New Journal and we have not received input yet 20 (2020) 100727

Contents lists available at ScienceDirect

Thermal Science and Engineering Progress

journal homepage: www.sciencedirect.com/journal/thermal-science-and-engineering-progress

Experimental investigation on the thermal performance of compact heat

exchanger and the rheological properties of low concentration mono and
hybrid nanofluids containing Al2O3 and CuO nanoparticles
Neeshoun Asokan a, *, Prem Gunnasegaran b, Vignesh Vicki Wanatasanappan b
a
College of Graduate Studies, Universiti Tenaga Nasional, Jalan IKRAM-UNITEN, 43000 Selangor, Malaysia
b
Institute of Power Engineering, Universiti Tenaga Nasional, Jalan IKRAM-UNITEN, 43000 Selangor, Malaysia

A R T I C L E I N F O A B S T R A C T

Keywords: The present work focused on preparing and studying the fluid and thermal properties such as density, viscosity,
Nanofluids specific heat capacity and thermal conductivity of low concentration hybrid nanofluids. The present work also
Rheological properties focused on comparing these properties and the thermal enhancements with other low concentration mono
Thermophysical properties
nanofluids. The nanofluids used for present work are Al2O3-CuO, Al2O3 and CuO combined with a 60:40 mixing
Thermal performance
Compact heat exchanger
ratio of distilled water and ethylene glycol acting as the base fluid. The volume concentration at which the
nanofluids are prepared are fixed 0.02%, 0.04% and 0.06%. The findings show that the thermal conductivity of
the hybrid samples is higher compare to its mono counterparts. The thermal conductivity of Al2O3-CuO showed
enhancements by 2.3% when compared to CuO and by 3.6% when compared to Al2O3. By using the prepared
samples as a working fluid in a compact fin and tube heat exchanger test rig, an experimental investigation is
conducted. This is to observe any form of improvements in terms of heat transfer by comparing the experimental
results of the hybrid nanofluids with its respective mono counterparts. Based on the experiments conducted, it is
found that the use of Al2O3-CuO hybrid nanofluids as a working fluid in a compact heat exchanger test rig
showed an increase in both Nusselt number and average heat transfer coefficients by 6.7% and 7.2% respectively
when compared to CuO nanofluids and by 17.9% and 12.1% respectively when compared to Al2O3 nanofluids.

1. Introduction conductivity, the conventional working fluids can be replaced by

nanofluids. Nanofluids, in general terms, are made up of metal or metal
Engineering applications such as air conditioning systems, thermal oxides. The suspending nanoparticles in the fluid needs to be below 100
power plants, automobiles and refrigeration are some of the few ex­ nm in size to be categorized as a nanofluid. The present study focuses on
amples in which heat exchangers (HX) are used. Heat exchanger is a comparing the thermophysical properties such as density, specific heat
device in which heat is transferred between two types of fluid without capacity, viscosity and thermal conductivity of low concentration
having to mix or come into direct contact. In terms of heat exchanger amounts of both mono and hybrid nanofluids. Numerous studies
systems, shell-and-tube heat exchangers (STHX) are one of the most used regarding the thermal conductivity of various hybrid nanofluid combi­
HX systems industrial related applications. STHXs are known to take up nations such as ZnO-Ag/water, TiO2-MWCNTs/water-ethylene glycol,
more than 30% of commonly used heat exchanger systems. the combination of water-based graphene wrapped carbon nanotubes
Conventional working fluid that are currently being used in heat with ZnO and CuO have been published. These studies shared a similar
exchanger systems are water, ethylene glycol and oil. These working finding in which the thermal conductivity of the hybrid nanofluids are
fluids have been used to cool various applications by transferring heat higher when compared to the corresponding base fluid [1–3].
away from the heat source. However, these conventional coolants Previous studies have also shown that the heat transfer coefficients of
possess low thermal conductivity properties which inhibits the overall mono of certain mono nanofluids are higher when compared to con­
thermal management performance of the heat exchanger system. ventional coolants such water and ethylene glycol. Water based mono
In order to overcome the limitation of the coolants’ low thermal nanofluids such as aluminium oxide (Al2O3), zirconium oxide (ZrO2)

* Corresponding author at: College of Engineering, Universiti Tenaga Nasional, Jalan IKRAM-UNITEN, 43000 Selangor, Malaysia.
E-mail addresses: neeshoun@uniten.edu.my (N. Asokan), prem@uniten.edu.my (P. Gunnasegaran), vignesh@uniten.edu.my (V. Vicki Wanatasanappan).

https://doi.org/10.1016/j.tsep.2020.100727
Received 12 August 2020; Received in revised form 13 September 2020; Accepted 14 September 2020
Available online 19 September 2020
2451-9049/© 2020 Elsevier Ltd. All rights reserved.
N. Asokan et al. Thermal Science and Engineering Progress 20 (2020) 100727

and titanium oxide (TiO2) have shown to exhibit better heat transfer 2. Materials and methods
coefficients when compared to distilled water. In some cases, when
pumping low concentration amount of copper oxide (CuO) nano­ 2.1. Experimental test rig and setup
particles into insulated copper tubes, it was found that as the heat
transfer coefficient increases with the Reynold’s number and the nano­ Fig. 1 shows the experimental test rig used for the experimentation
particles concentration of the fluid [4–6]. purposes of this research while Fig. 2 the schematic diagram of the test
Studies also indicated that not all mono nanofluids possess the same rig setup. The experimental set up consist of a storage tank, a pump, a
level of advantage in terms of high heat transfer capability. Under heater, a cross flow finned heat exchanger (car radiator), a forced draft
constant concentration values, water based CuO mono nanofluid possess fan, 3 resistance temperature detector (RTDs), a flowmeter, control
a higher heat transfer ratio when compared to TiO2 nanofluid. The main panel and power supply. The coloured lines indicate the heated working
issue with using mono nanofluid is the sacrifice between either good fluid that flows from the reservoir tank to the radiator while the blue
thermal properties (such as thermal conductivity and specific heat ca­ coloured lines indicate the cooled working fluid flowing from the radi­
pacity) or material properties (such as viscosity and density). All the ator back into the reservoir tank. The green lines indicate the electrical
desired characteristics tuned for a specific application are not possessed wirings of the entire experimental setup. Table 1 shows the dimensions
by any one mono nanoparticle. For example, Al2O3 nanoparticles are not of the car radiator and the air speed produced by the draft fan (Fig. 1).
chemically reactive and therefore stable, however has a lower thermal The control panel displays the temperature of the fluid in the storage
conductivity value compared copper and silver which, even though, tank, the temperature of the fluid at the inlet of the car radiator, and the
have high thermal conductivity values are highly reactive and unstable. temperature of the fluid at the outlet of the car radiator. The control
In order to solve this conundrum, the combination of two or more panel also allow to manually set the maximum temperature that fluid
nanoparticles together with a base fluid should produce a hybrid ma­ should reach in the storage tank and the controls the speed of the pump.
terial that possess the good characteristics of both the individual mono Table 2 shows the precision and uncertainty values of the detectors and
nanoparticles. These combinations are known as hybrid nanofluids meter indicators used in recording the necessary parameters such as
[7,8]. temperature readings, flowrate and air velocity.
Previous research have indicated that through the selective hybrid­ An initial maximum temperature of the storage tank and pump speed
izing of two different nanoparticles, the heat transfer rate possessed by is set before switching on the heater. At this moment, the pump is
the hybrid nanofluid is higher when compared to its mono nanoparticle switched off to allow the nanofluid sample inside the tank to heat up and
counterparts. Under varying concentrations with water acting as the reach the intended maximum temperature. Once the maximum tem­
base fluid, various combination of Al2O3-Cu, Al2O3-Ag, Al2O3 + multi- perature has been reached, the pump is switched on and the nanofluid
walled carbon nanotubes (MWCNT) and Al2O3-TiO2 mono nanoparticles sample will be pumped into the car radiator. The fluid that comes fluid
were exhibit higher heat transfer performance when compared to its exits the car radiator will return to the storage tank and the cycle repeats
mono nanofluid counterparts and also pure water. As the hybrid nano­ itself. The temperature readings were being recorded throughout this
composite volume concentration increases, the thermal conductivity process at a set time interval over a fix period.
and also the convective heat transfer coefficient of the fluid also in­ The experimental investigation is carried out by varying two factors
creases. The calculated Re number and Nu/Pr ratio were found to be which are the initial maximum temperature and the pump speed. The
directly proportional to one another. It was found that the combination maximum temperature that is used in the investigation is 100 ◦ C, while
of Al2O3-Cu and Al2O3-Ag with water shows the highest amount of heat the pump speed is set at 1050 RPM, 850 RPM and 650 RPM. This range
transfer coefficient values. Both the mass flow rate reductions’ and also of pump speed is selected because after undergoing a process of trial and
overall heat transfer coefficients’ values tend to increase when the error, it is found that any fluid flow under 650 RPM is not smooth and
particle volume concentration increases [9–12]. inconsistent and any fluid flow above 1050 RPM causes the temperature
Hybrid nanofluids can be formed by dispersing different nano­ reading to quickly fluctuate and therefore unable to be precisely
particles in a base fluid through single or two step method. The two-step observed and recorded. Each session of the experiment is conducted
method is more favorable due to being cost effective and the ability to over a period of 20 min and the temperature readings observed from the
prepare hybrid nanofluids on a larger scale. In some cases, during the control panel is recorded every 1-minute interval. The 20 min time
preparation stage of nanofluids the addition of surfactant/dispersant period is selected in order to prevent the heater from overheating as per
helps to stabilize the nanofluids by affecting surface characteristics of instructed by the test rig manufacturers. The flowrate is also recorded in
mixture. The surface tension between the contacts of the two material order to calculate the fluid velocity and Reynolds number of the fluid
can be reduced in order to achieve longer nanoparticles’ suspension flow within each tube. Eqs. (1)–(3) is used to calculate the mentioned
time. Research have been done, focusing on the effects on thermal values [15].
performance by adding surfactants such as Sodium Dodecyl Sulphate
4Atube
(SDS) and Polyvinyl Pyrrolidone (PVP) into preparing water-based Dh = (1)
Ptube
hybrid nanofluids. At fixed volume concentrations, the weight per­
centage of the surfactants added varied from 0.1% to 0.4%. The results where:
indicate that through addition of these surfactants, the heat transfer
performance are higher when compared to distilled water. In some Dh = Hydraulic diameter (m).
cases, when the weight percentage of PVP increases, the rate of heat Atube = Cross-sectional area of tube (m2).
transfer occurrence also increases [13,14]. Ptube = Perimeter of tube (m)
The present work aims to study both the fluid and thermal proper­
ties, at low volume concentrations of the prepared Al2O3-CuO hybrid V̇
υfluid = (2)
samples and compared the findings with both Al2O3 and CuO mono A×ηtube
nanofluids under similar volume concentrations. This research also aims
to experimentally study the changes in terms of thermal performance where:
such as heat transfer coefficients of both hybrid and mono nanofluids
under constant low volume concentrations using a compact heat υfluid = Fluid velocity (m/s).
exchanger test rig. V˙ = Flowrate (m3/s).
ηtube = Number of tubes

