e = 2.0, p/e = 10 rectangular w/e = 2.0, p/e = 10
150 rectangular w/e = 2.0, p/e = 10
PP [W]

600 e/d = 0.05

Nuav

e/d = 0.05
water
100 water/Al2O3 1%
400

water 50

200
water/Al2O3 1%
0
20000 30000 40000 50000 60000 20000 30000 40000 50000 60000
Re Re

b 1200 smooth b 600

smooth
square, p/e = 8
square, p/e = 8 500 smooth
1000
rectangular w/e = 0.5, p/e = 10 smooth
rectangular w/e = 0.5, p/e = 10 square, p/e = 12
rectangular w/e = 2.0, p/e = 10 square, p/e = 12
rectangular w/e = 2.0, p/e = 10 400
800 rectangular w/e = 0.5, p/e = 10
rectangular w/e = 0.5, p/e = 10
PP [W]

rectangular w/e = 2.0, p/e = 10

Nuav

300 rectangular w/e = 2.0, p/e = 10

600 e/d = 0.05
e/d = 0.05
200 water
400
water/Al2O3 4%
water 100
200
water/Al2O3 4%
20000 30000 40000 50000 60000 0
20000 30000 40000 50000 60000
Re Re
Fig. 9. Average Nusselt number profiles as a function of Re in the range 20,000e60,000 Fig. 10. Required pumping power profiles as a function of Re in the range
at e/d ¼ 0.05 for different rib shapes and significant pitches: a) 4 ¼ 0.00 and 0.01; b) 20,000e60,000 at e/d ¼ 0.05 for different rib shapes and significant pitches: a) 4 ¼ 0.00
4 ¼ 0.00 and 0.04. and 0.01; b) 4 ¼ 0.00 and 0.04.
290 O. Manca et al. / Applied Thermal Engineering 37 (2012) 280e292
evaluated at p/e ¼ 12 in the case of square ribs, both for e/d ¼ 0.02
and 0.05, as described by Fig. 6. Profiles show larger values for
increasing Reynolds number and particle volume concentrations. In
fact, for e/d ¼ 0.05 and Re ¼ 30,000, PP/PPs is equal to about 8, 10, 22
at most for 4 ¼ 0.00, 0.01 and 0.04, respectively. Smaller ribs
provide moderate losses, as depicted in Fig. 6b, which shows
a decrease of 2.5 times on average.
The rectangular turbulators present the maxima of PP/PPs ratio
at p/e ¼ 10, as pointed out by Fig. 7; moreover, ribs with w/e ¼ 2.0
result to have both lower thermal performances and losses in
comparison with square and rectangular elements with w/e ¼ 0.5.
They show values of about 7, 9 and 18 in Fig. 7a at Re ¼ 30,000 for
4 ¼ 0.00, 0.01 and 0.04, respectively whereas for the rectangular
ribs with w/e ¼ 0.5 values of PP/PPs are equal to about 9, 14 and 25.
Fig. 7b confirms the significant decrease of the required pumping
power in the configurations with small dimensionless height.
The use of special surfaces, like rough surfaces, in order to
enhance the heat transfer in thermal devices implicates the
increase of pumping power, sometimes significantly. The rib height
represents the dominant factor in losses, as underlined by Ref. [27],
but also the shape, the flow feature and the fluid type have their
influence [1,2,16,20,25]. Therefore, the introduction of nanoparticle
leads to the use of working fluids, featured by higher values of
density and viscosity (besides thermal conductivity) than water
and a significant growth of required pumping is expected. However,
no experimental tests have been accomplished for the configura-
tions, presented in this paper, which describes the first example
ever concerning the use of nanofluids in ribbed channels. However,
several authors have already reported the significant increase of
pumping power even in smooth tubes or channels, filled with
nanofluids [39,48e50,53e61].
The PEC index profiles, presented in Fig. 8, show that the most
efficient configurations are represented by ones with p/e ¼ 8. The
