Case Studies in Thermal Engineering 2 (2014) 50–61

Contents lists available at ScienceDirect

Case Studies in Thermal Engineering

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

Study of forced convection nanofluid heat transfer in the automotive

cooling system
Adnan M. Hussein a,b,1, R.A. Bakar a, K. Kadirgama a
a
Faculty of Mechanical Engineering, University Malaysia Pahang, Pekan 26600, Pahang, Malaysia
b
Al-Haweeja Institute, Foundation of Technical Education, Iraq

a r t i c l e in f o abstract

Article history: The heat transfer enhancement for many industrial applications by adding solid nano-
Received 15 November 2013 particles to liquids is significant topics in the last 10 years. This article included the friction
Accepted 2 December 2013 factor and forced convection heat transfer of SiO2 nanoparticle dispersed in water as a
Available online 11 December 2013
base fluid conducted in a car radiator experimentally and numerically. Four different
Keywords: concentrations of nanofluids in the range of 1–2.5 vol% have been used. The flowrate
Laminar changed in the range of 2–8 LPM to have Reynolds number with the range 500–1750.
Nanofluid The results showed that the friction factor decreases with an increase in flowrate and
Heat transfer increase with increasing in volume concentration. Furthermore, the inlet temperature to
Car radiator
the radiator has insignificantly affected to the friction factor. On the other side, Nusselt
CFD
number increases with increasing in flowrate, nanofluid volume concentration and inlet
temperature. Meanwhile, application of SiO2 nanofluid with low concentrations can
enhance heat transfer rate up to 50% as a comparison with pure water. The simulation
results compared with experimental data, and there is a good agreement. Likewise, these
results compared to other investigators to be validated.
& 2013 The Authors. Published by Elsevier Ltd. Open access under CC BY-NC-ND license.

1. Introduction

In the last 10 years, there has been more attention paid to enhance the convective heat transfer performance of nanofluids [1],
due to the recognition that in practical applications nanofluids. Experimental studies of friction factor and nanofluids heat
transfer enhancement with the flow velocity and nanofluid volume fraction inside heated tube under laminar flow condition has
been presented by [2–5]. There are many different applications of thermal-fluid systems, including automotive cooling systems
[6]. Base fluids, such as water, ethylene glycol and glycerol, have been used as conventional coolants in automobile radiators for
many years; however, these fluids have low thermal conductivities. The low thermal conductivities have thus prompted
researchers to search for fluids with higher thermal conductivities than that of conventional coolants. Therefore, nanofluids have
been used instead of the commonly used base fluids [7–9]. A numerical study on laminar heat transfer using CuO– and Al2O3–
ethylene glycol and water inside a flat tube of a car radiator was performed by Vajjha et al. [10]. A CFD model of the mass
flowrate of air passing across the tubes of a car radiator was presented by Peyghambarzadeh et al. [11]. The air flow was
simulated using the commercial software ANSYS 12.1, where the geometry was created in the software SOLID WORKS, followed
by creating both the surface mesh and the volume mesh accordingly. The results were compared and verified according to the

E-mail address: adnanphd2012@gmail.com (A.M. Hussein).

1
Tel.: þ 60 94246232; fax: þ60 94246222.

2214-157X & 2013 The Authors. Published by Elsevier Ltd. Open access under CC BY-NC-ND license.
http://dx.doi.org/10.1016/j.csite.2013.12.001
 A.M. Hussein et al. / Case Studies in Thermal Engineering 2 (2014) 50–61 51

Nomenclature μ viscosity [N s/m2]

ρ density [kg/m3]
C specific heat [W/kg oC] τ shear stress [N/m2]
D diameter [m] ϕ volume concentration
E energy [W]
f friction factor Subscripts
htc convection heat transfer coefficient [W/m2 oC]
k thermal conductivity [W/m oC] f liquid phases
Nu Nusselt Number [htcD/Knf] p solid particle
P pressure [N/m2] nf nanofluid
Pr Prandtl number [Cμ/Knf] h hydraulic
Re Reynolds number [ρnfDhu/Knf]
u velocity [m/s]