2
N. Asokan et al. Thermal Science and Engineering Progress 20 (2020) 100727

Fig. 1. Experimental Test Rig Setup.

ρfluid υfluid Dhydraullic where:

Refluid = (3)
μfluid
Nuexp = Experimental Nusselt number.
where: hexp = Experimental heat transfer coefficient (W/m2.◦ C).
Dh = Hydraulic diameter (m).
Refluid = Reynolds Number. k = Thermal conductivity (W/m.◦ C).
ρfluid = Density of fluid (kg/m3).
υfluid = Fluid velocity (m/s). The experimental Nusselt number and heat transfer coefficient
Dh = Hydraulic diameter (m). values are validated using the respective theoretical values. The theo­
μfluid = Viscosity of fluid (kg/m.s) retical Nusselt number values are calculated using Shah-London corre­
lation equation for laminar flow (Eq. (7)). The theoretical heat transfer
Due to the fixed values of the pump speeds, the flowrate values coefficient values are calculated using Eq. (8) [17].
recorded are 34.5 l/min, 32.5 l/min and 30 l/min for 1050 RPM, ( )1 ( )
850RPM and 650 RPM respectively. The average experimental heat Dh 3 Dh
Nutheoretical = 1.953 RePr for RePr ≥ 33.33 (7a)
transfer rate and heat transfer coefficient is calculated using Eq. (4) and L L
Eq. (5), respectively. The average experimental Nusselt Number is ( ) ( )〈
calculated using Eq. (6) as shown below [16]. Nutheoretical = 4.364 + 0.0722 RePr
Dh
for RePr
Dh
33.33 (7b)
L L
Q = ρV̇Cp (Tin − Tout ) (4)
where:
where:
Nutheoretica = Theoretical Nusselt number.
Q = Heat transfer rate (W). Re = Reynold’s Number
ρ = Density of fluid (kg/m3). Pr = Prandtl Number
V˙ = Flowrate (m3/s). Dh = Hydraulic diameter (m).
Cp = Specific heat capacity (J/◦ C.kg) L = Length of tube (m).
Tin = Inlet temperature (◦ C)
Tout = Outlet temperature (◦ C) Nutheoretical k
htheoretical = (8)
Dh
Q
hexp = (5)
As (Tin − Tout ) where:

where: htheoretical = Theoretical heat transfer coefficient (W/m2.◦ C).

Nutheoretical = Theoretical Nusselt number.
hexp = Experimental heat transfer coefficient (W/m2.◦ C). Dh = Hydraulic diameter (m).
Q = Heat transfer rate (W). k = Thermal conductivity (W/m.◦ C).
As = Surface area (m2).
Tin = Inlet temperature (◦ C)
Tout = Outlet temperature (◦ C) 2.2. Preparation of hybrid nanofluids
hexp Dhydraullic
Nuexp = (6) The present research is focused on the effects of three types of
k
working fluid on the compact heat exchanger test rig. These fluids are
water + ethylene glycol, mono nanofluids and hybrid nanofluid. The

3
N. Asokan et al. Thermal Science and Engineering Progress 20 (2020) 100727

RTD
Inlet

Radiator
+
Draft Fan

RTD Flow Meter

Outlet Indicator

RTD Pump
Tank
Heater

Heated Fluid Flow

Reservoir Tank
(Containing Working Fluid)
Cooled Fluid Flow

Electrical Wiring

Fluid Flow
Power Supply

Control Panel

Fig. 2. Schematic diagram of experimental setup.

Table 1 Table 2
Dimensions of Heat Exchanger Test Rig. Technical precision and uncertainty values of measuring device.
Descriptions Dimension Measuring Device Measuring Precision Uncertainty
Range
Tube Length 0.386 m
Tube Height 0.016 m Flow Meter 2–40 L/minute ±0.5 L/ ±0.17 L/min
Tube Width 0.002 m minute
Distance between Tubes 0.010 m Resistance Temperature 0–120 ◦ C ±1◦ C ±0.035 ◦ C
Number of Tubes 70 Detector
Fin Thickness 0.001 m Anemometer (Wind 0–20 m/s ±0.05 m/s ±0.007493 m/
Fin Length 0.010 m Velocity) s
Distance between Fins 0.005 m
Air Velocity 6.67 m/s

where:
combination of nanoparticles selected for the hybrid nanofluid prepa­ ∅ = Volume concentration (%).
ration are aluminium oxide and copper oxide that will be dispersed in a mnp = Mass of nanoparticles (g)
water/ethylene glycol (60:40) based mixture fluid. The mixing ratio of ρnp = Density of nanoparticles (kg/m3).
the nanoparticles have been fixed at 50:50 ratio. The hybrid nanofluid Vbase fluid = Volume of base fluid (m3)
samples are prepared in three variation of volume concentrations which The volume of each sample prepared is 8 L as the experimental test
are 0.02%, 0.04% and 0.06% using Eq. (9) [18]. rig requires a minimum of 6 L in order to operate. The hybrid nanofluid
samples are prepared in a two-step method. The nanoparticles are pur­
mnp1 mnp2
+ chased from Sigma-Aldrich. The properties of the purchased nano­
(9)
ρnp1 ρnp2
∅ = mnp1 mnp2 particles are shown in Table 3. The base mixture containing water and
ρnp1 + ρnp2 + Vbasefluid
ethylene glycol (60:40) is also studied in order to determine the

