Applied Thermal Engineering 30 (2010) 2685e2692

Contents lists available at ScienceDirect

Applied Thermal Engineering

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

Performance investigation of an automotive car radiator operated with

nanofluid-based coolants (nanofluid as a coolant in a radiator)
K.Y. Leong a, b, R. Saidur a, *, S.N. Kazi a, A.H. Mamun c
a
Department of Mechanical Engineering, University of Malaya, Kuala Lumpur, Malaysia
b
Department of Mechanical Engineering, National Defense University of Malaysia, Kuala Lumpur, Malaysia
c
Department of Mechanical Engineering, Bangladesh University of Engineering and Technology, Bangladesh

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

Article history: Water and ethylene glycol as conventional coolants have been widely used in an automotive car radiator
Received 30 March 2010 for many years. These heat transfer fluids offer low thermal conductivity. With the advancement of
Accepted 21 July 2010 nanotechnology, the new generation of heat transfer fluids called, “nanofluids” have been developed and
Available online 29 July 2010
researchers found that these fluids offer higher thermal conductivity compared to that of conventional
coolants. This study focused on the application of ethylene glycol based copper nanofluids in an auto-
Keywords:
motive cooling system. Relevant input data, nanofluid properties and empirical correlations were
Nanofluids
obtained from literatures to investigate the heat transfer enhancement of an automotive car radiator
Automotive radiator
Heat transfer enhancement
operated with nanofluid-based coolants. It was observed that, overall heat transfer coefficient and heat
transfer rate in engine cooling system increased with the usage of nanofluids (with ethylene glycol the
basefluid) compared to ethylene glycol (i.e. basefluid) alone. It is observed that, about 3.8% of heat
transfer enhancement could be achieved with the addition of 2% copper particles in a basefluid at the
Reynolds number of 6000 and 5000 for air and coolant respectively. In addition, the reduction of air
frontal area was estimated.
Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction 15e40% of heat transfer enhancement can be achieved by using

various types of nanofluids. With these superior characteristics, the
Continuous technological development in automotive indus- size and weight of an automotive car radiator can be reduced
tries has increased the demand for high efficiency engines. A high without affecting its heat transfer performance. This translates into
efficiency engine is not only based on its performance but also for a better aerodynamic feature for design of an automotive car frontal
better fuel economy and less emission. Reducing a vehicle weight area. Coefficient of drag can be minimized and fuel consumption
by optimizing design and size of a radiator is a necessity for making efficiency can be improved.
the world green. Addition of fins is one of the approaches to Therefore, this study attempts to investigate the heat transfer
increase the cooling rate of the radiator. It provides greater heat characteristics of an automotive car radiator using ethylene glycol
transfer area and enhances the air convective heat transfer coeffi- based copper nanofluids as coolants. Thermal performance of an
cient. However, traditional approach of increasing the cooling rate automotive car radiator operated with nanofluids is compared with
by using fins and microchannel has already reached to their limit a radiator using conventional coolants. The effect of volume frac-
[1]. In addition, heat transfer fluids at air and fluid side such as tion of the copper nanoparticles with basefluids on the thermal
water and ethylene glycol exhibit very low thermal conductivity. As performance and potential size reduction of a radiator were also
a result there is a need for new and innovative heat transfer fluids carried out. Copper nanoparticles were chosen in this study since it
for improving heat transfer rate in an automotive car radiator. has higher thermal conductivity compared to other nanoparticles
Nanofluids seem to be potential replacement of conventional such as alumina.
coolants in engine cooling system. Recently there have been
considerable research findings highlighting superior heat transfer
performances of nanofluids. Yu et al., [2] reported that about 2. Nanofluids in enhancing thermal conductivity

Eastman et al. [3] reported that the thermal conductivity of

* Corresponding author. Tel.: þ603 79674462; fax: þ603 79675317. ethylene glycol nanofluids containing 0.3% volume fraction
E-mail address: saidur@um.edu.my (R. Saidur). of copper particles can be enhanced up to 40% compared to that of

1359-4311/$ e see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.applthermaleng.2010.07.019
2686 K.Y. Leong et al. / Applied Thermal Engineering 30 (2010) 2685e2692