known physical situation and existing experimental data. The results serve as a good database for future investigations. New
correlations for the viscosity and thermal conductivity of nanofluids as a function of volumetric particle concentration and
temperature developed from the experiments were used in this paper. The convective heat transfer coefficient and shear stress of
the nanofluid showed marked improvement over the base fluid, showing higher magnitudes in the flat regions of the tube. The
results showed that increasing the nanofluid volume fraction increased the friction factor and convective heat transfer
coefficient; however, there was also an increase in pressure loss as the particle volume fraction increased. A numerical study that
analyzed mixed convection flows in a U-shaped grooved tube “in a radiator” was conducted by Park and Pak [12]. A modified
SIMPLE algorithm for the irregular geometry was developed to determine the flow and temperature field. The results have been
used as fundamental data for tube design by suggesting optimal specifications for radiator tubes. Two liquids, water and an
ethylene glycol/water mixture, used as the coolant fluid in a meso-channel heat exchanger were studied numerically by
Dehghandokht et al. [13]. The predicted results (heat transfer rate, pressure and temperature drops in the coolants) from the
numerical simulation were compared with the experimental data for the same geometrical and operating conditions and
showed good agreement. Additionally, the results showed the heat exchanger was enhanced, with heat transfer rate
approximately 20% higher than that of a straight slab of the same length; the enhanced heat exchanger has a good potential
to be used as a car radiator with reasonably enhanced heat transfer characteristics using an ethylene glycol/water mixture as the
coolant. The application of a copper–ethylene glycol nanofluid in a car cooling system has been studied by Leong et al. [14]. The
friction factor has been evaluated experimentally by [15–17], but heat transfer enhancement using water and ethylene glycol
water in a multi-port serpentine meso-channel heat exchanger has been determined as [18]. The overall heat transfer
coefficient of CuO/water nanofluids investigated experimentally under laminar flow regime in a car radiator by Naraki et al.
[19]. The results showed that the overall heat transfer coefficient with nanofluid was more than the base fluid. The overall heat
transfer coefficient increased with the enhancement in the nanofluid concentration from 0 to 0.4 vol%. Conversely, the overall
heat transfer coefficient decreased with increasing the nanofluid inlet temperature from 50 to 80 1C. The implementation of
nanofluid increases the overall heat transfer coefficient up to 8% at nanofluid concentration of 0.4 vol% in comparison with the
base fluid. Theoretical results of nanofluid heat transfer of Maiga et al. [20] have been verified with the correlations of Seider–
Tate [21] for base fluid only. Forced convection heat transfer of both CuO and Fe2O3 nanofluids flow through the car radiator
has been studied experimentally by Peyghambarzadeh et al. [22]. An experimental investigation for the determination of heat
transfer coefficient with SiO2 nanofluid under cooling and heating conditions at fluid inlet temperatures of 20, 50 and 70 1C for
Reynolds number range from 200 to 10,000 and 22 nm size diameter particles suspension in water has been conducted by
Ferrouillat et al. [23]. The heat transfer coefficients at particle volume concentrations of 2.3% and 19% have been determined as
10% and 50% respectively greater than the values obtained with water.
In this paper, an experimental and computational study of forced convection heat transfer in a car radiator with SiO2
nanoparticles suspension in water under laminar flow is presented. The test rig is setup to measure temperatures and
pressure drop between inlet and outlet of the radiator. Furthermore, the CFD model is simulated by finite volume method
and using FLUENT software to solve governing equations. The range of Reynolds number is 250–1750 and the volume
fraction of nanofluid is from 1% to 2.5%. The aim of this study is to enhance heat transfer in the car cooling system. The
comparison among experimental and simulation results are carried out then validated with other researchers data.

2. Experimental work

2.1. Experimental setup and procedure

The test rig shown in Fig. 1a has been used to measure heat transfer coefficient and friction factor in the automotive
engine radiator. This experimental setup includes a reservoir plastic tank, electrical heater, a centrifugal pump, a flow meter,
52 A.M. Hussein et al. / Case Studies in Thermal Engineering 2 (2014) 50–61

L d

Fig. 1. (a) Schematic diagram of experimental test rig and (b) flat tube configuration.