4
N. Asokan et al. Thermal Science and Engineering Progress 20 (2020) 100727

Table 3 allowed to dry. Once the sample is dried, moved to the machine.
Properties of mono nanoparticles.
Descriptions Aluminium Oxide Copper Oxide 2.3.2. Particle size analyzer testing
Particle size analyser depends on a well-known theory known as
Purity 99.8% 99.8%
Primary Particle Size 13 nm <50 nm “Brownian motion”. The theory discusses regarding the movement of
Density 3.95 g/cm3 6.31 g/cm3 particles in a liquid that is solely affected by repetitive bombardment
Molar Weight 101.96 g/mol 79.56 g/mol from its surrounding molecules that bring about the movement f the
particles. These particles in the liquid move about at random and uti­
lizing this theory allows comprehension of size of the particles. The
mixture’s fluid properties.
smaller particles diffuse rather quickly compared to their larger sized
In order to prevent agglomeration from occurring during the sample
counterparts due to larger surface area covered by the smaller sized
preparation stage, a surfactant known as Polyvinyl Pyrrolidone (PVP) is
particles. Thus, by photographing two pictures within a short span of
added. The ratio of PVP added is 1:10 of the total weight of nano­
time, the distance that the particle has propelled can be investigated.
particles added to the base fluid mixture [19]. The hybrid nanofluid
The movement of the heavier and larger sized particle in theory will be
mixture is first undergone magnetic stirring for two hours. The sample is
minimal comparatively with the lighter and smaller sized peer. The
then in an ice bath and an ultrasonicator probe is dipped into the so­
classification of the particle would enable subsequent deeper analysis of
lution. Then, the sample undergoes ultrasonication for 90 min at fre­
the particles be done. 1.5 ml of the 0.02% hybrid sample is required for
quency of 75 Hz. The sonificator probe is set to sonicate every 15 s
the purpose of this testing. The sample is placed into a cuvette and
followed by a 5 s break interval. Throughout this process the ice bath
sonicated for 1 min before being slotted into the Malvern Nano ZS ma­
must be changed and refilled in order to prevent the solution from
chine for analysis. The test is conducted 3 times in order to obtain an
overheating. Overheating inhibits the PVP from preventing agglomera­
average reading value.
tion from occurring.
2.3.3. Zeta potential analysis
Zeta potential is primarily used to obtain an indication of the sta­
2.3. Characteristics of hybrid nanofluid samples
bility of a colloidal system. It allows to obtain an overview of how the
interaction between the ions and the particles. Fluids contains ions such
A small amount of each of varying concentration of hybrid nanofluid
as cations and anions. The fluid movements are largely due to existence
samples are tested upon in order to determine the fluid properties such
of gravitational forces. The particles have two different layers sur­
as density, thermal conductivity, specific heat capacity and viscosity.
rounding them. Stern layer is closely bounded distance from the particle
The properties of the Al2O3-CuO hybrid nanofluid samples will be
whereas the diffuse layer is outside the stern layer. Within stern layer,
compared with the corresponding properties of both Al2O3 and CuO
the ions interact strongly with the particle comparatively with diffuse
mono nanofluid and also obtained from previous study as shown in
layer. A boundary exists within the diffuse layer itself in which ions and
Table 4 [20].
the particle comes together to form a stable entity. This boundary is
referred to as hydrodynamic shear. The movement of fluid forces the
2.3.1. Transmission electron microscopy (TEM) image testing
particle to move thus causing ions that surround the particle to subse­
TEM otherwise known as transmission electron microscope is used to
quently move in tandem with the particle. Movement of the ions in and
view specimens that are not visible to the naked eye (tissue sections,
out of the hydrodynamic shear creates a potential which could be
molecules. etc) where an image is projected through the electrons
measured.
passing by. TEM works on the same general concepts as light micro­
It is a general consensus that a value in between +30 mV and − 30
scopes, which uses electrons instead of photons. As the wavelength of
mV is considered to be unstable and any other value away from the said
electrons is far less than that of light, the optimum quality of the TEM
numerical number is considered to be stable. Prior statement is aligned
photos is much higher than that of the light microscope. TEM offers
with research made previously. Previous studies have indicated that any
high-resolution imaging and is used primarily for the analysis of tiny
values under the numerical values of 30 mV is considered unacceptable
areas or even single mineral platelets. The comparison of the TEM im­
stability, values ranging in between 30 mV and 45 mV is considered
ages relies on many aspects, including the variation in electron density
moderately stability and any values above 45 mV s considered to be of
of the materials of the sample (variation in the organic matrix and
excellent stability, meanwhile a value of above 30 mV is considered to
mineral electron density), the thickness and width of the sample, the
be of good stability of the colloidal system and any value beneath 20 mV
diameter of the analytical aperture, and the strength of the electrons.
is to be considered of poor stability. Even though that is the acceptable
TEM is used to collect structural and compositional knowledge that are
range, it must be taken as a hindsight that a numerical value with an
not be obtained using an appropriate approach. In action, TEM is seldom
enormously positive or negative value will bring negative repercussions.
used on its own and is usually part of an established diagnostic pro­
At these values, the particles will not come together as each particle has
cedure [21,22].
a high tendency to repel each other eliminating the intended purpose of
The performance of nanocomposites depends on the dispersion of
the experiment [23,24]. Similar to the particle size analyser testing, 1.5
nanofilms. TEM is the process used to calculate dispersion where the
ml of the nanofluid samples (both mono and hybrid) is placed into
hybrid solution is taken under 20 microlite and dissolved in 1.5 ml of
capillary cell and sonicated for 1 min before being slotted into a Malvern
ethanol. The ultra-thin tissue parts were placed on the copper grid and

Table 4
Fluid and thermal properties of working fluids.
Volume Concentration (%)

Al2O3 CuO

Properties (Base Fluid) 0.02 0.04 0.06 0.02 0.04 0.06

Density (kg/m3) 1084 1134 1184 1235 1192 1300 1409
Specific.Heat Capacity (J/kg.K) 3184 2997 2859 2731 2863 2625 2424
Thermal Conductivity (W/m.K) 0.3729 0.4203 0.4423 0.4657 0.4243 0.448 0.4721
Viscosity (kg/m.s) 0.00423 0.00539 0.00699 0.00906 0.00615 0.00971 0.01534

5
N. Asokan et al. Thermal Science and Engineering Progress 20 (2020) 100727

Nano ZS for testing. Each sample was tested 3 times in order to obtain an capacities of the calorimeter and pan on which the samples are placed
average zeta potential value. upon are ignored, the initial measured temperature is assumed as the
sample’s temperature and zero heat loss to surroundings.
2.3.4. Density Several models have been derived in order to calculate the theoret­
Properties such as pressure losses, fluid friction factor and Reynolds ical specific heat capacity value of nanofluids. The assumption made in
number are dependent on the density of fluid. In order to theoretically all these models is that, between the base fluid and the nanoparticles,
calculate the density, specific heat capacity and thermal conductivity of thermal equilibrium is present. The models used in the present work are
the nanofluids, one needs to first calculate the corresponding properties Khanafer and Vafai’s model as shown in Eq. (14), Pak and Cho’s model
of the nanoparticles that are being dispersed within the base fluid. The as shown in Eq. (15), and the model presented in Shekar and Sharma’s
mentioned properties of both Al2O3 and CuO nanoparticles combined is research as shown in Eq. (16) [26–28]. In all these models, the specific
calculated using Eq. (10) to Eq. (12) [25]. The values obtained from Eq. heat capacity of the nanofluid are dependent specific heat capacity of
(10) are then used in Eq. (13) in order to obtain the theoretical density the nanoparticles. This parameter can be calculated using Eq. (11).
value of the hybrid nanofluid samples [26].
ρbf .(1 − ∅) ρnp .∅
Cpnf = .Cpbf + .C (14)
φ1 ρnp1 + φ2 ρnp2 ρnf ρnf pnp
ρnp = (10)
∅
where:
where:
Cpnf = Specific heat capacity of nanofluid (J/◦ C.kg).
φ = Mixing ratio (%). Cpbf = Specific heat capacity of base fluid (J/◦ C.kg).
∅ = Volume concentration (%). Cpnp = Specific heat capacity of nanoparticles (J/◦ C.kg).
ρnp = Density of nanoparticles (kg/m3).
ρbf = Density of base fluid (kg/m3)
φ1 ρnp1 Cpnp1 + φ2 ρnp2 Cpnp2 ρnp = Density of nanoparticles (kg/m3)
Cpnp = (11)
ρnp ∅ ρnf = Density of nanofluid (kg/m3)
∅ = Volume concentration (%).
where:
ρnf = ρbf (1 − ∅) + ρnp ∅ (15)
Cpnp = Specific heat capacity of nanoparticles (J/ C.kg).
◦
where:
φ = Mixing ratio (%). ρnf = Density of nanoparticles (kg/m3).
∅ = Volume concentration (%).
ρnp = Density of nanoparticles (kg/m3). ρbf = Density of base fluid (kg/m3)
φ knp1 + φ2 knp2 ρnp = Density of nanoparticles (kg/m3)
knp = 1 (12)
∅ ∅ = Volume concentration (%).

