International Communications in Heat and Mass Transfer 38 (2011) 1283–1290

Contents lists available at ScienceDirect

International Communications in Heat and Mass Transfer

j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / i c h m t

Experimental study of heat transfer enhancement using water/ethylene glycol based

nanofluids as a new coolant for car radiators☆
S.M. Peyghambarzadeh a, S.H. Hashemabadi b,⁎, S.M. Hoseini a, M. Seifi Jamnani a
a
Department of Chemical Engineering, Mahshahr branch, Islamic Azad University, Mahshahr, Iran
b
CFD Research Laboratory, School of Chemical Engineering, Iran University of Science and Technology, Narmak, Tehran, 16846, Iran

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

Available online 16 July 2011 Traditionally forced convection heat transfer in a car radiator is performed to cool circulating fluid which
consisted of water or a mixture of water and anti-freezing materials like ethylene glycol (EG). In this paper,
Keywords: the heat transfer performance of pure water and pure EG has been compared with their binary mixtures.
Water/Al2O3 nanofluid Furthermore, different amounts of Al2O3 nanoparticle have been added into these base fluids and its effects on
Ethylene glycol/Al2O3 nanofluid the heat transfer performance of the car radiator have been determined experimentally. Liquid flow rate has
Heat transfer enhancement
been changed in the range of 2–6 l per minute and the fluid inlet temperature has been changed for all the
Car radiator
Cooling performance
experiments. The results demonstrate that nanofluids clearly enhance heat transfer compared to their own
Experimental study base fluid. In the best conditions, the heat transfer enhancement of about 40% compared to the base fluids has
been recorded.
© 2011 Elsevier Ltd. All rights reserved.

1. Introduction mixture is depressed significantly. The minimum freezing point is

observed when the EG percent in water is about 70%, as shown in Fig. 1.
After the publication of our previous paper [1] about the However, the boiling point for aqueous EG increases monotonically
application of water/Al2O3 nanofluids instead of pure water in the with increasing EG percentage. Thus, the use of EG not only declines
car radiator and recording the interesting heat transfer enhancement the freezing point but also elevates the boiling point such that the
of about 45%, we want to investigate the application of nanoparticle in operating range for the heat transfer fluid is broadened on both ends of
the mixture of water and anti-freeze materials (as the base fluid) the temperature scale [2].
which is conventionally used in the cars' radiators. It is common in the It has been proved that conventional fluids, such as water and EG
area of cold or hot weathers that some additives are added to the have poor convective heat transfer performance and therefore high
water in the automotive radiator which decrease freezing point and compactness and effectiveness of heat transfer systems are necessary
elevate boiling point of water. It keeps the radiator fluid from freezing to achieve the required heat transfer. Among the efforts for
when it is very cold and keeps the car from overheating on very hot enhancement of heat transfer the application of nanoparticle additives
days. Almost all of these additives are from glycol family specially to liquids is more noticeable and currently a large number of
ethylene glycol (EG). The major use of EG is as a medium for investigations are devoted to this subject [3–8]. Nanofluids are formed
convective heat transfer in, for example car radiators, liquid cooled by suspending metallic or non-metallic oxide nanoparticles (that are
computers, chilled water air conditioning systems, and the like. significantly smaller than 100 nm) in traditional heat transfer fluids.
Because water is a much better engine coolant, the mixture of water These so-called nanofluids display good thermal properties compared
and EG has been used. The trouble with water is that it freezes or boils with fluids conventionally used for heat transfer and fluids containing
at extreme temperatures. Anti-freezing agents like EG can withstand particles on the micrometer scale. These fluids are a new window
much greater temperature extremes, so by adding it to water we can which has been opened recently and it was confirmed by several
make a compromise. Most of the good cooling abilities of water are authors that these working fluids can enhance heat transfer perfor-
retained but the ability to withstand extreme temperatures comes mance [9,10].
from the anti-freeze. As can be seen in Fig. 1, a mixture of 60% EG and In the car radiators, the coolant media is pumped through the flat
40% water does not freeze to temperatures below − 45 °C. EG disrupts tubes while the air is drawn over the fins by forced convection,
hydrogen bonding when dissolved in water. Pure EG freezes at about thereby heat exchanges between the hot circulating fluid and air. The
−12 °C, but when intermixed with water, the freezing point of the application of nanofluids in these finned tube radiators may result in
several potential benefits including increased heating output for equal
☆ Communicated by W.J. Minkowycz.
liquid flow. These performance impacts, in turn, may be translated
⁎ Corresponding author. into a reduction in total required heat transfer area. Superior heat
E-mail address: hashemabadi@iust.ac.ir (S.H. Hashemabadi). transfer properties of nanofluids may also result in lower liquid flow

0735-1933/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.icheatmasstransfer.2011.07.001
1284 S.M. Peyghambarzadeh et al. / International Communications in Heat and Mass Transfer 38 (2011) 1283–1290