tubes, valves, a fan, a DC power supply; 10 thermocouples type T for temperature measurement, manometer tube with
mercury and heat exchanger (automobile radiator). An electrical heater (1500 W) inside a plastic storage tank (40 cm height
and 30 cm diameter) put to represent the engine and to heat the fluid. A voltage regular (0–220 V) provided the power to
keep the inlet temperature to the radiator from 60 to 80 1C. A flow meter (0–70 LPM) and two valves used to measure and
control the flowrate. The fluid flows through plastic tubes (0.5 in.) by a centrifugal pump (0.5 hp and 3 m head) from the
tank to the radiator at the flowrate range 2–8 LPM. The total volume of the circulating fluid is 3 l and constant in all the
experimental steps. Two thermocouples (copper–constantan) types T have been fixed on the flow line for recording the inlet
and outlet fluid temperatures. Eight thermocouples type T have been fixed to the radiator surface to ensure more of surface
area measurement. Two thermocouples type T also fixed in front of the fan and another side of radiator to measure air
temperatures. Very small thickness and high thermal conductivity of the copper flat tubes caused to make the inside
temperature of the tube with the outside one are equated. A handheld (  40 1C to 1000 1C) digital thermometer with the
accuracy of 70.1 1C used to read all the temperatures from thermocouples. Calibration of thermocouples and thermometers
carried out by using a constant temperature water bath and their accuracy estimated to be 0.15 1C [19]. Two small plastic
tubes with 0.25 in. diameter connected at inlet and outlet radiator and joined to U-tube mercury manometer with the
accurate scale 0.5 mm Hg to measure the pressure drop on the inlet and outlet of it. The car radiator has louvered fin and 32
flat vertical copper tubes with flat cross sectional area. The distances among the tube rows filled with thin perpendicular
copper fins. For the air side, an axial force fan (1500 rpm) installed close on axis line of the radiator. The DC power supply
(type Teletron 10–12 V) used to turn the axial fan instead of a car battery. The thermal properties of nanofluids have been
measured by [24] and used in this study. The pH values of nanofluid have been measured using OAKTON device shown in for
the nanofluid volume concentration of (1–2.5%). The pH values before and after experimental tests referring to nanofluids
stability and changing in thermal properties. If the pH values of the suspension decrease, the force among particles will
increase which the moving of the nanoparticles suspension lead to enhance the heat transfer process. So, to augment heat
transfer for many applications should keep low values of the nanofluid pH [25]. The pH meter has been calibrated using a
single point calibration technique, with a Hatch standard buffer solution of pH 7.00 70.02. According to [24] to find out the
pH values, samples with pH of 8.2, 8.6, 9.3 and 10.2 were prepared. Samples checked after finishing each test but found no
visible sedimentation. SiO2 nanoparticles with volume concentration (1, 1.5, 2.0 and 2.5 vol%) and 30 nm size diameter
added to base fluids (deionized water). The thermal properties of SiO2 nanoparticle and base fluids are shown in Table 1.

2.2. Experimental data analysis

According to Newton's cooling law the following procedure followed to obtain heat transfer coefficient and corresponding
Nusselt number as [26]:
Q ¼ hAΔT ¼ hAs ðT b T s Þ ð1Þ
 A.M. Hussein et al. / Case Studies in Thermal Engineering 2 (2014) 50–61 53

As is surface area of tube, Tb is the bulk temperature,

T in þT out
Tb ¼ ð2Þ
2
Tin and Tout are inlet and outlet temperatures, respectively and Ts is the tube wall temperature which is the mean value by two
surface thermocouples as
T 1 þ ::: þT 8
Ts ¼ ð3Þ
8
And heat transfer rate calculated by
Q ¼ m CΔT ¼ m CðT in  T out Þ ð4Þ

m is mass flowrate which determined as
m ¼ ρV  ð5Þ
The heat transfer coefficient can be evaluated by combining Eqs. (1) and (4)
m CðT in  T out Þ
hexp ¼ ð6Þ
As ðT b T s Þ
and Nusselt number can be calculated as
hexp Dh
Nu ¼ ð7Þ
k
Dh is the hydraulic diameter of the tube which estimated by description the problem undertaken as cylindrical geometry
coordinates shown in Fig. 1b, dimensions of the flat tube are major and minor diameter (D¼9 mm, d¼3 mm), the length (L) and
hydraulic diameter (Dh) of the flat tube are 345 mm and 4.68 mm. Reynolds number calculated regarded hydraulic diameter (Dh)
as [27]
h i
2
4  area 4  4π d þ ðD  dÞ  d
Dh ¼ Dh ¼ ð8Þ
perimeter π  d þ 2  ðD  dÞ
Reynolds number (Re) is determined as
ρnf Dh u
ReD ¼ : ð9Þ
μnf

Table 1
Thermal property of nanoparticle and base fluids.

Materials Density (Kg/m3) Specific heat (J/kg 1C) Thermal conductivity (W/m 1C) Viscosity (Pa s) References

SiO2 2220 745 1.4 – [24]

Pure water 998 4180 0.6067 0.0014 [24]

Table 2
Uncertainties of measured parameters.

No. Parameter Values Uncertainties (%)

1 Dh 4.68 mm 0.78
2 Re 250–1750 4.3
3 Nu 8–22 2.42

Fig. 2. Meshing by GAMBIT.