where: Cpnf =
Cpbf ρbf .(1 − ∅) + ∅Cpnp ρnp
(16)
(1 − ∅)ρbf + ∅ρnp
knp = Thermal conductivity of nanoparticles (W/m.◦ C).
φ = Mixing ratio (%). where:
∅ = Volume concentration (%).
Cpnf CPnf = Specific heat capacity of nanofluid (J/◦ C.kg).
ρnf = ρbf (1 − ∅) + ρnp ∅ (13) Cpnf CPf = Specific heat capacity of base fluid (J/◦ C.kg).
Cpnp Cpnf CPnf = Specific heat capacity of nanoparticles (J/◦ C.kg).
where:
ρbf = Density of base fluid (kg/m3)
ρnf = Density of nanoparticles (kg/m3). ρnp = Density of nanoparticles (kg/m3)
ρbf = Density of base fluid (kg/m3) ρnf = Density of nanofluid (kg/m3)
φ = Mixing ratio (%). ∅ = Volume concentration (%).
∅ = Volume concentration (%).
2.3.6. Viscosity
The experimental values of aforementioned property are measured Viscosity is described as the resistance of one fluid layer against
using a Rudolf density meter DDM 2909. The equipment is used in order another layer of either fluid or solid [29]. Therefore, in theory, the
to scale up the readings on a U-tube up from 2 to 10 times magnification higher the viscosity of a fluid, the higher the amount of pump power
in order to obtain a clearer reading. A minimum of 10 ml of the sample is needed in order to overcome the resistance force and push the fluid
needed in order to perform the testing. throughout an entire mechanism. The viscosity of all the hybrid nano­
fluid samples are measured using an In-House Method based on
2.3.5. Specific heat capacity Brookfield LV-DVII. The speed of which the samples are rotated during
The specific heat capacity of a fluid determines the amount of heat this experimentation ranges from 126 rpm to 135 rpm.
absorbed in order to increase the temperature of 1 g of the fluid by 1 ◦ C. Over the years, many researches have been done in order to deter­
This is an important aspect of nanofluids that are used as coolants, as it mine mathematical model which in turn aid in the predicting the dy­
influences the heat transfer rate flowing in and out of the fluid. The namic viscosities of nanofluids. These researches include Einstein’s
specific heat capacity of the hybrid nanofluid samples are measured model, as shown in Eq. (17), for determining the viscosity of nanofluids
using Differential Scanning Calorimetry (DSC) method. The model used which contains nanoparticle concentration of not more than 4% [30].
for this measurement is DSC Q20 by TA Instruments. The DSC method μnf = μbf (1 + 2.5∅) (17)
measured the heat flow rate difference between the samples and
chemically non-reactive references over the function of time and tem­ where:
perature. The assumptions made during this testing are the heat

6
N. Asokan et al. Thermal Science and Engineering Progress 20 (2020) 100727

Cpnf μnf = Viscosity of nanofluid (kg/m.s). Cpnf knf = Thermal conductivity of nanofluid (W/m.◦ C).
μbf = Viscosity of base fluid (kg/m.s). Cpnf kbf = Thermal conductivity of base fluid (W/m.◦ C).
∅ = Volume concentration (%). ∅= Volume concentration (%).

The Brickman’s model, as shown in Eq. (18), and the Batchelor 3. Results and discussion
model, as shown in Eq. (19), are also used as references in order to
determine the theoretical values of nanofluids viscosity [31,32]. 3.1. Rheological and thermal properties of hybrid and mono nanofluids
1
μnf = μbf (18) 3.1.1. Transmission electron microscopy (TEM) image testing
(1 − ∅)2.5
The TEM images of Al2O3-CuO nanofluid sample is shown in Fig. 3.
The black spots on the image are copper oxide nanoparticles while the
where:
white spots are aluminium oxide nanoparticles. The image also reveals
that the sample remained intact as it dispersed into the 1.5 ml ethanol.
Cpnf μnf = Viscosity of nanofluid (kg/m.s).
Signs of aggregation and agglomeration is observed during the disper­
μbf = Viscosity of base fluid (kg/m.s).
sion and during drying on the grid. However due to the presence of
∅ = Volume concentration (%).
ultrasonification and also PVP acting as a stabilizing agent in the prep­
μnf = μbf (1 + 2.5∅ + 6.5∅2 ) (19) aration stage of the nanofluids, these signs are reduced. Based on the
TEM images observed, the dispersion is diluted as much as possible, as
where: this minimizes the agglomeration during drying on the grid. It can be
found that both aluminium oxide and copper oxide occur in agglomer­
Cpnf μnf = Viscosity of nanofluid (kg/m.s). ated form and do not reside as an independent individual structure. This
μbf = Viscosity of base fluid (kg/m.s). dictates the relationship of aluminium oxide and copper oxide combined
∅ = Volume concentration (%). as one hybrid nanofluid.

2.3.7. Thermal conductivity 3.1.2. Particle size analyzer testing

Thermal conductivity is defined as the amount of time taken for heat Table 5 shows the tabulated results obtained from the particle size
to be transferred via conduction through one-unit area of material. A analyser testing. Based on Table 5, it can be clearly be seen that the
material that possess a high thermal conductivity value shows that the average of about 1152 nm with a peak at about 1111 nm. This shows
material is a good conductor of heat. The thermal conductivity of the that the colloidal stability of the system was greatly impacted and the
hybrid nanofluid samples are tested using KD2Pro. The KD2Pro contains zeta potential of +31.81 mV that was obtained displays as a proof to the
dual TR-1 needle like sensors that are inserted 1.5 cm parallel to each claim. The polydispersity index obtained was 0.589 which is a reason­
other into the samples. The sensor must be initially calibrated by testing ably acceptable value. Polydispersity index (PDI) refers to the degree of
on a standard fluid which in this case is glycerine. This is to ensure that non-uniformity of particles within a liquid. It gives an insight to the sizes
the temperature of the sensor is equal to the surrounding temperature of particles that exists within the tested sample. Previous study has
before proceeding with the sample testing. investigated the impact of particle size and polydispersity index. From
Various models have been published for the purpose of calculating the research done, the researchers deduced an appropriate value should
the theoretical thermal conductivity values. One of the earliest of models be within the range of 0.05–0.7. It is apt to conclude that the result
developed is by Maxwell as shown in Eq. (20). Other models that are also obtained is in compliant with the said range from the previous research
used as reference for the present work are Bruggeman’s model as shown [36].
in Eq. (21a) and Eq. (21b) and also the model proposed by Hamilton and
Crosser as shown in Eq. (22) [33–35]. 3.1.3. Zeta potential analysis
The results obtained from the particle size analyser testing show that
knp + 2kbf + 2(knp − kbf )∅
knf = kbf (20) the colloidal stability of the system was greatly impacted and the zeta
knp + 2kbf − (knp − kbf )∅
potential of +31.8 mV that was obtained displays as a proof to the claim.
In aligning with the past research papers, copper oxide as well as
where:
aluminium oxide too were in compliant with the range stated above. The
copper oxide has a zeta potential value of − 36.7 mV. aluminium oxide
Cpnf knf = Thermal conductivity of nanofluid (W/m.◦ C).
on the other hand has a zeta potential value of 38.5 mV. The contra­
Cpnf kbf = Thermal conductivity of base fluid (W/m.◦ C).
dicting polar sign is also a positive primitive method to forecast the
Cpnf knp = Thermal conductivity of nanoparticles (W/m.◦ C).
binding of both elements together. Binding of each element is pivotal in
∅ = Volume concentration (%). order to take advantage of the characteristics of both elements subse­
[ √̅̅̅̅ ] quently improving on the heat exchanged between the fluid and the
knf = 0.25 (3∅ − 1)knp + (2 − 3∅)kbf + Δ (21a)
system. Besides that, the zeta potential also showcases the well dispersed
[ ]2 particles in the suspension whilst the presence of any large flocculation
Δ = (3∅ − 1)knp + (2 − 3∅)kbf + 8knp .kbf (21b) cannot be observed proves as a positive merit to the present work. In
brief, the nanofluid was obtained was optimal for the research to be
where: done. As indicated earlier, Table 6 shows the zeta potential values of
each time the samples were tested.
Cpnf knf = Thermal conductivity of nanofluid (W/m.◦ C).
Cpnf kbf = Thermal conductivity of base fluid (W/m.◦ C). 3.1.4. Density
Cpnf knp = Thermal conductivity of nanoparticles (W/m.◦ C). Fig. 4 shows the density of the prepared Al2O3-CuO hybrid nanofluid
∅ = Volume concentration (%). samples against varying volume concentrations. As volume concentra­
( ) tion increases, the density of the samples also increases. This is due to
knf = 4.97∅2 + 2.72∅ + 1 kbf (22)
the increases in the amount of nanoparticles that are being suspended
within the base fluid. Fig. 5 shows the comparison of the density values
where:

7
N. Asokan et al. Thermal Science and Engineering Progress 20 (2020) 100727

Fig. 3. TEM Imaging of Al2O3-CuO nanofluid (0.02% volume concentration) (a) 100 nm (b) 20 nm.

Table 5 1500
Particle size analyzer results for hybrid nanofluid sample (0.02% vol.
concentration). Madhesh et. al [13]
1400
Parameter Testing 1 Testing 2 Testing 3 Average Present Work
Z-Average (d.nm) 1048 1132 1376 1132
Density, kg/m3
Polydispersity Index (PDI) 0.630 0.691 0.446 0.589 1300

1200
Table 6
Zeta potential values of each sample testing.
1100
Zeta Potential (mV)
Samples (0.02%) Testing 1 Testing 2 Testing 3 Average
1000
Al2O3 37.7 38.4 39.3 38.5
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
CuO − 38.9 − 35.8 − 35.5 − 36.7
Al2O3-CuO 33.7 28.8 32.9 31.8 Volume Concentration, %

Fig. 5. Comparison between calculated and present work density values of

hybrid nanofluid samples.
1600.0

1400.0
1500
1200.0
CuO
Density, kg/m3

1000.0 1400 Al₂O₃

800.0 Al₂O₃-CuO
1300
Density, kg/m3

600.0

400.0
1200
200.0

0.0
1100
0.02 0.04 0.06
Volume Concentration, %
1000
Fig. 4. Density of Al2O3-CuO hybrid nanofluid samples against varying volume 0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
concentrations.
Volume Concentration, %

from the prepared sample against the calculated theoretical values. The Fig. 6. Density of hybrid and mono nanofluids.
percentage difference between the measured density is within the
acceptable range. density values of Fe3O4-Graphene hybrid nanofluid are in between the
Fig. 6 shows the comparison between the density of the Al2O3-CuO density values of Fe3O4 and graphene mono nanofluids [37].
hybrid nanofluids with the Al2O3 and CuO mono nanofluids at various
volume concentrations of 0.02%, 0.04% and 0.06%. Based on the 3.1.5. Viscosity
mentioned figure, the density value of Al2O3-CuO hybrid nanofluid is in The results from experimental study on displaying the volume con­
between both Al2O3 and CuO mono nanofluids at volume concentrations centration against viscosity of the Al2O3-CuO nanofluid samples are
of 0.04% and 0.06% and only slightly higher at a volume concentration shown in Fig. 7. Fig. 8 shows the theoretical values are calculated from
of 0.02%. Similar finding is reported by another research, in which the correlations developed by other researchers such as Einstein, Brinkman

8
N. Asokan et al. Thermal Science and Engineering Progress 20 (2020) 100727

0.007 and 2% to 10% when compared with the Batchelor’s model.

Fig. 9 shows the comparison between the viscosity of the Al2O3-CuO
0.006 samples with the viscosity of both mono Al2O3 and CuO nanofluid.
Based on the mentioned figure, the viscosity of the hybrid nanofluid
0.005
samples are lower compared to the mono nanofluids. Similar findings
Viscosity, Pa.s

0.004 are reported from previous studies which compared the viscosity of
hybrid nanofluid with the respective mono counterparts. These re­
0.003 searches compared hybrid nanofluid combinations such as silica-
multiwall carbon nano tubes (MWCNT) and Al2O3-Cu with mono
0.002
nanofluid made up of silica, MWNCT, TiO2, Al2O3, ZnO, and CeO2
0.001 nanoparticles [38,39].

0 3.1.6. Specific heat capacity

0.02 0.04 0.06 Fig. 10 shows the specific heat capacity of the hybrid samples against
Volume Concentration, % vary concentrations. It is observed that, as the volume concentration
increases, the specific heat capacity of the samples decreases. This shows
Fig. 7. Viscosity of Al2O3-CuO hybrid nanofluid samples against varying vol­
that to increase the rate of which the nanofluid gains and losses heat, the
ume concentrations.
nanoparticle concentration within the base fluid should be increased.
This relationship between nanoparticle volume concentration and spe­
0.006 cific heat capacity determined from the present work are similar with
Einstein, Al₂O₃-CuO
the findings from other researchers [40–42].
Brinkman, Al₂O₃-CuO
As the Fig. 11 shows the comparison between the experimental
0.0055 Batchelor, Al₂O₃-CuO
specific heat capacity and their respective calculated values. Based on
Present Work
the figure, it can be seen that the specific heat capacity values calculated
using Pak and Cho’s model greatly deviates from the experimental value
Viscosity, Pa.s

of the present work. The values calculated from the other two models
0.005
also shows deviation, but within an acceptable margin ranging from 2%
to 14% difference. This finding is also similar with the results obtained
from a previous study, that showed specific heat capacity values
0.0045
calculated using Pak and Cho’s model for hybrid nanofluids showed
greater deviation with the experimental model [43]. In general, these
models do not account from the presence of nanolayers between nano­
0.004
particles and base fluid which causes the observed deviations.
0.01 0.02 0.03 0.04 0.05 0.06 0.07
Fig. 12 shows the comparison between the specific heat capacity
Volume Concentration, % values of Al2O3-CuO hybrid nanofluids, Al2O3 and CuO mono nano­
Fig. 8. Comparison between calculated and experimental values for the vis­
fluids. It is observed that the hybrid nanofluid samples possessed a lower
cosity of Al2O3-CuO hybrid nanofluid. specific heat capacity value compared to the other two mono nanofluids
across all three varying volume concentrations, This means that the
hybrid nanofluid sample will be able to gain and loss heat at a much
and Batchelor are also displayed on the mentioned figure. Based on these
faster rate compared to its mono nanofluid counterparts. Similar find­
correlations, it can be seen the two main parameters that determine the
ings, in which the hybrid nanofluid exhibits a lower specific heat ca­
viscosity of a nanofluid are the viscosity of the base fluid and the volume
pacity when compared to mono nanofluids, have been reported by other
concentration of the nanoparticles. As shown in Fig. 8, as the volume of
research works [44,45].
nanoparticle concentration increases, the viscosity of the nanofluid also
increases. Although there are deviations between the experimental
3.1.7. Thermal conductivity
value and the theoretical values, however these deviations are within
Fig. 13 shows the thermal conductivity of hybrid samples against
the acceptable margin ranging from 2% to 13% when compared with
vary concentrations. It is observed that, as the volume concentration
Einstein’s model, 2% to 11% when compared with Brinkman’s model
increases, the thermal conductivity of the samples increases. The in­
crease in the quantity of nanoparticles suspended within the base fluid
0.018
CuO
2600
0.015 Al₂O₃
Al₂O₃-CuO 2580
Specific Heat Capacity, J/kg.