Nomenclature Re Reynolds number

t fin thickness, m
A total heat transfer area, m2 T temperature, K
Af/A fin area on one side/total area U overall heat transfer coefficient W/m2 K
C heat capacity rate, W/K W mass flow rate
Cp specific heat J/kg K s minimum free flow area/frontal area
C* Cp, min/Cp, max hf fin efficiency of plate fin
Dh hydraulic diameter, m ho total surface temperature effectiveness
DP pressure drop, Pa m dynamic viscosity, Ns/m2
EG ethylene glycol n volumetric flow rate, m3/s
f fanning friction factor, dimensionless r density, kg/m3
G mass velocity, kg/m2s j volume fraction of particles
H total water flow length, m 3 heat exchanger effectiveness
h heat transfer coefficient, W/m2 K a total one side of heat transfer area/total volume
j Colburn factor, dimensionless
k thermal conductivity, W/m K Subscripts
L fin length for heat conduction from primary to fr frontal area
midpoint between plates for symmetric heating, m a air
NTU number of heat transfer units nf nanofluids
Nu Nusselt number f fluid (basefluid)
P pumping power p particles
Pr Prandtl number in inlet

ethylene glycol basefluid. Hwang et al. [4] found that thermal higher heat transfer coefficient compared to that of other types of
conductivity of the nanofluids depends on the volume fraction of nanofluids.
particles and thermal conductivity of basefluid and particles. Lee Ding et al. [9] found that convective heat transfer coefficient
et al. [5] measured the thermal conductivity of low volume of nanofluids has the highest magnitude at the entrance length of
concentration of aqueous alumina (Al2O3) nanofluids produced by a tube. It starts decreasing with axial distance and eventually
two-step method. Authors inferred that the thermal conductivity of accomplish at a constant value in the fully developed region. At
aqueous nanofluids increases linearly with the addition of alumina a given flow and particle concentration, aqueous carbon nano-
particles. Thermal conductivity of zinc dioxideeethylene glycol particles offer highest improvement. Zeinali et al. [10] experi-
(ZnOeEG) based nanofluids was investigated by Yu et al. [6]. They mentally investigated convective heat transfer to aluminaewater
obtained about 26.5% enhancement of thermal conductivity by (Al2O3/water) nanofluids in laminar flow inside a circular tube
adding 5% volume fraction of zinc dioxide nanoparticles in ethylene with constant wall temperature under different concentrations of
glycol. Present study concluded that size of nanoparticles and nanoparticles. They obtained augmentation of heat transfer
viscosity of the nanofluids played a vital role in thermal conduc- coefficient of nanofluid with increase of nanoparticle concen-
tivity enhancement ratio of them. tration. They also obtained greater heat transfer coefficient of
Mintsa et al. [7] investigated the effect of temperature, particle nanofluid in comparison to that of distilled water basefluid at
size and volume fraction on thermal conductivity of water based a constant Peclet number. Authors have reported that the heat
nanofluids of copper oxide and alumina. Authors suggested that transfer augmentation results are much higher in experimental
thermal characteristics can be enhanced with increase of particles’ observation than that of predicted results. Yu et al. [11] con-
volume fraction, temperature and particle size. Authors found that ducted heat transfer experiments of nanofluids containing
the smaller the particle size, the greater the effective thermal 170-nm silicon carbide particles at 3.7% volume concentration.
conductivity of the nanofluids at the same volume fraction. Contact The results showed that heat transfer coefficients of nanofluids
surface area of particles with fluid and Brownian motion can be are 50e60% greater than those of basefluids at a constant Rey-
increased when smaller particles are used in the same volume nolds number.
fraction. This consequently increased thermal conductivity of Kim et al. [12] investigated effect of nanofluids on the perfor-
nanofluids. mances of convective heat transfer coefficient of a circular straight
tube having laminar and turbulent flow with constant heat flux.
3. Nanofluids in enhancing forced convective heat transfer Authors have found that the convective heat transfer coefficient of
and relevant pressure drop alumina nanofluids improved in comparison to basefluid by 15%
and 20% in laminar and turbulent flow, respectively. This showed
Namburu et al. [8] numerically analyzed turbulent flow and heat that the thermal boundary layer played a dominant role in laminar
transfer to three types of nanofluids namely copper oxide (CuO), flow while thermal conductivity played a dominant role in turbu-
alumina (Al2O3) and silicon dioxide (SiO2) in ethylene glycol and lent flow. However, no improvement in convection heat transfer
water, flowing through a circular tube under constant heat flux. coefficient was noticed for amorphous particle nanofluids.
Results revealed that nanofluids containing smaller diameter of Ku et al. [13] conducted an experimental study on pressure drop
nanoparticles produce higher viscosity and Nusselt number. Nus- of nanofluids containing carbon nanotubes in a horizontal tube.
selt numbers are also increased at higher volume fraction of Authors had reported that the pressure drops of nanofluids became
particles. It is observed that at a constant heat flux (50 W/cm2) with almost the same as that of distilled water (basefluid). The viscosity
a constant Reynolds number (20,000), heat transfer coefficient of of nanofluids decreases with the shear rate and thus, the difference
6% CuO nanofluid has increased 1.35 times than that of the base- of pressure drop between nanofluids and distilled water at elevated
fluid. At the same particle volume fraction, CuO nanofluid produced flow rate also decreases.
 K.Y. Leong et al. / Applied Thermal Engineering 30 (2010) 2685e2692 2687