54 A.M. Hussein et al. / Case Studies in Thermal Engineering 2 (2014) 50–61

2.3. Uncertainty analysis

The uncertainty analysis performed by calculating the measurements error. The Reynolds number uncertainty range may
be come from the errors in the measurement of volume flowrate (Eq. (5)) and hydraulic diameter of the tubes (Eq. (8)). The
uncertainty of Nusselt number refers to the errors in the measurements of volume flowrate (Eq. (5)), hydraulic diameter
(Eq. (8)), and all the temperatures (Eqs. (2) and (3)). According to uncertainty analysis described by [28], the measurement
errors of the main parameters are summarized in Table 2. Furthermore, to check the reproducibility of the experiments,
some runs were repeated later which proved to be excellent.

0.40

Pure Water

Grid: 18x345

Grid: 36x345
0.30
Grid: 72x345

Grid: 108x345

Darcy Eq (13)
Friction Factor

0.20

0.10

0.00
0 500 1000 1500 2000
Reynolds Number

Pure Water
24
Grid: 18x345

Grid: 36x345

Grid: 72x345

20 Grid: 108x345

Shah-London Eq (14)
Nusselt Number

16

12

8
0 500 1000 1500 2000
Reynolds Number

Fig. 3. Grid independent test. (a) Friction factor and (b) Nusselt number.
 A.M. Hussein et al. / Case Studies in Thermal Engineering 2 (2014) 50–61 55

3. Theoretical analysis

3.1. Physical model

The assumptions of the study undertaken were steady, incompressible and Newtonian laminar fluid flow with constant
thermo-physical properties of the nanofluid. Additionally, heat conduction in the axial direction and wall thickness of the
tubes was neglected. Assuming, also, all thermal properties estimated at bulk temperature of the fluid.

3.2. Governing equations

The single-phase method can be adopted by using infinitesimal (less than 100 nm) solid particles in the liquid. The
dimensional conservation equations at steady state are the continuity, momentum and energy equations [26]
ð∇V Þ ¼ 0 ð10Þ

ρnf ð∇V ÞV ¼  ∇P þ μnf ∇2 V ð11Þ

ρnf C nf ðV∇ÞT ¼ knf ∇2 T ð12Þ

0.4
pure water
0.35
1%EXP.
0.3 1%CFD
Friction Factor

0.25 1.5%EXP.
1.5%CFD
0.2 2%EXP.
0.15 2%CFD
2.5%EXP.
0.1 2.5%CFD
0.05
0
0 500 1000 1500 2000
Reynolds Number

0.4 pure water

0.35 1%EXP.
1%CFD
0.3
Friction Factor

1.5%EXP.
0.25 1.5%CFD
2%EXP.
0.2 2%CFD
0.15 2.5%EXP.
2.5%CFD
0.1 Darcy Eq (13)
0.05
0
0 500 1000 1500 2000
Reynolds Number

0.4
pure water
0.35 1%EXP.
1%CFD
0.3
Friction Factor

1.5%EXP.
0.25 1.5%CFD
2%EXP.
0.2 2%CFD
2.5%EXP.
0.15
2.5%CFD
0.1 Darcy Eq (13)

0.05
0
0 500 1000 1500 2000
Reynolds Number
Fig. 4. Friction factor at different Reynolds number: (a) at 60 1C; (b) at 70 1C and (c) at 80 1C.
56 A.M. Hussein et al. / Case Studies in Thermal Engineering 2 (2014) 50–61

A low Reynolds number was used as an input parameter, and the pressure treatment was adopted using the SIMPLE scheme.
For all the governing solutions, when the residuals were lower than 10  6, solutions were considered converged. The results of the
simulation were compared with the equation for the friction factor which correlated by Darcy–Weisbach equation (Eq. (13)) and
Nusselt number which correlated by Shah–London equation (Eq. (14)) for laminar flow respectively as [29]
64
f¼ ð13Þ
Re
 ð1=3Þ  
Nuav ¼ 1:953 ReDh Pr DLh for RePrDh =L Z 33:33
    ð14Þ
Nuav ¼ 4:364 þ 0:0722 ReDh Pr DLh for RePrDh =L o 33:33

3.3. Boundary conditions

SiO2–water nanofluid volume concentrations of 1%, 1.5%, 2% and 2.5% at a base temperature of 25 1C were used as the
input fluids. For comparison purposes, water was also used as the working fluid. CFD studies were performed with uniform

