Professional Documents
Culture Documents
Used Oleymine To Achieve Stability Over 2 Months
Used Oleymine To Achieve Stability Over 2 Months
PII: S0894-1777(15)00304-0
DOI: http://dx.doi.org/10.1016/j.expthermflusci.2015.10.028
Reference: ETF 8618
Please cite this article as: D.K. Agarwal, A. Vaidyanathan, S. Sunil Kumar, Experimental Investigation on Thermal
Performance of Kerosene-Graphene Nanofluid, Experimental Thermal and Fluid Science (2015), doi: http://
dx.doi.org/10.1016/j.expthermflusci.2015.10.028
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers
we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting proof before it is published in its final form. Please note that during the production process
errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Experimental Investigation on Thermal Performance of Kerosene-
Graphene Nanofluid
a
Liquid Propulsion Systems Centre, ISRO, Thiruvananthapuram, Kerala, India - 695 547,
b
Indian Institute of Space Science and Technology, Thiruvananthapuram, Kerala, India - 695
547,
*
Corresponding author. email: sunil_plamood@yahoo.com, ph: +91 471 256 7535, fax: +91
Abstract
nanofluid containing graphene nanoplatelets (GNP) for its potential use in regenerative
0.02, 0.05, 0.1, 0.2 weight percentage (wt%) and 300, 500, 750 m2/g, specific surface area
(SSA) are prepared using ultrasonication and steric stabilization technique. Oleylamine is
found to be the suitable surfactant with an optimum oleylamine to GNP mass ratio of 0.6, for
maximum stability of nanofluid. Dynamic Light Scattering (DLS) technique along with
thermal conductivity measurements are used to identify the stability of nanofluid with time.
are observed for 750 SSA, 0.2 wt% kerosene-GNP nanofluid. Lower enhancement in thermal
conductivity and viscosity are noticed for nanofluid that is prepared with lower SSA GNPs.
The study shows less significant effect of temperature on the thermal conductivity of
conductivity is observed for 750 SSA nanofluid as compared to other sizes of GNPs.
coefficient is noticed for 750 SSA, 0.2wt% kerosene-GNP nanofluids. A correlation for
friction factor based on the experimental data is also determined. Merit number, used to
determine the total heat transfer performance shows the utility of these nanofluids as heat
transfer fluids.
Nomenclature
f Friction factor
I Current (A)
2
N Particle count
Nu Nusselt number
Pr Prandtl number
Re Reynolds number
T Temperature (ºC)
V Voltage (V)
Greek Symbol
ρ Density (kg/m3)
Subscripts
bf base fluid
c current value
f fluid
3
i initial value
nf nanofluid
np nanoparticle
s surface
4
1. Introduction
In rocket engines, the increased thrust requirement lead to increased heat flux on the thrust
chamber walls and therefore needs to be regeneratively cooled by circulating one of the
propellants along the outer surface of the nozzle wall through cooling channels. The cooling
efficiency of any fluids depends upon its thermo-physical properties and flow velocity. The
increase in flow velocity increases the pressure drop in the system which in turn increases
Regenerative cooling is used to protect the thrust chamber wall from the hot combustion
products. Kerosene is used as coolant in the regenerative passage before admitting it into the
combustion chamber. Kerosene with low thermal conductivity is also susceptible to coking if
the fluid temperature through the regenerative passage increases beyond a certain limit.
Therefore, to enhance the applicability limit of semi cryogenic engine at higher thrust levels
and reusability of the thrust chamber, it is pertinent to increase the thermal conductivity of its
nanoparticles in it. Choi [1] in 1995 at the Agronne National Laboratory, Chicago has carried
out the pioneering work of dispersing nano-sized particle in a base fluid and named this
that exhibit unique physical and chemical properties compared to those of larger particles of
the same materials. The presence of nanoparticles with large surface area is expected to
enhance the heat transfer and also stability of the nanofluid. This property of nanoparticle is
utilized to develop nanofluids with an unprecedented combination of two features that are
highly desirable for heat transfer systems; extreme stability and very high thermal
conductivity.
5
Over the past one and half decade, many researchers have investigated the nanofluids for
their enhanced heat transfer properties. Researchers have carried out wide range of
experimental studies [2] using spherical particles made of oxides, carbide and metal
nanoparticles in water or ethylene glycol. Most of these studies clearly indicate higher heat
transfer performance for nanofluids as compared to the respective base fluid. The enhanced
nanotube (MWCNT) was first reported by Choi et al. [3]. An increase of 160% in thermal
conductivity for 1 vol% MWCNTs dispersed in poly-(αolefin) oil was noticed. The main
reason for the selection of carbon nanotubes (CNT) was its very high thermal conductivity.