Nomenclature

A Peripheral area (m 2)
Cp Specific heat capacity (J/kg K)
d Hydraulic diameter of the tube (m)
dp Diameter of the nanoparticles (m)
f Friction factor
h Heat transfer coefficient (W/m 2K)
k Thermal conductivity (W/m K)
L Length of the tube (m)
ṁ Mass flow rate (kg/s)
Nu Nusselt number = (hdhy/k)
P Tube periphery (m)
Pe Peclet number = Re.Pr
Pr Prandtl number = (Cpμ/k)
Q Heat transfer rate (W)
Re Reynolds number = (ρudhy/μ)
Fig. 1. Boiling and freezing points of water/EG mixtures [2].
S Cross sectional area of the tube (m 2)
T Temperature (K)
VB Brownian velocity
number (Re b 270). The maximum promotion of Nusselt number for
z Axial distance from the tube inlet (m)
1 vol.% nanofluid was 8%. Jung et al. [15] conducted convective heat
transfer experiments for a nanofluid (Al2O3–water) in a rectangular
micro-channel under laminar flow conditions. Their results show the
Greek letters
heat transfer coefficient increases by more than 32% for 1.8 vol.%
α Thermal diffusivity (m 2/s)
nanoparticle. Sharma et al. [16] implemented 1 to 2.5 vol.% Al2O3 in
ρ Density (kg/m 3)
water in horizontal tube geometry and concluded while the Peclet
δ Distance between the centers of the particles (m)
number is between 3500 and 6000, up to 41% promotion in heat
μ Viscosity (kg/m.s)
transfer coefficient compared to pure water may have occurred. Ho et
φ Nanoparticle volume fraction (%)
al. [17] conducted an experiment for cooling in horizontal tube in
Φ Shape factor
laminar flow of Al2O3–water at 1 and 2 vol.% concentrations and
ψ Particle sphericity
concluded the interesting enhancement of 51% in heat transfer
coefficient. Nguyen et al. [18] performed their experiments in a
microprocessor type cooling heat exchanger and at 6.8 vol.% Al2O3 in
Subscripts water obtained 40% growing in heat transfer coefficient. Xie et al. [19]
ave Average reported the convective heat transfer enhancement of nanofluids as
b Bulk coolants in laminar flows inside a circular copper tube with constant
bf Base fluid wall temperature. Different nanofluids consisting of Al2O3, ZnO, TiO2,
exp Experimental and MgO nanoparticles were prepared with a mixture of 55 vol.%
in Input distilled water and 45 vol.% EG as base fluid. MgO, Al2O3, and ZnO
nf Nanofluid nanofluids exhibited superior enhancements of heat transfer coeffi-
out Output cient, with the highest enhancement up to 252% at a Reynolds number
p Particle of 1000 for MgO nanofluid. The performance of finned tube heating
w Wall units with nanofluids has been compared mathematically with a
conventional heat transfer fluid which comprised of 60% EG and 40%
water by Strandberg and Das [20]. Their model predicted an 11.6%
rate for a given rate of heat transfer, yielding a reduction in the liquid increase in finned tube heating output under certain conditions with
pumping power consumed compared to the base fluid. In order to the 4% Al2O3/60% EG nanofluid and an 8.7% increase with the 4%
have more understanding about the application of nanofluids in CuO/60% EG nanofluid compared to heating output with the base
various heat exchangers, a brief literature survey is performed in this fluid. Application of EG based copper nanofluids in an automotive
paper. cooling system has been studied by Leong et al. [21]. Relevant input
Pak and Cho [11] presented an experimental investigation of the data, nanofluid properties and empirical correlations were obtained
convective turbulent heat transfer characteristics of nanofluids from literatures to investigate the heat transfer enhancement of an
(Al2O3–water) with 1 to 3 vol.%. Their results show that Nusselt automotive car radiator operated with nanofluid-based coolants. It is
number for the nanofluids enhances with increasing of volume observed that, about 3.8% of heat transfer enhancement could be
concentration and Reynolds number. Heris et al. [12] examined and achieved with the addition of 2% copper nanoparticles in a base fluid
proved the enhancement of in-tube laminar flow heat transfer of at the Reynolds number of 6000 and 5000 for air and coolant
nanofluids (water–Al2O3) in a constant wall temperature boundary respectively. Some extensive reviews in the nanofluid heat transfer
condition. In other work, Heris et al. [13] presented an investigation of have also been published by Godson et al. [22], Kakaç et al. [23] and
the laminar flow convective heat transfer of Al2O3–water under Wang et al. [24]. The interested reader can refer to them for complete
constant wall temperature with 0.2 to 2.5 vol.% of nanoparticle for reviewing of the previous studies performed.
Reynolds number varying between 700 and 2050. They presented It should be emphasized that almost no document can be found to
again the Nusselt number for the nanofluid which is greater than the describe experimental evaluation of nanofluid performance in the car
base fluid. Lai et al. [14] studied the flow behavior of nanofluids radiator. In this paper, experimental comparisons have been accom-
(20 nm Al2O3 nanoparticle in water) in a millimeter-sized stainless plished between the heat transfer performance of pure water and
steel test tube, subjected to constant wall heat flux and a low Reynolds pure EG and some concentrations of their mixtures in the car radiator.
 S.M. Peyghambarzadeh et al. / International Communications in Heat and Mass Transfer 38 (2011) 1283–1290 1285