highest values of the index are provided by the rectangular ribs
with w/e ¼ 2.0, followed by the square turbulators and rectangular
ones with w/e ¼ 0.5, as pointed out by Fig. 8a and b. Furthermore,
profiles grow as volume particle concentration increases and height
decreases. In fact, PEC index is equal to 1.4 at most for Re ¼ 40,000,
e/d ¼ 0.02 and 4 ¼ 0.04 in the case of rectangular ribs with w/
e ¼ 2.0. The optimal p/e value is 8 but also for smaller p/e values PEC
index values are high.
The average Nusselt number increases as Re and 4 increase, as
pointed out in Fig. 9, which is concerned with the results obtained
for some significant pitches and different shapes with e/d ¼ 0.05.
The rectangular elements with w/e ¼ 0.5 show the highest thermal
Fig. 11. Stream function contours for square ribs, e/d ¼ 0.05, Re ¼ 50,000, 4 ¼ 0.04: a) p/e ¼ 4, 4 ¼ 0.01; b) p/e ¼ 4, 4 ¼ 0.04; c) p/e ¼ 8, 4 ¼ 0.01; d) p/e ¼ 8, 4 ¼ 0.04.
performances but also the worst behavior in terms of required
pumping power, as observed in Fig. 10. The minimum pumping
power is required by the rectangular elements with w/e ¼ 2.0 and it
is important to underline the significant increase of PP with Re and
particle volume concentration: at 4 ¼ 0.04 PP values are about two
times greater than ones calculated at 4 ¼ 0.01 and 10 times in
comparison with water cases, at an assigned value of Re.
Fig. 11 depicts the flow pattern in the inter-rib region at
Re ¼ 50,000 for square elements when the flow is stabilized in the
cases with p/e ¼ 4 and 8 and 4 ¼ 0.01 and 0.04. The stabilization
happens after 4e5 ribs, such as when the vortex structures tend to
repeat almost similarly from this zone, as already pointed out in
Ref. [66]. Fig. 11a and b refers to the behavior at small pitch ratios
(p/e < 6) and exhibits a d-type roughness behavior: a single eddy is
trapped between adjacent ribs and hinders the heat transfer in
a measure. Moreover, the outer flow is relatively disturbed by the
roughness elements. This behavior repeats at higher volume frac-
tion values but the vortex structures are more intense because of
the nanofluid properties. Other parameters influencing the vortex
intensity are the height of elements and Reynolds number. Fig. 11c
and d depicts the stream function contours at Re ¼ 50,000,
respectively, for p/e > 6 in order to describe the k-type roughness
behavior: the stream functions change because the y-velocity
gradient reduces along the fluid direction. As a result, the flow
separates then it reattaches after six-to-eight rib heights down-
stream from the turbulators. Near this point, the Nusselt number is
expected to be maximum as well as the friction factor. In this case
eddies penetrate into the bulk flow toward boundary layer and
promoting a more intense mixing. Moreover, the increase of
particle volume fraction does not lead to significant differences in
terms of stream function contours but velocity increases at a fixed
Reynolds number value, because of the viscosity increase not
balanced by the density growth.
Fig. 12 describes the dimensionless temperature fields for
significant pitches at Re ¼ 50,000 in the cases of square and rect-
angular ribs with w/e ¼ 0.5 and 2.0, characterized by e/d ¼ 0.05. The
dimensionless temperature, T*, is defined by the following relation:


T  Tm;in
T* ¼
 (41)
Tm;out  Tm;in
Fig. 12a depicts the temperature field, typically observed for the
configurations with square ribs, characterized by a k-type behavior
since the value of p/e is greater than 6. In the figure on the left,
referring to 4 ¼ 0.01, the fluid temperature is high near the rib-wall
 O. Manca et al. / Applied Thermal Engineering 37 (2012) 280e292
Fig. 12. Dimensionless temperature fields for different shapes, e/d ¼ 0.05, Re ¼ 50,000, 4 ¼ 0.01 and 0.04: a) square ribs, p/e ¼ 8; b) rectangular ribs w/e ¼ 0.5, p/e ¼ 10; c) rect-
angular ribs w/e ¼ 2.0, p/e ¼ 10.
juncture behind the turbulator. However, the fluid temperature is
slightly increased while the wall one is decreased in comparison
with the configuration with small pitches, where the core region is
poorly influenced. In fact, for the k-type behavior the mixing effect
is stronger and the difference between the average fluid tempera-
ture and wall one decreases and the heat transfer mechanism
becomes more efficient, as also described in Fig. 12b and c. If the
volume fraction values increase, the difference between the
average fluid temperature and wall one decreases, because the fluid
thermal conductivity has risen, as shown in the figures on the right.
Moreover, all the figures show that the temperature is lowest on
the windward face of the rib, and highest on the leeward face, while
heat flux is highest on the front face and lowest on the back face.
7. Conclusions
In this work a 2-D ribbed channel with square and rectangular
ribs, mounted on the principal walls and heated by a uniform heat
flux, was considered. Turbulators had e/d values equal to 0.02 and
0.05, and different pitches and Reynolds numbers were studied.
The fluid was a mixture of water and Al2O3 nanoparticles. Simu-
lations show that the introduction of nanoparticles augments both
the heat transfer coefficients and the pressure drops in the ribbed
channel. The highest Nusselt number values were evaluated for p/
e ¼ 8 and 10 for square and rectangular shapes, respectively.
Regarding the required pumping power values, maxima were
calculated at p/e ¼ 10 for both the square and rectangular elements.
The highest thermal performances were exhibited by the rectan-
gular turbulators with w/e ¼ 0.5 which provide also the highest
291
losses. The Nusselt number rises as Reynolds numbers increase as
well as the pumping power profiles. For increasing concentrations
of nanoparticles Nusselt number values, heat transfer rates and
pumping power increase and their values are 2.4, 2.66 and 31 times
the ones calculated for pure water in a smooth channel in the case
of rectangular ribs with w/e ¼ 2.0 and 4 ¼ 0.04 at most. The highest
values of PEC index are detected at p/e ¼ 8 and they increase as e/
d ratio decreases.
Acknowledgements
This work was supported by SUN with a 2008 grant and by MIUR
with Articolo D.M. 593/2000 Grandi Laboratori “EliosLab”.
References
[1] A.E. Bergles, Some perspectives on enhanced heat transfer, second-generation
heat transfer technology, J. Heat Transf. 110 (2000) 1082.
[2] R.L. Webb, N.H. Kim, Principles of Enhanced Heat Transfer, second ed. Taylor &
Francis Group, New York, NY, 2006.
[3] D. Westphalen, K. Roth, J. Brodrick, Heat transfer enhancement, ASHRAE J. 48
(2006) 68e71.
[4] S.U.S. Choi, Nanofluids: from vision to reality through research, J. Heat Transf.
131 (2009) 1e9.
[5] R.L. Webb, Advances in shell side boiling of refrigerants, J. Inst. Refrig. 87
(1991) 75e86.
[6] R.L. Webb, G.F. Robertson, Shell-side evaporators and condensers used in the
refrigeration industry, in: R.K. Shah, E.C. Subbarao, R.A. Mashelkar (Eds.), Heat