49 pure water
44 1%EXP.
1%CFD
39
Nusselt Number

1.5%EXP.
34 1.5%CFD
2%EXP.
29
2%CFD
24 2.5%EXP.
19 2.5%CFD
Shah-London Eq (14)
14
9
4
0 500 1000 1500 2000
Reynolds Number

pure water
54 1%EXP.
1%CFD
Nusselt Number

44 1.5%EXP.
1.5%CFD
2%EXP.
34
2%CFD
2.5%EXP.
24 2.5%CFD
Shah-London Eq (14)
14

4
0 500 1000 1500 2000
Reynolds Number

pure water
64
1%EXP.
1%CFD
54 1.5%EXP.
Nusselt Number

1.5%CFD
44 2%EXP.
2%CFD
34 2.5%EXP.
2.5%CFD
24 Shah-London Eq (14)

14

4
0 500 1000 1500 2000
Reynolds Number
Fig. 5. Nusselt number at different Reynolds number: (a) at 60 1C; (b) at 70 1C and (c) at 80 1C.
 A.M. Hussein et al. / Case Studies in Thermal Engineering 2 (2014) 50–61 57

velocity profile at the inlet, and a pressure outlet condition was used at the outlet of the radiator. The walls of the radiator
tubes were assumed to be perfectly smooth. The Reynolds number was varied from 500 to 1750 at each iteration step as
input data. The output data are the friction factor and Nusselt number.

3.4. Grid independence test

Grids independence was determined using the GAMBIT software and was found for 72  345 with subdivisions in the
axial directions and the face of the flat area. The grid independence test for the physical model performed to determine the
most suitable size of the mesh faces. In this study, rectangular cells were used to mesh the surfaces of the tube wall, and the
surfaces of the gap, as shown in Fig. 2. Grid independence was checked using different grid systems, and four mesh faces
were considered, 18  345, 36  345, 72  345 and 108  345 for pure water. The friction factor and Nusselt number have
been determined for all four mesh faces, and the results all agreed with each other and Eqs. (13) and (14). All four mesh faces
could have been used, and in this study, the mesh face with 12,000 cells was adopted because it was the best in terms of
accuracy, as shown in Fig. 3.

100
pure water
90 1%EXP.
1%CFD
80 1.5%EXP.
1.5%CFD
70 2%EXP.
Tout

2%CFD
60 2.5%EXP.
2.5%CFD
50

40

30
0 2 4 6 8 10
Flowrate (LPM)

110
pure water
100 1%EXP.
1%CFD
90 1.5%EXP.
1.5%CFD
80 2%EXP.
Tout

2%CFD
70 2.5%EXP.
2.5%CFD
60

50

40
0 2 4 6 8 10
Flowrate (LPM)

115
pure water
105 1%EXP.
1%CFD
95 1.5%EXP.
1.5%CFD
85 2%EXP.
Tout

2%CFD
75 2.5%EXP.
2.5%CFD
65

55

45
0 2 4 6 8 10
Fowrate (LPM)
Fig. 6. Outlet temperature at different volume flowrate: (a) at 60 1C, (b) at 70 1C and (c) at 80 1C.
58 A.M. Hussein et al. / Case Studies in Thermal Engineering 2 (2014) 50–61

3.5. CFD simulation

CFD simulations were performed using FLUENT software with the solver strategy, and the GAMBIT software used to
analyze the problems. The governing equations have been solved by using the control volume approach for simulating the
governing equations numerically. Using FLUENT software for CFD analysis has been done in the literature, and a detailed
description of the mathematical model can be found in [30]. There are distinct steps to design a region using CFD: in the first
step, the pre-process stage, the geometry of the problem was constructed as a flat tube, and the computational mesh was
generated in GAMBIT. The next step involves creating a physical model and choosing the boundary conditions and other
parameters required to define the model setup and solving stage. All scalar values and velocity components of the problem
were calculated at the center of the control volume interfaces, where the grid schemes are used intensively. Throughout the
iterative process, the residuals were continually monitored. When the residuals for all governing equations were less than
10  6, all solutions were assumed to be converging. Finally, the results were obtained when the FLUENT iterations
converged, which was defined by a set of convergence criteria. The friction factor and Nusselt number inside the flat tube
could then be defined throughout the computational domain in the post-process stage.