How do anti-freeze materials like EG affect the heat transfer minute. A recycle line included a globe valve was prepared to obtain
performance of the radiator? What happens when you increase the the predetermined flow rate. Storage tank (height of 35 cm and
EG concentration? Furthermore, when small amounts of alumina diameter of 30 cm) has volume of 30 l and the working liquid would
nanoparticle are added to water or EG or their mixtures, does the rate fill 25%. Consequently, the total volume of the circulating liquid is
of heat transfer change compared with the base fluids? What will be constant in all the experiments. A flow meter (Technical Group LZM-
the effects of operating parameters like nanoparticle concentration, 15Z Type) was used to control and manipulate the flow rate with the
flow rate, and temperature of circulating fluid on the heat transfer precision of 0.1 l per minute.
performance? These are the main questions which have been For heating the working fluid, an electrical heater and a controller
answered along this paper. were used to maintain the temperature between 40 and 80 °C. Two
RTDs (Pt-100 Ω) were implemented on the flow line to record
2. Experimental radiator fluid inlet and outlet temperatures. Two other J-type
thermocouples were also used for radiator wall temperature
2.1. Experimental rig and procedure measurement. These thermocouples were installed at the center of
the radiator surfaces (both sides). Due to very small thickness and
In order to measure the liquid side heat transfer coefficients in the very high thermal conductivity of the flat tubes, it is reasonable to
car radiator, a flow loop shown in Fig. 2(A) has been used. This equate the inside temperature of the tube with the outside one. The
experimental rig includes a storage tank, a heater, a pump, a flow measured temperatures from these thermocouples and RTDs have
meter, a forced draft fan, a cross flow finned tube heat exchanger (car been shown on two digital monitors with the accuracy of 0.1 °C. All
radiator), and flow lines. The test fluid flows through the five layer used thermocouples and RTDs were thoroughly calibrated by using a
insulated tubes (0.75 inch diameter) from the feed tank to the constant temperature water bath, and their accuracy was estimated to
radiator by a centrifugal pump with constant flow rate of 10 l per be ±0.2 °C.

Fig. 2. A) Schematic of experimental set up. B) Schematic drawing of the applied louvered fin and flat tube radiator and their dimensions.
1286 S.M. Peyghambarzadeh et al. / International Communications in Heat and Mass Transfer 38 (2011) 1283–1290

Car radiator configuration is louvered fin and flat tube as shown in Table 2

Fig. 2(B). This air cooler includes 34 vertical aluminum tubes with Range of experimental operational conditions.

elliptical cross section. The distances between the tube rows have Parameters Water based nanofluids EG based nanofluids
been filled with thin perpendicular aluminum fins. For cooling the
Nanoparticle type γ-Al2O3 γ-Al2O3
liquid, an axial forced fan (Techno Pars 1400 rpm) was installed close Nanoparticle concentration 0 to 1 0 to 1
on axis line of the radiator and consequently air and water have (vol.%)
indirect cross flow contact. Flow rate (l/min) 2 to 5 3 to 6
Reynolds number 9000 to 23,000 1200 to 2500
Gamma alumina nanoparticles (0.1, 0.3, 0.5, 0.7, and 1 vol.%) have
Flow type Turbulent Laminar
been added to different base fluids including pure water, pure EG and Inlet temperature 35 to 50 °C 45 to 60 °C
their mixtures (5, 10, 20 vol.% EG). The mean grain size of this gamma
alumina is 20 nm and some of its other properties are shown in
Table 1. There was no dispersant or stabilizer added to the nanofluid.
sphere with volume equal to that of the particle, to the surface area of
This is due to the fact that the addition of any agent may change the
the particle, and in this paper Φ is considered to be 3.
fluid properties [11] and the authors were interested to simulate the
For calculation of water based nanofluid viscosities, the following
easiest actual condition encountered in the car radiator. Additionally,
correlation has been applied [26–28]:
creating highly turbulent flow condition in the radiator tubes and
connecting pipes can improve the stabilization of the nanoparticle in

2
water. Table 2 shows the ranges of all the experimental parameters. μ nf = μ bf 123φ + 7:3φ + 1 ð4Þ

2.2. Uncertainty analysis

For EG based and mixture based nanofluids, the correlation
Uncertainty analysis is carried out by calculating the error of the proposed by Masoumi et al. [29] has been used:
measurements. The uncertainty range of Reynolds number comes
from the errors in the measurement of volume flow rate and hydraulic ρp VB d2p
diameter of the tubes and the uncertainty of Nusselt number refers to μ nf = μ bf + ð5Þ
72Cδ
the errors in the measurements of volume flow rate, hydraulic
diameter, and all the temperatures. According to uncertainty analysis
described by Moffat [25], the measurement errors of the main Where the second term is the apparent viscosity arising from the
parameters are summarized in Table 3. Furthermore, to check the effects of nanoparticles in the fluid. The distance between the centers
reproducibility of the experiments, some runs were repeated later of the nanoparticles, δ and correction factor (C) are calculated
which proved to be excellent. respectively:

rffiffiffiffiffiffiffi
3. Estimation of nanofluid physical properties π
δ= 3
d ð6Þ
6φ p
By assuming that the nanoparticles are well dispersed within the
base fluid, i.e. the particle concentration can be considered uniform −1
C = μ bf ðaφ + bÞ ð7Þ
throughout the system; the effective thermophysical properties of the
mixtures can be evaluated using some classical formulas as usually
used for two phase flow. The following correlations have been used to Where a and b are experimental parameters, which for the engine
predict nanofluid density, specific heat, and thermal conductivity coolant–Al2O 3 nanofluids were estimated to be 0.00004 and
respectively at different temperatures and concentrations [26–28]: 7.1274×10 − 7, respectively [30]. Various correlations were proposed
for temperature dependence of the nanofluid viscosity. Kole and Dey
ρnf = φρp + ð1−φÞρ bf ð1Þ [30] showed only that the following correlation proposed by Namburu
et al. [31] can give an acceptable agreement to the temperature




ρCp = φ ρCp + ð1−φÞ ρCp ð2Þ dependence of viscosity of the Al2O3/conventional coolant nanofluids.
nf p bf



 log μ nf = M expð NT Þ ð8Þ
kp + ðΦ−1Þkbf −φðΦ−1Þ kbf −kp
knf =
 kbf ð3Þ
kp + ðΦ−1Þkbf + φ kbf −kp
Where two parameters (M and N) are functions of nanoparticle
concentration [30]. Table 4 depicts pure water and pure EG physical
Where Φ is empirical shape factor given by Φ = 3/ψ, and ψ is the properties. As can be seen in Table 4, these two base fluids show
particle sphericity, that is defined as the ratio of the surface area of a significant differences in the physical properties. The mentioned
water/EG mixture properties in Table 4 were used by Dai et al. [35] to
develop computational models for the calculations of water/EG
Table 1 physical properties. Their findings were also used whenever no
Some characteristics of alumina nanoparticle.
experimental data existed.
Specification Value

Appearance White powder

Purity + 99% Table 3
Grain size (nm) 20 The uncertainty of the measured parameters.
Specific surface area (m2/g) 200
Parameter Value Uncertainty
Silicon (Si) (ppm) 3.5
Calcium (Ca) (ppm) 1.6 d (mm) 6.53 1.6%
Iron (Fe) (ppm) 0.2 Re 1200 to 23,000 5.2%
Cobalt (Co) (ppm) 0.8 Nu 24 to 120 18%
 S.M. Peyghambarzadeh et al. / International Communications in Heat and Mass Transfer 38 (2011) 1283–1290 1287

Table 4
Physical properties of pure water, pure EG and mixture of water–EG at 40 °C.

Physical Water EG Water 95% Water 90% Water 80%

properties
EG 5% EG 10% EG 20%

ρ (kg/m3) 992 1101 995a 1002 1008

μ (kg/m.s) 0.00065 0.0095 0.00101b 0.00165 0.0019
k (W/m °C) 0.633 0.256 0.615b 0.6 0.58
Cp (J/kg.°C) 4174 2382 4157c 4090 4020
α (m2/s) 1.5 × 10− 7 9.8 × 10− 8 1.49 × 10− 7 1.46 × 10− 7 1.43 × 10− 7
Pr 4.3 93 6.8 11.2 13.1
a
The density of water/EG mixtures have been obtained from [32].
b
The experimental thermal conductivities and viscosities of water/EG mixtures have
been obtained from [33].
c
The heat capacities of water/EG mixtures have been extracted from [34].

4. Calculation of heat transfer coefficient

The heat transfer coefficient and corresponding Nusselt number

can be derived as follows [1]:

h:d ṁCp ðTin −Tout Þ

Nu = = ð9Þ
k AðTb −Tw Þ

where ṁ is mass flow rate which is the product of density and volume
flow rate of the fluid, Tb is bulk temperature which is assumed to be
the average values of inlet and outlet temperatures of the fluid
moving through the radiator, and Tw is tube wall temperature which
is the mean value measured by two surface thermocouples. In Eq. (9),
k is fluid thermal conductivity and d is hydraulic diameter of the flat
tube. It should also be mentioned that all the physical properties were
calculated at the fluid bulk temperature. Fig. 3. The results for: A) pure water (Tin = 50 °C) in comparison with Dittus–Boelter
correlation [36]. B) Pure EG (Tin = 40 °C) in comparison with the Vajjha et al. correlations
5. Results and discussions [37].