0.012 2560
Viscosity, Pa.s

2540
0.009
2520

0.006 2500
2480
0.003
2460

0 2440
0.02 0.04 0.06
0 0.01 0.02 0.03 0.04 0.05 0.06 0.07
Volume Concentration, %
Volume Concentration, %
Fig. 10. Specific heat capacity of Al2O3-CuO hybrid nanofluid samples against
Fig. 9. Viscosity of hybrid and mono nanofluids. varying volume concentrations.

9
N. Asokan et al. Thermal Science and Engineering Progress 20 (2020) 100727

3750 0.6
Khanafer and Vafai Maxwell
3500 Pak and Cho
Bruggeman
Specific Heat Capacity, J/kg.

Shekar and Sharma 0.55

Thermal Conductivity, W/m.K

3250 Present Work Hamilton and Crosser
0.5 Present Work
3000

2750 0.45

2500
0.4
2250
0.35
2000
0 0.02 0.04 0.06 0.08
0.3
Volume Concentration of Al2O3-CuO nanofluid, % 0 0.02 0.04 0.06 0.08
Volume Concentration of Al2O3-CuO nanofluid, %
Fig. 11. Comparison between calculated and experimental values for the spe­
cific heat capacity of Al2O3-CuO hybrid nanofluid.
Fig. 14. Comparison between calculated and experimental values for the
thermal conductivity of Al2O3-CuO hybrid nanofluid.

3500
Al₂O₃ values of thermal conductivity of the hybrid nanofluid samples. When
CuO compared to the calculated values, the experimental thermal conduc­
Specific Heat Capacity, J/kg.

3250
Al₂O₃-CuO
tivity of the hybrid nanofluid samples show deviations ranging between
3000 3% and 13% based on both Maxwell and Bruggeman model while the
maximum deviation calculated when compared to Hamilton and
2750
Crosser’s model is 3%. These deviations are within the acceptable range
of margin.
2500
Fig. 15 shows the comparison between the thermal conductivity
2250 values of the Al2O3-CuO hybrid samples and the respective mono
nanofluid counterparts. The thermal conductivity of Al2O3-CuO samples
2000 are higher by a maximum enhancement of 2.3% when compared to CuO
0 0.02 0.04 0.06 0.08 and 3.6% when compared to Al2O3 mono nanofluids across all three
Volume Concentration of nanofluid, % varying volume concentrations.

Fig. 12. Comparison between the specific heat capacity values of hybrid and
3.2. Experimental results
mono nanofluids.

The present work focused on determining the experimental values of

0.6 heat transfer coefficient and Nusselt numbers of the working fluids. The
experiment initially began by using water and ethylene glycol mixture
0.5 (base fluid) as the working fluid. The results recorded are then compared
Thermal Conductivity, W/m.K

with its respective theoretical values for the purpose of validation.

0.4 Fig. 16 and Fig. 17 shows the comparison between the experimental and
calculated values of the Nusselt number and heat transfer coefficient of
0.3 base fluid mixture against flowrate. Fig. 16 shows that as the flowrate
increases, the Nusselt number of the fluid also increases. This is because
0.2
an increase in flowrate leads to an increase in Reynolds number which in
turn, based on the correlation of Eq. (7a) and (7b), is proportional to the
0.1
Nusselt number of the fluid. The deviations between these Nusselt
number values ranges between 8% and 12% which is within the
0
0.02 0.04 0.06 acceptable margin. Fig. 17 shows a similar trend in which as the flowrate
Volume Concentration, %
0.55
Fig. 13. Thermal conductivity of Al2O3-CuO hybrid nanofluid samples against
Al₂O₃
varying volume concentrations.
Thermal Conductivity, W/m.K

CuO
0.5
Al₂O₃-CuO
also leads to a larger total surface area of the working fluid. Another
possible rationality behind this trend may be due to the increase of
0.45
Brownian movements within the hybrid nanofluid samples. Brownian
movement is described as the random motion of particles suspended
within a fluid. A previous study was conducted specifically regarding the 0.4
effects of Brownian motion on nanofluids’ thermal conductivity values.
The research indicated that an increase in Brownian movement is indi­
0.35
rectly responsible to the nanofluids’ thermal conductivity enhancement.
0 0.02 0.04 0.06 0.08
The deviations between the experimental values and the theoretical
Volume Concentration, %
ones can be explained by the negligence of the Brownian motion particle
effect in these models [46]. Fig. 15. Comparison between thermal conductivity of hybrid and
Fig. 14 shows the comparison between the measured and calculated mono nanofluids.

10
N. Asokan et al. Thermal Science and Engineering Progress 20 (2020) 100727

10.00 can also be seen, that across most variations of volume concentrations,
Kristiawan et. al. [19]
the average Nusselt numbers of the Al2O3-CuO hybrid nanofluid samples
8.00 are higher when compared to the mono nanofluid counterparts and also
Present Work
the base fluid mixture. Nusselt number indicates the ratio of heat
Nusselt Number

6.00
transfer via convection to conduction at the fluid boundary layer. Higher
Nusselt number indicates an increase in active heat transfer via con­
vection. This trend is similar with the findings of other researchers and is
4.00
caused by the high thermal conductivity values of the hybrid compared
to the mono nanofluids [47,48].
2.00 Fig. 19 show the average heat transfer coefficient of the fluids against
flowrate at various volume concentrations. The heat transfer coefficient
0.00 increases with the flowrate. Across all volume concentrations, the heat
29.0 30.0 31.0 32.0 33.0 34.0 35.0 transfer coefficient values of the Al2O3-CuO hybrid nanofluid samples
Flowrate l/min are higher when compared to the mono nanofluid counterparts and the
base fluid mixture. The highest enhancement in terms of heat transfer
Fig. 16. Validation of water + ethylene glycol mixture’s Nusselt number with
Shah-London correlation.
coefficient values when using the hybrid nanofluid samples are 6.7%,
when compared to CuO nanofluids, and 17.9% when compared to Al2O3
nanofluids. When compared with the base fluid mixture, the hybrid
1680 samples show a maximum increase of 23.2%. The addition of a combi­
nation of Al2O3 and CuO nanoparticle into a base mixture of ethylene
Heat Transfer Coeffcient, W/m2.k

1640 Kristiawan et. al [19]

glycol and water (40:60) have shown to improve the overall heat
1600
Present Work transfer performance of the fluid. Similar findings have been reported in
other research works which supports the results and observations of the
1560
present work [49–51].
1520
3.3. Uncertainty analysis
1480

1440 Table 7 shows the experimentally measured fluid and thermal

properties of the hybrid samples. The lowest percentage uncertainty
1400 observed was valued at 0.01% while the highest percentage of uncer­
29.0 30.0 31.0 32.0 33.0 34.0 35.0
tainty observed was 0.5% for the specific heat capacity for the volume
Flowrate l/min
concentration of 0.04% and 0.06%.
Fig. 17. Validation of water + ethylene glycol mixture’s average heat trans­
fer coefficient. 4. Conclusion

increases, the heat transfer coefficient of fluid also increases. The heat In conclusion, the objective of measuring and comparing the fluid
transfer coefficient is proportional towards the Nusselt number of the properties of the hybrid nanofluids such as density, viscosity and specific
fluid, as indicated by Eq. (6) and Eq. (8). The experimental values of heat heat capacity are fulfilled, where the deviations between the experi­
transfer coefficient match this trend and show deviations ranging be­ mental and correlation values are within the acceptable margin.
tween 1% and 4%. Furthermore, the second objective of the present work have also been
Fig. 18 shows the relationship between flowrate and average Nusselt fulfilled in which the thermal performance of both mono and hybrid are
number for the base fluid and both mono and hybrid nanofluids at compared with one another. Based on the findings of the present work,
volume concentration of 0.02%, 0.04% and 0.06% respectively. The the Al2O3-CuO hybrid nanofluid samples prepared is able to show
trend shown in the mentioned figure matches the trend pattern of observable enhancement in terms of thermophysical fluid properties.
Fig. 16, in which both the parameters are proportional to one another. It The hybrid nanofluid samples are able to enhance the thermal conduc­
tivity of the working fluid by 2.3% when compared to CuO nanofluids

8.0 2000
0.02 (Al₂O₃-CuO) 0.02 (CuO)
0.02 (Al₂O₃) 0.04 (Al₂O₃-CuO) 1900 0.02 (Al₂O₃-CuO) 0.02 (CuO)
0.04 (CuO) 0.04 (Al₂O₃) 0.02 (Al₂O₃) 0.04 (Al₂O₃-CuO)
Heat Transfer Coeffcient, W/m 2.k

0.06 (Al₂O₃-CuO) 0.06 (CuO) 1800

0.04 (CuO) 0.04 (Al₂O₃)
Nusselt Number

7.0 0.06 (Al₂O₃) 0.0 (W + EG) 0.06 (Al₂O₃-CuO) 0.06 (CuO)

1700
0.06 (Al₂O₃) 0.0 (W + EG)
1600

1500

6.0 1400

1300

1200

5.0 1100
29.0 30.0 31.0 32.0 33.0 34.0 35.0 29.0 30.0 31.0 32.0 33.0 34.0 35.0
Flowrate, l/min Flowrate, l/min

Fig. 18. Average Nusselt number against flowrate of both mono and Fig. 19. Average heat transfer coefficient against flowrate for both mono and
hybrid samples. hybrid samples.