4. Results and discussion

The friction factor at different Reynolds number and nanofluid volume concentration is shown in Fig. 4. It appears that
the friction factor decreases with increasing in Reynolds number and nanofluid volume concentration. Fig. 4a–c illustrates
the effect of inlet temperature on the friction factor. It seems that slightly effect of inlet temperature and nanofluid volume
concentration on the friction factor. A reason related to the slow flow is generated high friction factor. The friction factor at
Reynolds number less than 1000 has been given the maximum deviation is 82% but after that the maximum deviation

0.35

0.3 1%EXP at 60 oC
1%CFD at 60 oC
0.25 1%EXP at 70 oC
Friction factor

1%CFD at 70 oC
0.2
1%EXP at 80 oC
1%CFD at 80 oC
0.15
Darcy Eq (13)
0.1

0.05

0
0 2 4 6 8 10
Flowrate (LPM)

35
1%EXP at 60 oC
30 1%CFD at 60 oC
1%EXP at 70 oC
25 1%CFD at 70 oC
Nusselt Number

1%EXP at 80 oC
20 1%CFD at 80 oC
Shah-London Eq (14)
15

10

0
0 2 4 6 8 10
Flowrate (LPM)
Fig. 7. The effect of inlet temperature on: (a) friction factor and (b) Nusselt number.
 A.M. Hussein et al. / Case Studies in Thermal Engineering 2 (2014) 50–61 59

appears 40%. Fig. 5 shows Nusselt number at different Reynolds number and nanofluid volume concentration. The Nusselt
number increases with increasing of Reynolds number and nanofluid volume concentration. The deviation is approximately
40% when adding the nanoparticles on base fluid (pure water). Likewise, Fig. 5a–c demonstrates the effect of inlet radiator
temperature on Nusselt number. The highest values of Nusselt number found at inlet temperature 80 1C followed by at 70 1C
and finally at 60 1C inlet temperature. The maximum values of Nusselt number are 17.8, 21 and 25 at 60, 70 and 80 1C
respectively. This refers to high heat transfer from the radiator when high inlet temperature would apply. Fig. 6 shows the
outlet temperature of the radiator (Tout) with different volume flowrate circulating and nanofluid volume concentrations.
It seems that the outlet temperature increases with increasing in the volume flowrate and decreasing in nanofluid volume
concentration. The outlet temperature at pure water is the highest which referred the heat transfer increasing from the
radiator. It should be noted that all the data in Fig. 6a, b and c obtained when the radiator inlet temperatures were 60, 70
and 80 1C, respectively. It also showed that the outlet temperature increases with increasing of inlet temperature and the
difference between inlet and outlet temperature for each run. This means the heat transfer increases with increasing of inlet
temperature too. Fig. 7a shows the friction factor at different volume flowrates and different inlet temperatures. It appears
that the friction factor decreases with increasing of volume flowrate and slightly decreases with increasing of the inlet
temperature. Darcy equation (Eq. (13)) is also indicated as solid black line for pure water. The friction factor of 1% nanofluid
increases by approximately 15% as compared with pure water. On the other side, Fig. 7b illustrates Nusselt number at
different Reynolds numbers and inlet temperatures. The Shah–London equation (Eq. (14)) is also indicated as solid black line
for pure water. The Nusselt number increases with increasing of volume flowrate and inlet temperature. The Nusselt
number of 1% nanofluid at 80 1C has 52% deviation than pure water but 32% at 60 1C. Fig. 8 indicates the percentage of heat
transfer enhancement depending on nanofluid volume concentration and inlet temperature to the radiator. The percentage
enhancement can be evaluated by

Nunf  Nuw
Enhancement % ¼  100 ð15Þ
Nunf

It is observed that the heat transfer enhancement increases with increasing of nanofluid volume concentration and inlet
temperature respectively. The values of heat transfer enhancement are from 31% to 46% for nanofluid volume concentration

50
45
40
Enhancement %

35
30
25
20
15
10
5
0

60

50
Enhancement %

40

30

20

10

0
EXP at CFD at EXP at CFD at EXP at CFD at
60 oC 60 oC 70 oC 70 oC 80 oC 80 oC

Fig. 8. (a) The effect of nanofluid volume concentration and (b) the effect of inlet temperature.
60 A.M. Hussein et al. / Case Studies in Thermal Engineering 2 (2014) 50–61

from 1% to 2.5% while the values of heat transfer enhancement are from 39% to 56% for inlet temperature from 60 to 80 1C.
This refers to nanofluid volume concentration affected more than inlet temperature on heat transfer enhancement, but
significantly of using all of them. Fig. 9 demonstrates that the experimental and simulation data agreed with other
investigator data to be validated. Fig. 9a shows the friction factor results from the experimental and simulation work and
experimental data of [15,17] and the curve fitting data of [16]. The present work agreed with data of [15,16] with deviation,
not more than 9% because their data has been done on Al2O3 nanofluid, but the present work has been done on SiO2
nanofluid. On the other hand, Fig. 9b shows the validation of Nusselt number from experimental and simulation work with
curve fitting data of [20]. It seems that there is a good agreement with the maximum deviation 7% may be related to size
diameter all type of nanofluids.