Duangthongsuk and Wongwises [14] conducted an experi- Table 2

mental study on the heat transfer performance and pressure drop Surface characteristics of a compact heat exchangereradiator [17].

of TiO2ewater nanofluids, flowing in a turbulent flow regime. They Serial Description Air side Coolant
reported that the pressure drop of nanofluids was slightly higher number side
than that of the basefluid and increased with the increasing volume 1 Tube arrangement Staggered
concentrations. 2 Fin type Ruffled
3 Fin pitch 4.46 fins/cm
4 Fin thickness 0.01 cm
5 Hydraulic diameter, Dh 0.351 cm 0.373 cm
4. Input data and operating characteristics 6 Free flow area frontal area, s 0.78 0.129
7 Heat transfer area/total volume, a 886 m2/m3 138 m2/m3
Necessary input data were taken from literature to perform the 8 Fin area/total area, b 0.845

analysis in this study. The characteristics of the radiator considered

in this study are obtained from Vasu et al. [15] and Kays and London
[16] as shown in Tables 1 and 2. The radiator with given specifi- relevant data are shown in Tables 1e3 and Fig. 1. Calculations were
cations is mounted on the turbo-charged diesel engine of type TBD done on air and coolant sides.
232V-12. It is a cross-flow compact heat exchanger with unmixed
air and fluid as coolants. This compact heat exchanger consists of 5.1. Air side calculation
644 brass flat tubes and 346 continuous copper fins. [15]. Compact
heat exchanger is a unique and special class of heat exchanger Initially, air side calculations were performed to determine air
having a large heat transfer area per unit volume. In addition, flat heat capacity, air heat transfer coefficient, fin efficiency and total
tubes are more popular in automotive applications due to the lower surface temperature effectiveness. These data were needed to
drag profile compared to round tubes [15]. These two features will calculate heat exchanger effectiveness, NTU number and overall
enhance the cooling rate and at the same time minimizing the flow heat transfer coefficient for the nanofluids’ side calculation. Air
resistance [15]. Other important thermal properties of nanofluids properties are based on temperature at 300 K. In addition, relevant
are calculated from empirical correlations shown in Section 5. input data from previous section were substituted into the math-
Thermo-physical properties of ethylene glycol are presented in ematical formulations. The mathematical formulations are shown
Table 3 and the thermal conductivities of nanofluids used in this below.
analysis are obtained from Eastman et al. [3] as shown in Fig. 1. The
calculated thermo-physical properties of nanofluids were assumed (a) Air heat capacity rate, Ca can be expressed as Eq. (1):
constant and their stability also remained invariable after their
application in an automotive car radiator. The thermal conductivity Ca ¼ Wa Cp;a (1)
obtained from Eastman et al. [3] is considered as benchmark value.
where,
Higher value could be obtained in real application because the
coolant of an automotive radiator is usually operated at a higher
Cp,a ¼ 1007 J/kg K at 300 K.
temperature. This condition will enhance the Brownian motion of
the nanoparticles and subsequently increase thermal conductivity
of the nanofluids. The density, viscosity and heat capacity of Wa ¼ GAfr sa (2)
nanofluids were assumed to generate no substantial effects on
those properties of basefluid because only a maximum of 2% (b) Heat transfer coefficient, ha can be expressed as Eq. (3):
volume fraction of nanofluids were considered in this study.
ja Ga Cp;a
ha ¼ 2=3
(3)
Pra
5. Mathematical formulation of ethylene glycol based copper
nanofluids in an automotive car radiator where,

Mathematical correlations shown in Section 5.1 are taken from 0:174

ja ¼ (4)
references [15e17,19]. In this paper a comparison is made between Re0:383
a
the heat transfer performance of radiator by operating with
ethylene glycol and nanofluid coolants. It highlighted not only the W
G ¼ (5)
influence of nanofluids but also volume fraction of copper nano- Afr sa
particles to the heat transfer rate of a radiator. Described equations
are being incorporated to aid the comparison. Thermal perfor-
Ga Dh;a
mances of a radiator can be calculated using Eqs. (1)e(27), the Rea ¼ (6)
ma
where, ma ¼ 0.00001846 Ns/m2 at 300 K.
Table 1
Core geometry of flat tubes, continuous fins and operating conditions of a compact
heat exchangereradiator [15].