5.1. Heat transfer to pure water and pure EG

with these empirical correlations over the Reynolds number range used
Before running the experiments on the nanofluids as a coolant for car in this study. The experimental data in different water temperatures at
radiator, some tests with pure water and pure EG were done in order to the radiator inlet including 35, 45, and 50 °C have 10% absolute average
check the reliability and accuracy of the experimental setup. Fig. 3(A) error with respect to Dittus–Boelter [36] relation and for all the different
shows the experimental results for water flow through the radiator at EG temperatures at the radiator inlet including 40, 45, and 50 °C, Vajjha
constant inlet temperature of 50 °C. It is shown that the higher Reynolds et al.'s [37] relation has 6% absolute average error.
number increases the heat transfer coefficient of pure water. The
experimental data has been compared with following empirical
correlation suggested by Dittus–Boelter [36] in turbulent flow: 5.2. Heat transfer to water based and EG based nanofluids

0:8 0:3
Nu = 0:0236Re Pr ð10Þ The nanofluid is implemented in different Al2O3 concentrations,
i.e. 0.1, 0.3, 0.5, 0.7, and 1 vol.% and at different flow rates of 2, 3, 4, 5,
Two correlations were developed by Vajjha et al. [37] for and 6 l per minute. In order to consider the effect of temperature on
nanofluids from the numerical analysis of the flat tube geometry thermal performance of the radiator, different inlet temperatures
which were shown in Eq. (11): have been applied for each concentration. The inlet temperatures
include 35, 45, and 50 °C for the water based nanofluids and 45, 50,
8    
>
> Dh 0:3 d and 60 °C for EG based nanofluids. Fig. 4(A) and (B) shows the heat
>
< Nu = 1:9421 × RePr z RePr ≥33:33
z transfer enhancement due to the replacement of water as a
    ð11Þ conventional coolant with water and EG based nanofluids respective-
>
> D d
>
: Nu = 6:1 + 0:003675 × RePr h RePr ≤33:33 ly. As can be seen, the ratio of the nanofluid Nusselt number to the
z z
base fluid Nusselt number (Nunf/Nubf) has increased by enhancement
in the concentrations of nanoparticle at constant Reynolds number for
The average Nusselt number can be calculated as follow: both nanofluids. By the addition of only 1 vol.% of Al2O3 nanoparticle
L
into the water or EG, an increase of about 40% in comparison with the
1 pure water and pure EG Nusselt number was recorded.
Nuave: = ∫ Nu: dz ð12Þ
L0 The Nusselt number increases monotonously with Reynolds
number, but Nunf/Nubf presents different variation tendencies with
Fig. 3(B) represents the comparison of experimental results for pure Reynolds number for these two nanofluids. For the water based
EG entering the radiator at constant temperature of 40 °C with the nanofluid it is obvious from Fig. 4(A) that Nunf/Nuw increases with
prediction of Vajjha et al. [37] correlation in laminar flow. As can be seen Reynolds number and in higher concentrations of nanoparticle the
in Fig. 3(A) and 3(B), the experimental results show good agreements effect of Reynolds number becomes pronounced. This result is in
1288 S.M. Peyghambarzadeh et al. / International Communications in Heat and Mass Transfer 38 (2011) 1283–1290

that an increase in the fluid inlet temperature (in the range of our
experiments) slightly improves the heat transfer performance. Inspect-
ing the results reveals that increasing the inlet temperature of water
based nanofluids from 35 °C to 50 °C can enhance Nusselt number up to
16%. For EG based nanofluids, the temperature elevation from 45 to
60 °C creates maximum enhancement of 7%. This variation in Nusselt
number may be attributed to the effect of temperature on the physical
properties and also to the increased effect of test liquid radiation to the
internal wall of the tubes.

5.4. Heat transfer to water/EG binary mixture nanofluids

A conventional fluid usually used in the car radiator is the mixture

of water/EG in different concentrations depending on the regional
weather. In order to have more insights to the influence of
nanoparticle addition in the radiator coolant, three different concen-
trations of water/EG binary mixtures which include 5, 10, and 20 vol.%
EG were prepared as the base fluids. Four different values of Al2O3
nanoparticle (0, 0.05, 0.15, and 0.3 vol.%) were added to each
concentration of water/EG mixtures and finally the effect of flow
rate on the heat transfer performance was studied for each case. All
the obtained experimental data are summarized in Table 5. The results
obtained for pure water and pure EG based nanofluids (which were
shown previously) were also presented in this table for more
comfortable comparison of the results. Furthermore, due to the
large variations in the physical properties of the base fluids, it is not
possible to imply Nusselt number as a function of Reynolds number.
Reynolds number greatly reduced when EG concentration enhances.
At extreme conditions, Reynolds number changes between 9000 and
23,000 for water based nanofluids while it changes between 1200 and

Fig. 4. Variations of dimensionless Nusselt number at different Reynolds numbers as a

function of nanoparticle concentration (Tin = 45 °C). A) Water based nanofluids. B) EG
based nanofluids.

contradiction to the EG based nanofluids in which Nunf/NuEG exhibited

irregular trends against Reynolds number, as can be seen in Fig. 4(B).
It was shown that when small amounts of Al2O3 nanoparticles are
added to the base fluids, the density and the thermal conductivity
increase and the specific heat decreases slightly while the viscosity
increases more markedly compared to the base fluid [1]. These
variations, however, are too small (of about 4%) to explain heat transfer
enhancement of up to 40% gained in this study. Heris et al. [12] have
done experiments with Al2O3 nanoparticles in water under laminar flow
up to turbulence. They found the heat transfer enhancement as high as
40% with Al2O3 particles while the thermal conductivity enhancement
was less than 15%. Many researchers have suggested that in fact
Brownian motion is one of the most important factors in the
enhancement of heat transfer. The presence of nanoparticles and their
random motion within the base fluid cause the thickness of thermal
boundary layer to reduce and it has important contribution to such heat
transfer improvement [38]. This random motion of ultra-fine particles
would create a slip velocity between the solid particles and the fluid
medium [23]. Xuan and Roetzel [26] also suggested the role of small
perturbations in the temperature and velocity formulation to account
for the Brownian motion.