5. Conclusions

The friction factor and forced convection heat transfer enhancement of SiO2 suspended in water carried out. Significant
increases of the friction factor and heat transfer enhancement observed with the volume concentration of nanoparticles
addition. The maximum values of friction factor increased to 22% for SiO2 nanoparticles dispersed in water with 2.5% volume
concentration. A highest Nusselt number enhance up to 40% obtained for SiO2 nanoparticles in water. The simulation results
showed that the friction factor and Nusselt number action of the nanofluids were highly depended on the volume
concentration and Reynolds number. On the other side, the experimental study concluded that there was a significant
increasing in Nusselt number and agreeing with the numerical study. The friction factor decreases with increasing of volume
flowrate and the inlet temperature. The friction factor at Reynolds number less than 1000 has a maximum deviation 82% but
after that the maximum deviation becomes 40%. A highest Nusselt number enhancement is 56% obtained for SiO2
nanoparticles in water as a base fluid. The experimental results showed that the Nusselt number behavior of the nanofluids

Present EXP Work (pure water)

0.5
Present CFD Work (pure water)
EXP data [17]

0.4 EXP data [15]

Present EXP Work (1%SiO2-W)
Friction Factor

Present CFD Work (1%SiO2-W)

0.3 Darcy Eq (13)
Curve fitting of [16]

0.2

0.1

0
0 500 1000 1500 2000
Reynolds Number

CFD Work (pure water)

22 EXP Work (pure water)

EXP Work (1%SiO2-W)
CFD Work (1%SiO2-W)
Nusselt Number

17 Shah-London Eq (14)
Sieder-Tate Corelation [21]
Curve fitting of [20]

12 Curve fitting of [22]

2
0 500 1000 1500 2000
Reynolds Number
Fig. 9. Validation of (a) friction factor and (b) Nusselt number.
 A.M. Hussein et al. / Case Studies in Thermal Engineering 2 (2014) 50–61 61

was highly depended on the volume concentration and the volume flowrate. The Nusselt number of 1% SiO2 nanofluid at
80 1C has 52% deviation than pure water but 32% at 60 1C. There is a good agreement among present data of friction factor
with data of [15–17] and Nusselt number data with data of [18,20]. The results proved that SiO2 nanofluid have high
potential for hydrodynamic flow and heat transfer enhancement and are highly appropriate to industrial and practical
applications. When adding nanoparticles to the base fluid the possible enhancement of car engine cooling rates will improve
otherwise the engine heat remove or reduce size cooling system. The small cooling systems have led to benefit almost every
aspect of car performance and increased in fuel economy.