Serial Description Air Coolant Table 3

number Thermal physical properties of ethylene glycol [18].
1 Fluid inlet 20e55  C 70e95  C
Thermal physical property Ethylene glycol (at 82.5  C)
temperature (assume Ta ¼ 37.5  C) (assume Ta ¼ 82.5  C)
2 Core width 0.6 m Density (kg/m3) 1076
3 Core height 0.5 m Specific heat (J/kgK) 2664
4 Core depth 0.4 m Viscosity (Ns/m2) 0.003036
5 Tube size 1.872 cm  0.245 cm Conductivity (W/m K) 0.261
2688 K.Y. Leong et al. / Applied Thermal Engineering 30 (2010) 2685e2692

1
mnf ¼ mf (13)
ð1  4Þ2:5
Wnf
Gnf ¼ (14)
Afr snf

Wnf
vnf ¼ (15)
rnf

mnf Cp;nf
Prnf ¼ (16)
knf
 
1  4 rf Cp;f þ 4rp Cp;p
Cp;nf ¼ (17)
rnf

rnf ¼ ð1  4Þrf þ 4rp (18)

Fig. 1. Thermal conductivity of ethylene glycol based copper nanofluids.

(b) Heat capacity rate, Cnf can be expressed as Eq. (19):

(c) Fin efficiency of plate fin, hf can be expressed as Eq. (7): Cnf ¼ Wnf Cp;nf (19)

tanh mL
h¼ (7) (c) Heat exchanger effectiveness for cross-flow unmixed fluid,
mL 3 can be expressed as Eq. (20):
where,
  h h i i
rffiffiffiffiffiffiffiffi 1
2ha 3 ¼ 1  exp *
ðNTUÞ0:22 exp  C * ðNTUÞ0:78  1 (20)
m ¼ ; (8) C
kt
where,
k ¼ 3994 W/m K is the thermal conductivity of copper fin.
Cminimun
(d) Total surface temperature effectiveness, can be expressed as C* ¼ (21)
Cmaximum
Eq. (9):
  Af Ua Afr;a
ho ¼ 1:0  1:0  hf  (9) NTU ¼
Ca
(22)
A
(d) Overall heat transfer coefficient, based on air side can be
5.2. Nanofluid side calculation expressed as Eq. (23), where wall resistance and fouling factors
are neglected.
The parameters needed for nanofluid side calculation are
nanofluid heat transfer coefficient, nanofluid heat capacity rate, 1 1 1
heat exchanger effectiveness for cross-flow unmixed fluid, heat ¼ þ  (23)
Ua ho ha anf
transfer coefficient based on air side, pressure drop, pumping hnf
aa
power and total heat transfer rate. Similar to previous section, the
relevant input data are substituted into the mathematical formu- (e) Pressure drop can be expressed as Eq. (24):
lations as shown below.
G2nf  fnf  H
(a) Heat transfer coefficient can be expressed as Eq. (10): DPnf ¼   (24)
D
2  rnf  h;nf 4
Nunf knf
hnf ¼ (10)
Dh;nf where,

knf is obtained from Eastman et al. [3] where,  0:25

fnf ¼ 0:079  Renf (25)
Nunf ¼ 0:023Re0:8 0:3
nf Prnf (11)
(f) Pumping power can be expressed as Eq. (26):
Gnf Dh;nf
Renf ¼ (12) P ¼ Vnf  DP
mnf (26)

mnf, Cp,nf and rnf were calculated based on correlations obtained

from Tsai and Chein [19]. Authors in their study also used these (g) Total heat transfer rate can be expressed as Eq. (27):
correlations to analyze the performance of nanofluid-cooled
 
microchannel heat sink. Similar correlations for Cp,nf and rnf were
Q ¼ 3Cmin Tnf;in  Ta;in (27)
used by other researchers [20].
 K.Y. Leong et al. / Applied Thermal Engineering 30 (2010) 2685e2692 2689

5.3. Test conditions

The analysis was performed based on a radiator specifications and

conditions of fluids shown in Tables 1 and 2. However, nanoparticles
volume fraction, air and coolant Reynolds number were varied in
order to determine the thermal performance of the radiator using
nanofluids. The test conditions for each analysis are explained below.