Transfer Equipment Design, Hemisphere Pub. Corp, Washington, 1988, pp.
559e570.
[7] M.H. Jaber, R.L. Webb, P. Stryker, An experimental investigation of enhanced
tubes for steam condensers, ASME Paper (1991) 1e8.
292 O. Manca et al. / Applied Thermal Engineering 37 (2012) 280e292
[8] B. Sunden, G. Xie, Gas turbine blade tip heat transfer and cooling: a literature
survey, Heat Transf. Eng. 31 (2010) 527e554.
[9] R. Karwa, S.C. Solanky, J.S. Saini, Thermo-hydraulic performance of solar air
heaters having integral chamfered rib roughness on absorber plates, Energy
26 (2001) 161e176.
[10] K.B. Muluwork, S.C. Solanky, J.S. Saini, Study of heat transfer and friction in
solar air heaters roughened with staggered discrete ribs, Proceedings of the
fourth ISHMT-ASME Heat and Mass Transfer Conference 2000, Pune, India,
pp. 391e400.
[11] R. Karwa, S.C. Solanky, J.S. Saini, Heat transfer coefficient and friction factor
correlations for the transient flow regime in rib-roughened rectangular ducts,
Int. J. Heat Mass Transf. 42 (1999) 1597e1615.
[12] C.K. Lee, S.A. Abdel Moneim, Computational analysis of heat transfer in
turbulent flow past a horizontal surface with a 2-D ribs, Int. Commun. Heat
Mass Transf. 26 (2001) 161e170.
[13] L. Wang, B. Sunden, Experimental investigation of local heat transfer in
a square duct with various-shaped ribs, Int. J. Heat Mass Transf. 43 (2006)
759e766.
[14] P. Promvonge, C. Thianpong, Thermal performance assessment of turbulent
channel flows over different shaped ribs, Int. Commun. Heat Mass Transf. 35
(2008) 1327e1334.
[15] S.K. Saha, Thermal and friction characteristics of turbulent flow through
rectangular and square ducts with transverse ribs and wire-coil inserts, Exp.
Therm. Fluid Sci. 34 (2010) 575e589.
[16] B.K. Lee, N.H. Cho, Y.D. Choi, Analysis of periodically fully developed turbulent
flow and heat transfer by keε equation model in artificially roughened
annulus, Int. J. Heat Mass Transf. 31 (1988) 1797e1806.
[17] R. Manceau, S. Parneix, D. Laurence, Turbulent heat transfer predictions
using the model on unstructured meshes, Int. J. Heat Fluid Flow 21 (2000)
320e328.
[18] D.K. Tafti, Evaluating the role of subgrid stress modeling in a ribbed duct for
the internal cooling of turbine blades, Int. J. Heat Fluid Flow 26 (2005)
92e104.
[19] A. Slanciauskas, Two friendly roles for the turbulent heat transfer enhance-
ment, Int. J. Heat Mass Transf. 44 (2001) 2155e2161.
[20] T.M. Liou, J.J. Hwang, S.H. Chen, Simulation and measurement of enhanced
turbulent heat transfer in a channel with periodic ribs on one principal wall,
Int. J. Heat Mass Transf. 36 (1993) 507e517.
[21] G. Rau, M. Cakan, D. Moeller, T. Arts, The effect of periodic ribs on the local
aerodynamic and heat transfer performance of a straight cooling channel,
J. Turbomach. 120 (1998) 368e375.
[22] D.N. Ryu, D.H. Choi, V.C. Patel, Analysis of turbulent flow in channels rough-
ened by two-dimensional and three-dimensional blocks. Part I: resistance, Int.
J. Heat Fluid Flow 28 (2007) 1098e1111.
[23] D.N. Ryu, D.H. Choi, V.C. Patel, Analysis of turbulent flow in channels rough-
ened by two-dimensional and three-dimensional blocks. Part II: heat transfer,
Int. J. Heat Fluid Flow 28 (2007) 1112e1124.
[24] A. Chaube, P.K. Sahoo, S.C. Solanki, Analysis of heat transfer augmentation and
flow characteristics due to rib roughness over absorber plate of a solar air
heater, Renew. Energy 31 (2005) 317e331.
[25] O. Manca, S. Nardini, D. Ricci, Numerical investigation of air forced convection
in channels with differently shaped transverse ribs, Int. J. Numer. Method
Heat Fluid Flow 21 (2010) 618e639. doi:10.1108/09615531111135855.