References

[1] Abbasian AA, Amani J. Experimental study on the effect of TiO2–water nanofluid on heat transfer and pressure drop. Exp Therm Fluid Sci 2012;42:
107–15.
[2] Hojjat M, Etemad SG, Bagheri R, Thibault J. Laminar convective heat transfer of non-Newtonian nanofluids with constant wall temperature. Heat Mass
Transf 2011;47:203–9.
[3] Kayhania MH, Soltanzadeh H, Heyhat MM, Nazari M, Kowsary F. Experimental study of convective heat transfer and pressure drop of TiO2–water
nanofluid. Int Commun Heat Mass Transf 2012;39:456–62.
[4] Trivedi PK, Vasava NB. Study of the effect of mass flow rate of air on heat transfer rate in automobile radiator by CFD simulation using CFX. Int J Eng
Res Technol 2012;1(6):1–4.
[5] Saidur R, Leong KY, Mohammad HA. A review on applications and challenges of nanofluids. Renew Sustain Energy Rev 2011;15:1664–8.
[6] Oliet C, Oliva A, Castro J, Segarra CDP. Parametric studies on automotive radiators. Appl Therm Eng 2007;27:2033–43.
[7] Lee S, Choi SUS, Li S, Eastman J. Measuring thermal conductivity of fluids containing oxide nanoparticles. J Heat Transf 1999;121(2):280–9.
[8] Heris SZ. Convective heat transfer of a Cu/water nanofluid flowing through a circular tube. Exp Heat Transf 2009;22:217–27.
[9] Choi SUS, Wang X, Xu W. Thermal conductivity of nanoparticle–fluid mixture. J Thermophys Heat Transf 1999;13(4):470–84.
[10] Vajjha RS, Das DK. Experimental determination of thermal conductivity of three nanofluids and development of new correlations. Int J Heat Mass
Transf 2009;52:4675–82.
[11] Peyghambarzadeh SM, Hashemabadi SH, Jamnani MS, Hoseini SM. Improving the cooling performance of automobile radiator with Al2O3/water
nanofluid. Appl Therm Eng 2011;31:1833–8.
[12] Park KW, Pak HY. Flow and heat transfer characteristics in flat tubes of a radiator. Numer Heat Transf 2002;41:19–40.
[13] Dehghandokht M, Khan MG, Fartaj A, Sanaye S. Flow and heat transfer characteristics of water and ethylene glycol–water in a multi-port serpentine
meso-channel heat exchanger. Int J Therm Sci 2011;50:1615–27.
[14] Leong KY, Saidur R, Kazi SN, Mamun AM. Performance investigation of an automotive car radiator operated with nanofluid-based coolants (nanofluid
as a coolant in a radiator). Appl Therm Eng 2010;30:2685–92.
[15] Chandrasekar M, Suresh S, Chandra A. Experimental studies on heat transfer and friction factor characteristics of Al2O3/water nanofluid in a circular
pipe under laminar flow with wire coil inserts. Exp Therm Fluid Sci 2010;34:122–30.
[16] Hwang KS, Jang SP, Choi SUS. Flow and convective heat transfer characteristics of water-based Al2O3 nanofluids in fully developed laminar flow
regime. Int J Heat Mass Transf 2009;52:193–9.
[17] Sivashanmugam P, Suresh S. Experimental studies on heat transfer and friction factor characteristics of laminar flow through a circular tube fitted with
helical screw-tape inserts. Appl Therm Eng 2006;26:1990–7.
[18] Dehghandokht M, Khan MG, Fartaj A, Sanaye S. Flow and heat transfer characteristics of water and ethylene glycol water in a multi-port serpentine
meso-channel heat exchanger. Int J Therm Sci 2011;50:1615–27.
[19] Naraki M, Peyghambarzadeh SM, Hashemabadi SH, Vermahmoudi Y. Parametric study of overall heat transfer coefficient of CuO/water nanofluids in a
car radiator. Int J Therm Sci 2013;66:82–90.
[20] Maiga S, Palm SJ, Nguyen CT, Roy G, Galanis N. Heat transfer enhancement by using nanofluids in forced convection flows. Int J Heat Fluid Flow
2005;26(4):530–46.
[21] Sieder EN, Tate GE. Heat transfer and pressure drop of liquids in tubes. Ind Eng Chem 1939;28:1429–36.
[22] Peyghambarzadeh SM, Hashemabadi SH, Naraki M, Vermahmoudi Y. Experimental study of overall heat transfer coefficient in the application of dilute
nanofluids in the car radiator. Appl Therm Eng 2013;52:8–16.
[23] Ferrouillat S, Bontemps A, Ribeiro JP, Gruss JA, Soriano O. Hydraulic and heat transfer study of SiO2/water nanofluids in horizontal tubes with imposed
wall temperature boundary conditions. Int J Heat Fluid Flow 2011;32:424–39.
[24] Hussein AM, Bakar RA, Kadirgama K, Sharma KV. Experimental measurements of nanofluids thermal properties. Int J Autom Mech Eng 2013;7:850–64.
[25] Xie WY, Yang L, Lifei C. Discussion on the thermal conductivity enhancement of nanofluids Huaqing. Nanoscale Res Lett 2011;6:1–12.
[26] Bejan A. Convection heat transfer. New York: John Wiley & Sons Inc.; 2004.
[27] Mohammed HA, Hasan HA, Wahid MA. Heat transfer enhancement of nanofluids in a double pipe heat exchanger with louvered strip inserts. Int
Commun Heat Mass Transf 2013;40:36–46.
[28] Hyder HB, Abdullah S, Zulkifli R. Effect of oxides nanoparticle materials on the pressure loss and heat transfer of nanofluids in circular pipes. J Appl Sci
2012;12(13):1396–401.
[29] Vajjha RS, Das DK, Namburu PK. Numerical study of fluid dynamic and heat transfer performance of Al2O3 and CuO nanofluids in the flat tubes of a
radiator. Int J Heat Fluid Flow 2010;31:613–21.
[30] Fluent Incorporated. Fluent 6.2 User Manual; 2006.