(a) Influence of volume fraction of copper nanoparticles on the

thermal performance of an automotive car radiator: In this
analysis, air and coolant Reynolds number were kept fixed at
4000 and 5000, respectively. However, concentrations of
copper nanoparticles were increased from 0% to 0.2%. Coolant
mass and volume flow rate, coolant Prandtl and Nusselt
number, overall heat transfer coefficient based on air side and
total heat transfer rate of the radiator were then determined.
(b) Influence of air Reynolds number on the thermal performance
of a radiator: Air Reynolds number was varied from 4000 to
6000 while coolant Reynolds number was kept fixed at 5000. Fig. 2. Effect of copper volume fraction to coolant mass and volumetric flow rate at
The analysis also included a comparison of the thermal constant air and coolant Reynolds number.
performance of a radiator with nanofluids at different volume
fractions. This part focused on the air heat transfer coefficient,
overall heat transfer coefficient based on air side and total heat However, volumetric flow rate of nanofluids is decreased with
transfer rate of an automotive radiator. increase of volume fraction of copper nanoparticles as calculated
(c) Influence of coolant Reynolds number on the thermal perfor- using Eq. (15) and presented in Fig. 2. In this study, it is found that
mance of a radiator: Coolant Reynolds number was varied from Prandtl number of nanofluids based coolant decreases exponentially
5000 to 7000 while air Reynolds number was kept fixed at 4000. with volume fraction of copper nanoparticles, mainly due to higher
The analysis also included a comparison of thermal performance thermal conductivity of nanofluids. Nanofluids exhibit higher
of the radiator with nanofluids at different volume fractions. This thermal conductivity due to increase in Brownian motion, formation
analysis focused on the overall heat transfer coefficient based on of nanolayer etc. Prandtl number of nanofluids is calculated using Eq.
air side and total heat transfer rate of an automotive radiator. (16) and the knf is an influential parameter for Prandtl number. It
(d) Comparison of coolant pressure drop and pumping power: In may be stated that higher value of this property will lead to lower
this section, coolant flow rate was kept fixed at 0.2 m3/s and air value of Prandtl number. These relationships are shown in Fig. 3.
Reynolds number at 4000 but the volume fraction of copper Fig. 3 also showed lower coolant Nusselt number with the addition
nanoparticles was varied. It focused on the effects of volume of copper particles, based on Eq. (11). In this equation, the coolant
fraction of copper nanoparticles on the coolant pressure drop Reynolds number is kept constant while the coolant Prandtl number
and pumping power. is varied. This study also found that ethylene glycol based copper
nanofluids demonstrated higher overall heat transfer coefficient for
6. Results and discussions air side as calculated using Eq. (23). The relationship is shown in
Fig. 4 where overall heat transfer coefficient for air side is increased
6.1. Influence of volume fraction of copper particles to thermal with copper nanoparticles. For instance, an overall heat transfer
performance of an automotive car radiator coefficient, 164 W/m2 K can be achieved for 2% Cu þ EG nanofluid
compared to 142 W/m2 K for the basefluid. This indicates lower air
In this section, analysis of thermal performance of an auto- side area is required to achieve 142 W/m2 K overall heat transfer
motive car radiator at constant Reynolds number of air (4000) and
coolant (5000) have been carried out. With the increase of volume
fraction of copper nanoparticles, the dynamic viscosity of nano-
fluids has been increased. Dynamic viscosity of nanofluids in this
study was calculated using the correlation developed by Tsai and
Chein [15] as shown in Eq. (13). This parameter influences mass
flow rate of nanofluids in an automotive car radiator. From Eq.
(12), it is found that the mass velocity of nanofluids increased with
the increase in volume fraction of copper nanoparticles due to
higher dynamic viscosity of the nanofluids. It can also be
explained from Eq. (28).

Renf  mnf
Gnf ¼ (28)
Dh;nf

In this equation mnf was varied and other parameters were kept
constant. This value is then substituted into Eq. (14) to calculate the
coolant mass flow rate, which can be manipulated as shown in
Eq. (29).
Fig. 3. Effect of copper volume fraction to coolant Prandtl and Nusselt number at
Wnf ¼ Gnf  Afr  snf (29) constant air and coolant Reynolds number.
2690 K.Y. Leong et al. / Applied Thermal Engineering 30 (2010) 2685e2692

170
Overall Heat Transfer Coefficient based

165
on Air Side (W/m2K)

160

155

150

145

140
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 Fig. 6. Effect of air Reynolds number to air heat transfer coefficient.
Volume Fraction of Copper Particles (%)