5.3. Effect of temperature on heat transfer coefficient of nanofluids

Fig. 5(A) and (B) compares the nanofluid Nusselt numbers at Fig. 5. Effect of inlet temperature on the Nusselt numbers. A) Water based nanofluid at
different inlet temperatures in order to analyze the effect of temperature the nanoparticle concentration of 1 vol.%. B) EG based nanofluid at the nanoparticle
variation on the heat transfer performance of the car radiator. It is clear concentration of 0.7 vol.%.
 S.M. Peyghambarzadeh et al. / International Communications in Heat and Mass Transfer 38 (2011) 1283–1290 1289

2500 for EG based nanofluids. It should be mentioned that all the data
in Table 5 have been obtained at the constant liquid inlet temperature
to the radiator namely 45 °C. Data in Table 5 obviously indicate that
higher Nusselt numbers have been obtained for the pure water
compared to pure EG at constant flow rate (more than twice). This is
basically due to the sharp differences in the physical properties of the
two pure liquids. For the sake of clarity, some of the important
physical properties of water and EG were compared in Table 2 at the
temperature of 45 °C. The Prandtl number (Pr) as an important
parameter affecting the heat transfer coefficient differs greatly in
these two liquids and it can be the main reason for the lower Nusselt
number in EG compared to water. Furthermore, it is obvious from
Table 5 that the addition of EG into water decreases the Nusselt
numbers. It is the reason for mixing of EG with water; water does a
much better job at the cooling of the engine. In all the experimental
data shown in Table 5, it is proven that the addition of nanoparticle to
each base fluid enhances the heat transfer coefficient. This effect
manifests itself at water/EG binary mixtures with lower concentra-
tions of EG. Different correlations were introduced for the prediction
of nanofluid forced convection heat transfer in the literature up to
now. Xuan and Li [39] suggested two empirical correlations for
laminar and turbulent flow of the nanofluids in tube:

0:218

0:754 0:333 0:4
Nu = 0:4328 1 + 11:285φ Pe Re Pr Re b 2100 ð13Þ

0:001

0:6886 0:9238
Nu = 0:0059 1 + 7:6286φ Pe Re Re b 2100 ð14Þ

The experimental data obtained in the present investigation have

been compared with these two correlations. Fig. 6(A) and (B)
compare the experimental results for the Nusselt number of Al2O3/
water and Al2O3/EG nanofluids respectively with the prediction of
Xuan and Li correlations. Very good agreement can be seen in these
two figures. Calculating the absolute average errors reveals that the
prediction error for water based nanofluids is 7% and for EG based Fig. 6. Comparison of measured Nusselt numbers with those predicted from Xuan and Li
nanofluids is 12.5%. As can be seen in Fig. 6, Xuan and Li correlations correlation [39]. A) Water based nanofluid. B) EG based nanofluid.
almost over-predict Nusselt number for these nanofluids. The over-
prediction of these empirical correlations has been mentioned before
by other investigators in the case of other nanofluids [40].
total heat transfer rates have been observed with the nanoparticle
6. Conclusion addition. A highest Nusselt number enhancement up to 40% was
obtained at the best conditions for both nanofluids. The experimental
In this paper, the convective heat transfer enhancement of water results have demonstrated that the heat transfer behaviors of the
and EG based nanofluids as the coolants inside flat aluminum tubes of nanofluids were highly depended on the particle concentration and
the car radiator has been investigated. Significant increases of the the flow conditions and weakly dependent on the temperature. Our
results indicate that nanofluids have great potential for heat transfer
enhancement and are highly suited to apply in practical heat transfer
processes. This provides promising ways for engineers to develop
Table 5
Experimental Nusselt numbers measured for different concentrations of EG/water highly compact and effective radiators for cars. These higher heat
mixtures as the base fluid (inlet temperature 45 °C). transfer coefficients obtained by using nanofluid instead of water
allow the working fluid in the car radiator to be cooler. The addition of
φ (%)
nanoparticles to the coolant has the potential to improve automotive
Flow rate EG 5% water 95% EG 10% water 90% and heavy-duty engine cooling rates or equally causes to remove the
(l/min)
0 0.05 0.15 0.3 0 0.05 0.15 0.3 engine heat with a reduced-size cooling system. Smaller cooling
3 63.93 66.58 71.03 76.67 63.28 66.22 71.94 77.50 systems result in smaller and lighter radiators, which in turn benefit
4 79.26 81.56 86.50 92.24 79.45 81.98 87.53 92.54 almost every aspect of vehicle performance and lead to increased fuel
5 89.15 92.59 97.19 103.16 91.27 93.93 99.99 104.38 economy.
6 101.87 104.95 109.31 114.56 99.99 108.07 111.55 114.32