11
N. Asokan et al. Thermal Science and Engineering Progress 20 (2020) 100727

Table 7 [10] H.R. Allahyar, F. Hormozi, B. ZareNezhad, Experimental investigation on the

Thermophysical properties of the hybrid samples at 30 ◦ C. thermal performance of a coiled heat exchanger using a new hybrid nanofluid, Exp.
Therm Fluid Sci. 76 (2016) 324–329.
Volume Density Viscosity Thermal Specific Heat [11] D. Madhesh, S. Kalaiselvam, Experimental analysis of hybrid nanofluid as a
Concentration (g/cm3) (m.Pa.s) Conductivity, Capacity (j/ coolant, Procedia Eng. 97 (2014) 1667–1675.
(%) (W/m.k) kg.k) [12] S.K. Singh, J. Sarkar, Energy, exergy and economic assessments of shell and tube
condenser using hybrid nanofluid as coolant, Int. Commun. Heat Mass Transf. 98
0.02 1.211 ± 4.350 ± 0.452 ± 0.001 2558 ± 5 (September) (2018) 41–48, https://doi.org/10.1016/j.
0.001 0.001 icheatmasstransfer.2018.08.005.
0.04 1.284 ± 5.006 ± 0.458 ± 0.001 2537 ± 13 [13] Muhammad Usman Sajid, Hafiz Muhammad Ali, Thermal conductivity of hybrid
0.001 0.001 nanofluids: a critical review, Int. J. Heat Mass Transf. 126 (2018) 211–234.
0.06 1.389 ± 5.509 ± 0.475 ± 0.001 2502 ± 13 [14] F. Hormozi, B. ZareNezhad, H.R. Allahyar, An experimental investigation on the
0.001 0.001 effects of surfactants on the thermal performance of hybrid nanofluids in helical
coil heat exchangers, Int. Commun. Heat Mass Transfer 78 (2016) 271–276.
[15] M. Carl, D. Guy, B. Leyendecker, A. Miller, and X. Fan, “THE THEORETICAL AND
EXPERIMENTAL INVESTIGATION OF THE HEAT TRANSFER PROCESS OF AN
and by 3.6% when compared to Al2O3 nanofluids. The use of hybrid AUTOMOBILE RADIATOR 2012 ASEE Gulf Southwest Annual Conference April 4-
nanofluids as a working fluid in compact heat exchanger rig (radiator) 6, El Paso, Texas,” vol. 1, no. 128, pp. 1–12, 2012.
also shows enhancements in terms of Nusselt number and overall heat [16] Adnan M. Hussein, R.A. Bakar, K. Kadirgama, Study of forced convection nanofluid
heat transfer in the automotive cooling system, Case Stud. Thermal Eng. 2 (2014)
transfer coefficient values by 6.7% and 7.2% respectively when 50–61.
compared to CuO nanofluids and by 17.9% and 12.1% respectively [17] B. Kristiawan, B. Santoso, W.E. Juwana, R.M. Ramadhan, I. Riandana, Numerical
when compared to Al2O3 nanofluids. This goes to show, even at low investigation of laminar convective heat transfer for TiO2/water nanofluids using
two-phase mixture model (Eulerian approach), AIP Conf. Proc. 1788 (2017),
concentrations (0.02%–0.06%), Al2O3-CuO hybrid nanofluid does
https://doi.org/10.1063/1.4968255.
perform better as working fluid when compared to its respective mono [18] N.S.M. Sahid, M.M. Rahman, K. Kadirgama, M.A. Maleque, Experimental
counterparts. investigation on properties of hybrid nanofluids (TiO2 and ZnO) in water–ethylene
glycol mixture, J. Mech. Eng. Sci. 11 (4) (2017) 3087–3094.
[19] S. Simpson, A. Schelfhout, C. Golden, and S. Vafaei, “Nanofluid thermal
Funding conductivity and effective parameters,” Appl. Sci., vol. 9, no. 1, 2018, doi:
10.3390/app9010087.
[20] E. K. Ungar and L. R. Erickson, “Assessment of the use of nanofluids in spacecraft
This research work was funded by the Ministry of Higher Education active thermal control systems,” AIAA Sp. Conf. Expo. 2011, no. April, 2011, doi:
(MOHE), Malaysia under the Fundamental Research Grant [FRGS/1/ 10.2514/6.2011-7328.
2018/TK09/UNITEN/03/1] and Universiti Tenaga Nasional (UNITEN) [21] Stefano Casciardi, Renata Sisto, Marco Diociaiuti, The analytical transmission
electron microscopy: a powerful tool for the investigation of low-dimensional
BOLD2025 Phase 2 Grant [https://doi.org//10436494/B/2019058]
carbon nanomaterials, J. Nanomater. 2013 (2013) 1–15.
[22] Paula Guzmán, Victoria Fernández, Mohamed Khayet, María Luisa García,
CRediT authorship contribution statement Agustín Fernández, Luis Gil, Ultrastructure of plant leaf cuticles in relation to
sample preparation as observed by transmission electron microscopy, Sci. World J.
2014 (2014) 1–9.
Neeshoun Asokan: Methodology, Investigation, Data curation, [23] Amin Asadi, Meisam Asadi, Alireza Rezaniakolaei, Lasse Aistrup Rosendahl,
Writing - original draft. Prem Gunnasegaran: Supervision, Project Masoud Afrand, Somchai Wongwises, Heat transfer efficiency of Al2O3-MWCNT/
thermal oil hybrid nanofluid as a cooling fluid in thermal and energy management
administration. Vignesh Vicki Wanatasanappan: Conceptualization,
applications: an experimental and theoretical investigation, Int. J. Heat Mass
Visualization, Validation, Resources, Funding acquisition. Transf. 117 (2018) 474–486.
[24] Sadegh Aberoumand, Amin Jafarimoghaddam, Experimental study on synthesis,
stability, thermal conductivity and viscosity of Cu–engine oil nanofluid, J. Taiwan
Declaration of Competing Interest Inst. Chem. Eng. 71 (2017) 315–322.
[25] Evangelos Bellos, Christos Tzivanidis, Thermal analysis of parabolic trough
collector operating with mono and hybrid nanofluids, Sustain. Energy Technol.
The authors declare that they have no known competing financial Assess. 26 (2018) 105–115.
interests or personal relationships that could have appeared to influence [26] Khalil Khanafer, Kambiz Vafai, A critical synthesis of thermophysical
characteristics of nanofluids, Int. J. Heat Mass Transf. 54 (19-20) (2011)
the work reported in this paper.
4410–4428.
[27] Bock Choon Pak, Young I. Cho, Hydrodynamic and heat transfer study of dispersed
References fluids with submicron metallic oxide particles, Exp. Heat Transfer 11 (2) (1998)
151–170.
[28] Y. Raja Sekhar, K.V. Sharma, Study of viscosity and specific heat capacity
[1] N.N. Esfahani, D. Toghraie, M. Afrand, A new correlation for predicting the thermal
characteristics of water-based Al 2 O 3 nanofluids at low particle concentrations,
conductivity of ZnO–Ag (50%–50%)/water hybrid nanofluid: an experimental
J. Exp. Nanosci. 10 (2) (2015) 86–102.
study, Powder Technol. 323 (2018) 367–373.
[29] F.C. Ciornei, S. Alaci, D. Amarandei, L. Irimescu, I.C. Romanu, L.I. Acsinte, Method
[2] A. Akhgar, D. Toghraie, N. Sina, M. Afrand, Developing dissimilar artificial neural
and device for measurement of dynamic viscosity, IOP Conf. Ser.: Mater. Sci. Eng.
networks (ANNs) to prediction the thermal conductivity of MWCNT-TiO2/Water-
174 (2017) 012041, https://doi.org/10.1088/1757-899X/174/1/012041.
ethylene glycol hybrid nanofluid, Powder Technol. 355 (2019) 602–610.
[30] A. Einstein, “Einstein’s Dissertation on the Determination of Molecular Dimensions
[3] P. Sati, R.C. Shende, S. Ramaprabhu, An experimental study on thermal
Paper 1. A New Determination of Molecular Dimensions,” Energy, pp. 177–177.
conductivity enhancement of DI water-EG based ZnO(CuO)/graphene wrapped
[31] B. H.C., “Journal of Chemical Physics,” J. Chem. Phys., vol. 20, no. 4, p. 571, 1952.
carbon nanotubes nanofluids, Thermochim. Acta 666 (June) (2018) 75–81,
[32] G.K. Batchelor, The effect of Brownian motion on the bulk stress in a suspension of
https://doi.org/10.1016/j.tca.2018.06.008.
spherical particles, J. Fluid Mech. 83 (1) (1977) 97–117.
[4] R. Bubbico, G.P. Celata, F. D’Annibale, B. Mazzarotta, C. Menale, Comparison of
[33] G. Chrystal, A treatise on electricity and magnetism an elementary treatise on
the heat transfer efficiency of nanofluids, Chem. Eng. Trans. 43 (2015) 703–708,
electricity, Nature 25 (637) (1882) 237–240.
https://doi.org/10.3303/CET1543118.
[34] Sidi El Bécaye Maïga, Samy Joseph Palm, Cong Tam Nguyen, Gilles Roy,
[5] M. Abdollahi-Moghaddam, K. Motahari, A. Rezaei, Performance characteristics of
Nicolas Galanis, Heat transfer enhancement by using nanofluids in forced
low concentrations of CuO/water nanofluids flowing through horizontal tube for
convection flows, Int. J. Heat Fluid Flow 26 (4) (2005) 530–546.
energy efficiency purposes; an experimental study and ANN modeling, J. Mol. Liq.
[35] K. Pietrak, T. Wiśniewski, A review of models for effective thermal conductivity of
271 (2018) 342–352.
composite materials, J. Power Technol. 95 (1) (2015) 14–24.
[6] M. S. Parashurama, D. A. Dhananjaya, and N. K. R. R, “Experimental Study of Heat
[36] M. Danaei et al., “Impact of particle size and polydispersity index on the clinical
Transfer in a Radiator using Nanofluid,” vol. 3, no. 2, pp. 307–311, 2015.
applications of lipidic nanocarrier systems,” Pharmaceutics, vol. 10, no. 2, 2018,
[7] T. Salameh, M. Tawalbeh, M. El, and H. Assad, “Experimental and Numerical Study
doi: 10.3390/pharmaceutics10020057.
on Heat Transfer Enhancements of Concentric Tube Heat Exchanger Using Water
[37] Saeed Askari, Hadis Koolivand, Mahnaz Pourkhalil, Roghayyeh Lotfi,
Based Nanofluids,” 2018 5th Int. Conf. Renew. Energy Gener. Appl., pp. 232–237,
Alimorad Rashidi, Investigation of Fe 3 O 4/Graphene nanohybrid heat transfer
2018, doi: 10.1109/ICREGA.2018.8337627.
properties: experimental approach, Int. Commun. Heat Mass Transfer 87 (2017)
[8] J.A. Ranga Babu, K.K. Kumar, S. Srinivasa Rao, State-of-art review on hybrid
30–39.
nanofluids, Renew. Sustain. Energy Rev. 77 (October) (2017) 551–565, https://
[38] Vikas Kumar, Arun Kumar Tiwari, Subrata Kumar Ghosh, Effect of variable spacing
doi.org/10.1016/j.rser.2017.04.040.
on performance of plate heat exchanger using nanofluids, Energy 114 (2016)
[9] S. Suresh, K.P. Venkitaraj, P. Selvakumar, M. Chandrasekar, Effect of Al2O3–Cu/
1107–1119.
water hybrid nanofluid in heat transfer, Exp. Therm Fluid Sci. 38 (2012) 54–60.