Fig. 4. Effect of copper volume fraction to overall heat transfer coefficient based on air
air side of the radiator is obtained and shown in Fig. 7. From the
side at constant air and coolant Reynolds number.
study, air heat transfer coefficient is found proportional with air
Reynolds number. It can be explained using Eqs. (3) and (6). Eq. (6)
coefficient if 2% Cu þ EG nanofluid is used. Hence, estimated can be manipulated as shown the Eq. (30):
reduction of air side area up to 15.31% at 2% Cu þ EG particles was
Rea ma
achieved. This study also found that heat transfer rate is increased Ga ¼ (30)
exponentially as the volume fraction of copper particles are
Dh;a
increased as shown in Fig. 5. This improvement is calculated using In Eq. (30), the hydraulic diameter and air viscosity were kept
Eq. (27). It can be deduced that effectiveness of the radiator is constant. Therefore, air mass velocity is increased when air Rey-
increased with the application of nanofluids. However the nolds number increases. The higher value of air mass velocity is
percentage of effectiveness does not increase substantially, although then substituted into Eq. (3) to determine the air heat transfer
the improvement of overall heat transfer coefficient is significant. coefficient. In Eq. (3) the air heat capacity and Prandtl number were
kept constant. Although the Colburn factor is decreased with air
Reynolds number, the increment of air mass velocity is much
6.2. Influence of air Reynolds number on thermal performance of higher and subsequently air heat transfer coefficient is increased.
a radiator Nanofluids with higher copper volume fraction generates higher
overall heat transfer coefficient than that of a basefluid. Same
The effect of air Reynolds number on the thermal performance scenario happened for heat transfer rate where it is proportional to
of a radiator is discussed in this section. A coolant’s volumetric and air Reynolds number as shown in Fig. 8. About 3.8% of heat transfer
mass flow rates, Nusselt and Prandtl numbers did not experience improvement can be achieved with addition of 2% copper particles
any change since the coolant Reynolds number was kept fixed at at 6000 and 5000 Reynolds number for air and coolant respectively.
5000. Only air Reynolds number and fraction of copper nano- Based on the overall heat transfer coefficient and heat transfer rate
particles were varied in this section. With the increase of air Rey- improvement, percentage reduction of air frontal area can be
nolds numbers, the air heat transfer coefficient was increased as estimated, at these Reynolds numbers. Percentage improvement of
shown in Fig. 6. This will influence the thermal performance of the heat transfer rate is increased with air Reynolds number. Although
radiator system. Substituting higher value of air heat transfer
coefficient into Eq. (23), overall heat transfer coefficient based on

222

221.5

221
Heat Transfer (kW)

220.5

220

219.5

219

218.5

218

217.5

217

216.5
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2
Volume Fraction of Copper Particles (%)

Fig. 5. Effect of copper volume fraction to heat transfer rate at constant air and coolant Fig. 7. Effect of air Reynolds number and copper volume fraction to overall heat
Reynolds number. transfer coefficient based on air side.
 K.Y. Leong et al. / Applied Thermal Engineering 30 (2010) 2685e2692 2691

320

310

300

290
Heat Transfer (KW)

280

270

260

250

240

230

220

210

200
3500 4000 4500 5000 5500 6000 6500
Air Reynolds Number

EG Nanofluid (2.0% Copper Particles)

Fig. 8. Effect of air Reynolds number and copper volume fraction to heat transfer rate Fig. 10. Effect of coolant Reynolds number to heat transfer rate of radiator.
of radiator.

thermostat is also playing an important role to control the coolant

results indicated that higher air Reynolds number leads to better
Reynolds number. From this study, it indicates that as the coolant
heat dissipation process, design of the radiator must ensure the
Reynolds numbers are increased, the mass and volumetric flow
engine operation at optimum temperature. Driving conditions or its
rates are also increased. Overall heat transfer coefficient based on
speed and engine load must be considered. For instance, car’s
air side is increased with coolant Reynolds number as shown in
engine needs to operate at higher load when driving up hill and at
Fig. 9. The magnitude of this property for nanofluids is higher than
the same time air Reynolds number is low due to lower air velocity.
that of a basefluid. Therefore, heat transfer area reduction for the
Hence, there is possibility for engine to get overheated at this
same value of overall heat transfer coefficient can be achieved by
condition. However when driving downhill, an engine only
using nanofluids. Heat transfer enhancement was also observed
requires operating at lower load and at the same time high air
with coolant Reynolds number. For instance, with the addition of
Reynolds number is observed. Eventually engine might be over-
2% copper particles, 1.4% improvement of heat transfer rate has
cooled. Therefore, these aspects must take into consideration when
been achieved at 4000 and 7000 Reynolds number for air and
designing automotive radiator.
coolant respectively. It is also observed that the percentage of
improvement is decreased with decrease of coolant Reynolds
6.3. Influence of coolant Reynolds number on thermal performance number. Fig. 10 shows heat transfer rate of a radiator using nano-
of a radiator fluid is higher than that of a radiator using ethylene glycol.