φ (%) References
Flow rate EG 20% water 80% Pure EG Pure water [1] S.M. Peyghambarzadeh, S.H. Hashemabadi, M. Seifi Jamnani, S.M. Hoseini,
(l/min) Improving the cooling performance of automobile radiator with Al2O3/water
0 0.05 0.15 0.3 0 0.3 0 0.3
nanofluid, Applied Thermal Engineering 31 (2011) 1833–1838.
3 61.59 63.45 68.78 71.78 24.69 27.01 67.76 76.77 [2] S. Rebsdat, D. Mayer, “Ethylene Glycol” in Ullmann's Encyclopedia of Industrial
4 79.35 82.43 87.29 90.81 30.23 32.63 83.72 99.28 Chemistry, Wiley-VCH, Weinheim, 2002.
5 90.48 92.55 98.83 102.40 33.67 36.00 96.46 115.69 [3] S.M. Fotukian, M. Nasr Esfahany, Experimental study of turbulent convective heat
transfer and pressure drop of dilute CuO/water nanofluid inside a circular tube,
6 90.77 101.38 107.23 110.55 37.47 40.20 - -
International Communications in Heat and Mass Transfer 37 (2010) 214–219.
1290 S.M. Peyghambarzadeh et al. / International Communications in Heat and Mass Transfer 38 (2011) 1283–1290