[26] K.M. Kim, B.S. Kim, D.L. Lee, H.K. Moon, H.H. Cho, Optimal design of transverse
ribs in tubes for thermal performance enhancement, Energy 35 (2010)
2400e2406.
[27] H. Liu, J. Wang, Numerical investigation on synthetical performances of fluid
flow and heat transfer of semiattached rib-channels, Int. J. Heat Mass Transf.
54 (2011) 575e583.
[28] S.E.B. Maiga, C.T. Nguyen, N. Galanis, G. Roy, T. Mare, M. Coqueux, Heat
transfer enhancement in turbulent tube flow using Al2O3 nanoparticle
suspension, Int. J. Numer. Method Heat Fluid Flow 16 (2006) 275e292.
[29] S. Özerinç, S. Kakaç, A.G. Yazıcıog lu, Enhanced thermal conductivity of
nanofluids: a state-of-the-art review, Microfluid Nanofluid 8 (2010) 145e170.
[30] Y.T. Yang, F.H. Lai, Numerical study of heat transfer enhancement with the use
of nanofluids in radial flow cooling system, Int. J. Heat Mass Transf. 53 (2010)
5895e5904.
[31] Q. Li, Y. Xuan, Heat transfer enhancement of nanofluids, Int. J. Heat Fluid Flow
21 (2000) 58e64.
[32] V. Trisaksri, S. Wongwises, Critical review of heat transfer characteristics of
nanofluids, Renew. Sustain. Energy Rev. 11 (2007) 512e523.
[33] L. Godson, B. Raja, D.M. Lal, S. Wongwises, Enhancement of heat transfer using
nanofluids e an overview, Renew. Sustain. Energy Rev. 14 (2010) 629e641.
[34] D. Wen, L. Lin, S. Vafaei, K. Zhang, Review of nanofluids for heat transfer
applications, Particuology 7 (2009) 141e150.
[35] J.A. Eastman, S.R. Phillipot, S.U.S. Choi, P. Keblinski, Thermal transport in
nanofluids, Annu. Rev. Mater. Res. 34 (2004) 219e246.
[36] L. Cheng, Nanofluid heat transfer technologies, Recent Pat. Eng. 3 (2009) 1e7.
[37] J.U. Ahamed, R. Saidur, H.H. Masjuki, A review on exergy analysis of vapor
compression refrigeration system, Renew. Sustain. Energy Rev. 15 (2011)
1593e1600.
[38] P. Keblinski, R. Prasher, J. Eapen, Thermal conductance of nanofluids: is the
controversy over? J. Nanopart. Res. 10 (2008) 1089e1097.
[39] I. Gherasim, G. Roy, C.T. Nguyen, D. Vo-Ngoc, Heat transfer enhancement and
pumping power in confined radial flows using nanoparticle suspensions
(nanofluids), Int. J. Therm. Sci. 50 (2011) 369e377.
[40] S.M.S. Murshed, K.C. Leong, C. Yang, Thermophysical and electrokinetic
properties of nanofluids e a critical Review, Appl. Therm. Eng. 28 (2008)
2109e2125.
[41] C.H. Chon, K.D. Kihm, S.P. Lee, S.U.S. Choi, Empirical correlation finding the
role of temperature and particle size for nanofluid (Al2O3) thermal conduc-
tivity enhancement, Appl. Phys. Lett. 87 (2005) 1e3.
[42] H.U. Kang, S.H. Kim, J.M. Oh, Estimation of thermal conductivity of nanofluids
using experimental effective particle volume, Exp. Heat Transf. 19 (2006)
181e191.
[43] X. Zhang, H. Gu, M. Fujii, Effective thermal conductivity and thermal diffu-
sivity of nanofluids containing spherical and cylindrical nanoparticles, Exp.
Therm. Fluid Sci. 31 (2007) 593e599.
[44] D. Venerus, et al., Viscosity measurements on colloidal dispersions (nano-
fluids) for heat transfer applications, Appl. Rheol. 20 (2010) 445e462.
[45] J. Buongiorno, Convective transport in nanofluids, J. Heat Transf. 128 (2006)
240e250.
[46] N. Sohrabi, N. Masoumi, A. Behzadmehr, S.M.H. Sarvari, A simple analytical
model for calculating the effective thermal conductivity of nanofluids, Heat
Transf. e Asian Res. 39 (2010) 141e150.
[47] C. Sitprasert, P. Dechaumphai, V. Juntasaro, A thermal conductivity model for
nanofluids including effect of the temperature-dependent interfacial layer,
J. Nanopart. Res. 11 (2009) 1465e1476.
[48] S.J. Palm, G. Roy, C.T. Nguyen, Heat transfer enhancement with the use of