This section presents the effect of coolant Reynolds number on

the thermal performance of a radiator at a fixed air Reynolds 6.4. Comparison of coolant pressure drop
number (4000). Coolant Reynolds number plays vital role in
determining the radiator’s thermal performance. Engine might be In this section analyses of the coolant pressure drop and
overcooled or overheated if coolant Reynolds number is not pumping power of a radiator at a fixed coolant flow rate (0.2 m3/s)
properly controlled. The main function of a radiator is to ensure
that engine is operating at optimum temperature by not only
controlling the air Reynolds number but also the coolant Reynolds
number. Although the coolant pump is usually driven by an engine,

Fig. 9. Effect of coolant Reynolds number to overall heat transfer coefficient based on Fig. 11. Influence of copper volume fraction to coolant pressure drop at fixed coolant
air side. volumetric flow rate.
2692 K.Y. Leong et al. / Applied Thermal Engineering 30 (2010) 2685e2692

0.4% heat transfer enhancement were observed for pure

ethylene glycol and ethylene glycol with 2% copper nano-
particles respectively when coolant Reynolds number was
increased from 5000 to 7000.
(c) Estimated 18.7% reduction of air frontal area is achieved by
adding 2% copper nanoparticles at Reynolds number of 6000
and 5000 for air and coolant respectively.
(d) Additional 12.13% pumping power is needed for a radiator
using nanofluid of 2% copper particles at 0.2 m3/s coolant
volumetric flow rate compared to that of the same radiator
using only pure ethylene glycol coolant.