[4] W. Duangthongsuk, S. Wongwises, An experimental study on the heat transfer [21] K.Y. Leong, R. Saidur, S.N. Kazi, A.H. Mamun, Performance investigation of an
performance and pressure drop of TiO2–water nanofluids flowing under a automotive car radiator operated with nanofluid-based coolants (nanofluid as a
turbulent flow regime, International Journal of Heat and Mass Transfer 53 (2010) coolant in a radiator), Applied Thermal Engineering 30 (2010) 2685–2692.
334–344. [22] L. Godson, B. Raja, D. Mohan La, S. Wongwises, Enhancement of heat transfer using
[5] M. Emami Meibodi, M. Vafaie-Sefti, A.M. Rashidi, A. Amrollahi, M. Tabasi, H. Sid nanofluids—an overview, Renewable and Sustainable Energy Reviews 14 (2010)
Kalal, The role of different parameters on the stability and thermal conductivity of 629–641.
carbon nanotube/water nanofluids, International Communications in Heat and [23] S. Kakaç, A. Pramuanjaroenkij, Review of convective heat transfer enhancement
Mass Transfer 37 (2010) 319–323. with nanofluids, International Journal of Heat and Mass Transfer 52 (2009)
[6] P. Naphon, P. Assadamongkol, T. Borirak, Experimental investigation of titanium 3187–3196.
nanofluids on the heat pipe thermal efficiency, International Communications in [24] X.Q. Wang, A.S. Mujumdar, A review on nanofluids — Part II: experiments and
Heat and Mass Transfer 35 (2008) 1316–1319. applications, Brazilian Journal of Chemical Engineering 25 (4) (2008) 631–648.
[7] H. Demir, A.S. Dalkilic, N.A. Kürekci, W. Duangthongsuk, S. Wongwises, Numerical [25] R.J. Moffat, Describing the uncertainties in experimental results, Experimental
investigation on the single phase forced convection heat transfer characteristics of Thermal and Fluid Science 1 (1988) 3–17.
TiO2 nanofluids in a double-tube counter flow heat exchanger, International [26] Y. Xuan, W. Roetzel, Conceptions for heat transfer correlation of nanofluids,
Communications in Heat and Mass Transfer 38 (2) (2011) 218–228. International Journal of Heat and Mass Transfer 43 (2000) 3701–3708.
[8] W. Duangthongsuk, S. Wongwises, Heat transfer enhancement and pressure drop [27] X. Wang, X. Xu, S.U.S. Choi, Thermal conductivity of nanoparticles–fluid mixture,
characteristics of TiO2–water nanofluid in a double-tube counter flow heat Journal of Thermophysics and Heat Transfer 13 (4) (1999) 474–480.
exchanger, International Journal of Heat and Mass Transfer 52 (2009) 2059–2067. [28] W. Duangthongsuk, S. Wongwises, Effect of thermophysical properties models on
[9] D. Wen, Y. Ding, Experimental investigation into convective heat transfer of the predicting of the convective heat transfer coefficient for low concentration
nanofluids at the entrance region under laminar flow conditions, International nanofluid, International Communications in Heat and Mass Transfer 35 (2008)
Journal of Heat and Mass Transfer 47 (2004) 5181–5188. 1320–1326.
[10] M.S. Liu, M.C.C. Lin, I.T. Huang, C.C. Wang, Enhancement of thermal conductivity [29] N. Masoumi, N. Sohrabi, A. Behzadmehr, A new model for calculating the effective
with CuO for nanofluids, Chemical Engineering and Technology 29 (1) (2006) viscosity of nanofluids, Journal of Physics D: Applied Physics 42 (2009) 055501.
72–77. [30] M. Kole, T.K. Dey, Viscosity of alumina nanoparticles dispersed in car engine
[11] B.C. Pak, I.Y. Cho, Hydrodynamic and heat transfer study of dispersed fluids with sub- coolant, Experimental Thermal and Fluid Science 34 (2010) 677–683.
micron metallic oxide particles, Experimental Heat Transfer 11 (1998) 151–170. [31] P.K. Namburu, D.P. Kulkarni, D. Misra, D.K. Das, Viscosity of copper oxide
[12] S.Z. Heris, S.Gh. Etemad, M. Nasr Esfahany, Experimental investigation of oxide nanoparticles dispersed in ethylene glycol and water mixture, Experimental
nanofluids laminar flow convective heat transfer, International Communications Thermal and Fluid Science 32 (2007) 397–402.
in Heat and Mass Transfer 33 (4) (2006) 529–535. [32] J.Y. Huot, E. Battistel, R. Lumry, G. Villeneuve, J.F. Lavallee, A. Anusiem, C. Jolicoeur,
[13] S.Z. Heris, M. Nasr Esfahany, S.Gh. Etemad, Experimental investigation of A comprehensive thermodynamic investigation of water-ethylene glycol mix-
convective heat transfer of Al2O3/water nanofluid in circular tube, International tures at 5, 25, and 45 °C, Journal of Solution Chemistry 17 (7) (1988) 601–636.
Journal of Heat and Fluid Flow 28 (2) (2007) 203–210. [33] T. Sun, A.S. Teja, Density, viscosity, and thermal conductivity of aqueous ethylene,
[14] W.Y. Lai, B. Duculescu, P.E. Phelan, R.S. Prasher, Convective heat transfer with diethylene, and triethylene glycol mixtures between 290 K and 450 K, Journal of
nanofluids in a single 1.02-mm tube, Proceedings of ASME International Chemical Engineering Data 48 (2003) 198–202.
Mechanical Engineering Congress and Exposition (IMECE 2006), 2006. [34] Z. Nan, B. Liu, Z. Tan, Calorimetric investigation of excess molar heat capacities for
[15] J.Y. Jung, H.S. Oh, H.Y. Kwak, Forced convective heat transfer of nanofluids in water + ethylene glycol from T = 273.15 to T = 373.15 K, The Journal of Chemical
microchannels, Proceeding of ASME International Mechanical Engineering Thermodynamics 34 (6) (2002) 915–926.
Congress and Exposition (IMECE 2006), 2006. [35] J. Dai, L. Wang, Y. Sun, L. Wang, H. Sun, Prediction of thermodynamic, transport
[16] K.V. Sharma, L. Syam Sundar, P.K. Sarma, Estimation of heat transfer coefficient and vapor–liquid equilibrium properties of binary mixtures of ethylene glycol and
and friction factor in the transition flow with low volume concentration of Al2O3 water, Fluid Phase Equilibria 301 (2011) 137–144.
nanofluid flowing in a circular tube and with twisted tape insert, International [36] F.W. Dittus, L.M.K. Boelter, Heat Transfer in Automobile Radiators of Tubular Type,
Communications in Heat and Mass Transfer 36 (2009) 503–507. University of California Press, Berkeley, CA, 1930, pp. 13–18.
[17] C.J. Ho, L.C. Wei, Z.W. Li, An experimental investigation of forced convective [37] R.S. Vajjha, D.K. Das, P.K. Namburu, Numerical study of fluid dynamic and heat
cooling performance of a microchannel heat sink with Al2O3/water nanofluid, transfer performance of Al2O3 and CuO nanofluids in the flat tubes of a radiator,
Applied Thermal Engineering 30 (2009) 96–103. International Journal of Heat and Fluid Flow 31 (2010) 613–621.
[18] C.T. Nguyen, G. Roy, C. Gauthier, N. Galanis, Heat transfer enhancement using [38] S.E.B. Maiga, S.J. Palm, C.T. Nguyen, G. Roy, N. Galanis, Heat transfer enhancement
Al2O3–water nanofluid for an electronic liquid cooling system, Applied Thermal by using nanofluids in forced convection flows, International Journal of Heat and
Engineering 27 (2007) 1501–1506. Fluid Flow 26 (2005) 530–546.
[19] H. Xie, Y. Li, W. Yu, Intriguingly high convective heat transfer enhancement of [39] Y. Xuan, Q. Li, Investigation on convective heat transfer and flow features of
nanofluid coolants in laminar flows, Physics Letters A 374 (2010) 2566–2568. nanofluids, Journal of Heat Transfer 125 (2003) 151–155.
[20] R. Strandberg, D.K. Das, Finned tube performance evaluation with nanofluids and [40] B. Farajollahi, S.Gh. Etemad, M. Hojjat, Heat transfer of nanofluids in a shell and
conventional heat transfer fluids, International Journal of Thermal Sciences 49 tube heat exchanger, International Journal of Heat and Mass Transfer 53 (2010)
(2010) 580–588. 12–17.