12
N. Asokan et al. Thermal Science and Engineering Progress 20 (2020) 100727

[39] Mohammadali Baghbanzadeh, Alimorad Rashidi, Amir Hassan Soleimanisalim, [45] H.W. Xian, N.A.C. Sidik, S.R. Aid, T.L. Ken, Y. Asako, Review on preparation
Davood Rashtchian, Investigating the rheological properties of nanofluids of techniques, properties and performance of hybrid nanofluid in recent engineering
water/hybrid nanostructure of spherical silica/MWCNT, Thermochim Acta 578 applications, J. Adv. Res. Fluid Mech. Therm. Sci. 45 (1) (2018) 1–13.
(2014) 53–58. [46] M. R. Azizian, H. Ş. Aybar, and T. Okutucu, “Effect of nanoconvection due to
[40] Na. P.K., K. D.P., D. A., and D. D.K., “Experimental investigation of viscosity and Brownian motion on thermal conductivity of nanofluids,” Proc. 7th IASME /
specific heat of silicon dioxide nanofluid,” Micro Nano Lett., vol. 2, no. 3, pp. WSEAS Int. Conf. Heat Transf. Therm. Eng. Environ. HTE ’09, pp. 53–56, 2009.
67–71, 2007, doi: 10.1049/mnl. [47] N.A. Usri, W.H. Azmi, Rizalman Mamat, K. Abdul Hamid, G. Najafi, Thermal
[41] S.A. Angayarkanni, Vijutha Sunny, John Philip, Effect of nanoparticle size, conductivity enhancement of Al2O3 nanofluid in ethylene glycol and water
morphology and concentration on specific heat capacity and thermal conductivity mixture, Energy Proc. 79 (2015) 397–402.
of nanofluids, J Nanofluids 4 (3) (2015) 302–309. [48] M.N.F. Mohamad, W.A.W. Hamzah, K.A. Hamid, R. Mamat, Heat transfer
[42] R. Hentschke, On the specific heat capacity enhancement in nanofluids, Nanoscale performance of Tio2-SiO2 nanofluid in water-ethylene glycol mixture, J. Mech.
Res. Lett. 11 (1) (2016) 1–11, https://doi.org/10.1186/s11671-015-1188-5. Eng. 5 (1) (2018) 39–48.
[43] Eric C. Okonkwo, Ifeoluwa Wole-Osho, Doga Kavaz, Muhammad Abid, Comparison [49] M. Nuim Labib, Md. J. Nine, Handry Afrianto, Hanshik Chung, Hyomin Jeong,
of experimental and theoretical methods of obtaining the thermal properties of Numerical investigation on effect of base fluids and hybrid nanofluid in forced
alumina/iron mono and hybrid nanofluids, J. Mol. Liq. 292 (2019) 111377, convective heat transfer, Int. J. Therm. Sci. 71 (2013) 163–171.
https://doi.org/10.1016/j.molliq.2019.111377. [50] Hyder H. Balla, Shahrir Abdullah, Wan MohdFaizal, Rozli Zulkifli,
[44] Ifeoluwa Wole-Osho, Eric C. Okonkwo, Doga Kavaz, Serkan Abbasoglu, An Kamaruzaman Sopian, Numerical study of the enhancement of heat transfer for
experimental investigation into the effect of particle mixture ratio on specific heat hybrid CuO-Cu nanofluids flowing in a circular pipe, J. Oleo Sci. 62 (7) (2013)
capacity and dynamic viscosity of Al2O3-ZnO hybrid nanofluids, Powder Technol. 533–539.
363 (2020) 699–716. [51] Abdolreza Moghadassi, Ehsan Ghomi, Fahime Parvizian, A numerical study of
water based Al2O3 and Al2O3–Cu hybrid nanofluid effect on forced convective
heat transfer, Int. J. Therm. Sci. 92 (2015) 50–57.

13