References

Fig. 12. Influence of copper volume fraction to pumping power at fixed coolant [1] D.P. Kulkarni, R.S. Vajjha, D.K. Das, D. Oliva, Application of aluminum oxide
volumetric flow rate. nanofluids in diesel electric generator as jacket water coolant, Applied
Thermal Engineering 28 (14e15) (2008) 1774e1781.
[2] W. Yu, D.M. France, S.U.S. Choi, J.L. Routbort, Review and Assessment of
Nanofluid Technology for Transportation and Other Applications (No. ANL/
and an air Reynolds number of (4000) with varying copper nano-
ESD/07-9). Energy System Division, Argonne National Laboratory, Argonne,
particles volume fraction is incorporated. Initially, the coolant 2007.
volumetric flow rate was converted to mass flow rate by using Eq. [3] J.A. Eastman, S.U.S. Choi, S. Li, W. Yu, L.J. Thompson, Anomalously increased
(15). This equation is manipulated as shown in Eq. (31) to calculate effective thermal conductivities of ethylene glycol-based nanofluids contain-
ing copper nanoparticles, Applied Physics Letters 78 (6) (2001) 718e720.
the coolant mass flow rate. [4] Y. Hwang, J.K. Lee, C.H. Lee, Y.M. Jung, S.I. Cheong, C.G. Lee, Stability and
thermal conductivity characteristics of nanofluids, Thermochimica Acta 455
Wnf ¼ vnf  rnf (31) (1e2) (2007) 70e74.
[5] J.-H. Lee, K.S. Hwang, S.P. Jang, B.H. Lee, J.H. Kim, S.U.S. Choi, Effective
Coolant mass flow rate is found to be increased with the addi- viscosities and thermal conductivities of aqueous nanofluids containing low
tion of copper nanoparticles due to its higher density. Then, coolant volume concentrations of Al2O3 nanoparticles, International Journal of Heat
and Mass Transfer 51 (11e12) (2008) 2651e2656.
mass velocity is calculated using Eq. (14). This value is substituted [6] W. Yu, H. Xie, L. Chen, Y. Li, Investigation of thermal conductivity and viscosity
into Eq. (12) to get the coolant Reynolds number. The pressure drop of ethylene glycol based ZnO nanofluid, Thermochimica Acta 491 (1e2) (2009)
has been calculated by using fanning friction factor based on Eq. 92e96.
[7] H.A. Mintsa, G. Roy, C.T. Nguyen, D. Doucet, New temperature dependent
(25), which then substituted into Eq. (24). thermal conductivity data for water-based nanofluids, International Journal of
It was observed that the coolant pressure drop increased with Thermal Sciences 48 (2) (2009) 363e371.
the addition of copper nanoparticles. The result reveals that [8] P.K. Namburu, D.K. Das, K.M. Tanguturi, R.S. Vajjha, Numerical study of
turbulent flow and heat transfer characteristics of nanofluids considering
a pressure drop of 110.97 kPa was obtained by adding 2% copper variable properties, International Journal of Thermal Sciences 48 (2) (2009)
particles compared to a pressure drop of 98.93 kPa for a basefluid. 290e302.
Due to this extra pressure drop, a higher coolant pumping power is [9] Y. Ding, H. Chen, Y. He, A. Lapkin, M. Yeganeh, L. Siller, Forced convective heat
transfer of nanofluids, Advanced Powder Technology 18 (6) (2007) 813e824.
needed. The pumping power is calculated using Eq. (26). Calculated [10] H.S. Zeinali, M. Nasr Esfahany, S.G. Etemad, Experimental investigation of
results indicate that about 12.13% increase in pumping power is convective heat transfer of Al2O3/water nanofluid in circular tube, Interna-
observed at 2% addition of copper nanofluids compared to a base- tional Journal of Heat and Fluid Flow 28 (2) (2007) 203e210.
[11] W. Yu, D.M. France, D.S. Smith, D. Singh, E.V. Timofeeva, J.L. Routbort, Heat
fluid. These trends are shown in Figs. 11 and 12. Increase in density
transfer to a silicon carbide/water nanofluid, International Journal of Heat and
increases pressure drop of flowing liquids. Adding particles in Mass Transfer 52 (15e16) (2009) 3606e3612.
a base liquid increases density of the fluid and augments pressure [12] D. Kim, Y. Kwon, Y. Cho, C. Li, S. Cheong, Y. Hwang, Convective heat transfer
drop at a low percentage as observed in the present study. Similar characteristics of nanofluids under laminar and turbulent flow conditions,
Current Applied Physics 9 (2) (2009) e119ee123.
results were reported by Ko et al. [13] and Duangthongsuk and [13] G.H. Ko, K. Heo, K. Lee, D.S. Kim, C.K.Y. Sohn, M. Choi, An experimental study
Wongwises [14]. on the pressure drop of nanofluids containing carbon nanotubes in a hori-
zontal tube, International Journal of Heat and Mass Transfer 50 (2007)
4749e4753.
7. Conclusion [14] W. Duangthongsuk, S. Wongwises, An experimental study on the heat
transfer performance and pressure drop of TiO2ewater nanofluids flowing
Following conclusions can be drawn from this study: under a turbulent flow regime, International Journal of Heat and Mass
Transfer 53 (2010) 334e344.
[15] V. Vasu, K.R. Krishna, A.C.S. Kumar, Thermal design analysis of compact heat
(a) Heat transfer rate is increased with increase in volume exchanger using nanofluids, International Journal of Nanomanufacturing 2 (3)
concentration of nanoparticles (ranging from 0% to 2%). About (2008) 271e287.
[16] W.M. Kays, A.L. London, Compact Heat Exchanger, third ed. McGraw-Hill, Inc.,
3.8% heat transfer enhancement was achieved with addition of
United States, 1984.
2% copper particles at 6000 and 5000 Reynolds number for air [17] D.G. Charyulu, G. Singh, J.K. Sharma, Performance evaluation of a radiator in
and coolant respectively. a diesel engine e a case study, Applied Thermal Engineering 19 (1999)
625e639.
(b) Thermal performance of a radiator using nanofluid or ethylene
[18] F.P. Incopera, D.P. Dewitt, T.L. Bergman, A.S. Lavine, Fundamentals of Heat and
glycol coolant is increased with air and coolant Reynolds Mass Transfer, sixth ed. John Wiley & Sons, New York, 2007.
number.About 42.7% and 45.2% heat transfer enhancement [19] T.H. Tsai, R. Chein, Performance analysis of nanofluid-cooled microchannel
were observed for pure ethylene glycol and ethylene glycol heat sinks, International Journal of Heat and Fluid Flow 28 (2007) 1013e1026.
[20] R. Strandberg, D.K. Das, Finned tube performance evaluation with nanofluids
with 2% of copper nanoparticles respectively when air Rey- and conventional heat transfer fluids, International Journal of Thermal Science
nolds number was increased from 4000 to 6000. Only 0.9% and 49 (3) (2010) 580e588.