nanofluids in radial flow cooling systems considering temperature dependent
properties, Appl. Therm. Eng. 26 (2009) 2209e2218.
[49] Y. Feng, C. Kleinstreuer, Nanofluid convective heat transfer in a parallel-disk
system, Int. J. Heat Mass Transf. 53 (2010) 4619e4628.
[50] O. Manca, P. Mesolella, S. Nardini, D. Ricci, Numerical study of a confined slot
impinging jet with nanofluids, Nanoscale Res. Lett. 6 (2011) 1e16.
doi:10.1186/1556-276X-6-188.
[51] A. Akbarinia, M. Abdolzadeh, R. Laur, Critical investigation of heat transfer
enhancement using nanofluids in microchannels with slip and non-slip flow
regimes, Appl. Therm. Eng. 31 (2011) 556e565.
[52] Y.M. Xuan, Q. Li, Investigation on convective heat transfer and flow features of
nanofluids, J. Heat Transf. 125 (2003) 151e155.
[53] D. Wen, Y. Ding, Experimental investigation into convective heat transfer of
nanofluids at the entrance region under laminar flow conditions, Int. J. Heat
Mass Transf. 47 (2004) 5181e5188.
[54] S.K. Hwang, S.P. Jang, S.U.S. Choi, Flow and convective heat transfer charac-
teristics of water-based Al2O3 nanofluids in fully developed laminar flow
regime, Int. J. Heat Mass Transf. 52 (2009) 193e199.
[55] Z. Heris, M.N. Esfahany, S.G. Etemad, Experimental investigation of convective
heat transfer of Al2O3/water nanofluid in circular tube, Int. J. Heat Fluid Flow
28 (2007) 203e210.
[56] W. Williams, J. Buongiorno, L.W. Hu, Experimental investigation of turbulent
convective heat transfer and pressure loss of alumina/water and zirconia/
water nanoparticle colloids (nanofluids) in horizontal tubes, J. Heat Transf.
130 (2008) 1e6.
[57] B.C. Park, Y.I. Cho, Hydrodynamic and heat transfer study of dispersed fluids
with submicron metallic oxide particles, Exp. Heat Transf. 11 (1998) 151e170.
[58] S.E.B. Maiga, C.T. Nguyen, N. Galanis, G. Roy, Heat transfer behaviours of nano-
fluids in a uniformly heated tube, Superlattices Microstruct. 35 (2004) 543.
[59] M.H. Fard, M.N. Esfahany, M.R. Talaie, Numerical study of convective heat
transfer of nanofluids in a circular tube two-phase model versus single-phase
model, Int. Commun. Heat Mass Transf. 37 (2010) 91e97.
[60] V. Bianco, F. Chiacchio, O. Manca, S. Nardini, Numerical investigation of
nanofluids forced convection in circular tubes, Appl. Therm. Eng. 29 (2009)
3632e3642.
[61] S.E.B. Maiga, S.J. Palm, C.T. Nguyen, G. Roy, N. Galanis, Heat transfer
enhancement by using nanofluids in forced convection flows, Int. J. Heat Fluid
Flow 26 (2005) 530e546.
[62] M. Izadi, A. Behzadmehr, D. Jalali-Vahida, Numerical study of developing
laminar forced convection of a nanofluid in an annulus, Int. J. Therm. Sci. 48
(2009) 2119e2129.
[63] M. Rostamani, S.F. Hosseinizadeh, M. Gorji, J.M. Khodadadi, Numerical study
of turbulent forced convection flow of nanofluids in a long horizontal duct
considering variable properties, Int. Commun. Heat Mass Transf. 37 (2010)
1426e1431.
[64] R.L. Hamilton, O.K. Crosser, Thermal conductivity of heterogeneous two
component system, Ind. Eng. Chem. Fundam. 1 (1962) 187e191.
[65] Fluent v.6.3 User Guide, Fluent Corporation, Lebanon, New Hampshire, 2006.
[66] O. Manca, S. Nardini, D. Ricci, Numerical analysis of water forced convection in
channels with differently shaped transverse ribs, J. Appl. Math. 2011 (2011).
doi:10.1155/2011/323485 article ID 323485.
[67] F.R. Menter, Two equation eddy-viscosity turbulence models for engineering
applications, AIAA J. 32 (1994) 1598e1605.
[68] W.M. Rohsenow, J.P. Hartnett, Y.I. Cho, Handbook of Heat Transfer, third ed.
McGraw-Hill, New York, NY, 1998.
[69] F. Kreith, M.S. Bohn, Principle of Heat Transfer, fifth ed. PWS Publishing
Company, Boston, MA, 1997.