Subsequently, many studies were carried out using MWCNT [4,5] and single walled carbon
nanotube SWCNT [6,7] and most of the studies reported significantly higher heat transfer
Graphene, a single atomic layer of graphite has two-dimensional form of carbon and found to
exhibit high crystal quality as it has ballistic electronic transport at room temperature.
Graphene attracted much attention since it was discovered by Novoselov et al. [8] in year
2004. Graphene has superior thermal conductivity and therefore would expect to result
higher heat transfer performance as a nanofluid. According to Hamilton and Crosser [9],
when the particle to liquid thermal conductivity ratio of a suspension is above 100, the
particle shape can play a substantial effect on its effective thermal conductivity. More
recently Balandin et al. [10] reported very high value of thermal conductivity of 5000 W/mk
for single-layer graphene. This has generated increased interest among the scientific
community to use this novel material as suspension in base fluid to be used as a heat transfer
fluid.
6
The first study on graphene based nanofluid was simultaneously reported by two researcher
groups, Yu et al. [11] and Baby and Ramaprabhu [12]. Yu et al. [11] experimentally
determined the thermal conductivity of Graphene Oxide based nanofluid and reported 30.2%,
62.3% and 76.8% enhancement in thermal conductivity for 5 vol% Graphene oxide
nanosheets in water, propyl glycol and liquid paraffin respectively. Baby and Ramaprabhu
[12] used functionalized thermal exfoliated graphene oxide (f-TEG) in Deionized (DI) water
and Ethylene glycol (EG) as nanofluid. They found significant increase in thermal
conductivity of water based nanofluid as compared to those with EG. The reported value for
0.056% vol% f-TEG water nanofluid was 14% at 25⁰C and 64% at 50⁰C. Further, in their
study, the authors [13] found 76% enhancement in heat transfer coefficient at 0.01 vol% f-
TEG water nanofluid for a Reynolds number of 4500. In their subsequent work [14] the
thermal performance of Ag/HEG nanofluid are determined and nanofluids was found to be
Mehrali et al. [15] prepared water-GNP nanofluids without using surfactant and observed
determined the rate of sedimentation of GNPs with time and reported the effect of specific
surface area (SSA) and concentration of sedimentation rate. They also observed increased
Effect of temperature on thermal conductivity was first reported by Das et al. [16] for metal
oxide based nanofluids. The authors observed 2 to 4 fold increase in thermal conductivity of
nanofluid as the temperature was increased from 21⁰C to 51⁰C. Subsequently, large number
of researchers carried out study in this area and some of them reported significant
enhancement of thermal conductivities with temperature [12, 17-21] whereas, some have
7
Baby and Ramprabhu [12] observed significant effect of temperature on thermal conductivity
for water- graphene nanofluid. Ijam et al. [17] prepared GNP - glycerol + water nanofluids
and monitored the stability of nanofluids with time. It was found that their nanofluids were
stable for more than 5 months. The study also showed significant effect of temperature on the
thermal conductivity of nanofluids. Wang et al. [18] prepared stable graphene based
nanofluid in ionic liquid without surfactant and reported 15.5% and 18.6% increase in
thermal conductivity at 25⁰C and 65 ⁰C respectively for 0.06 wt% of graphene. They found
decrease in viscosity for graphene based nanofluid as compared to the base fluid. Thermal
conductivity of nanofluid at very high temperature is measured by Liu et al. [19] for
Contrary to above mentioned findings, Yu et al [11] did not observe any increase in thermal
conductivity for graphene oxide nanofluids with temperature. Kole and Dey [22] determined
the thermal conductivity and viscosity at various volume fractions of graphene nanosheets in
water + EG fluid and also reported the insignificant effect of temperature on thermal
conductivity of the nanofluids. The increase in viscosity reported for their nanofluid was up
Brownian motion of nanosheets depends upon the fluid temperature whereas the percolation
8
behaviour in between the metal oxides and carbon nanotubes. A role of percolation and sheet
Recent studies of, Li et al. [27] and Sonawane et al. [20] respectively with kerosene and
kerosene- lipophilic Cu nanofluid and reported the effect of surface modification on the
thermal properties. In the study of Sonawane et al. [20] with ATF-alumina nanofluid, the
thermo-physical properties and heat transfer coefficient at various volume fractions were
A study carried out by Justin et al. [28] on combustion characteristics of graphene based
fuels, indicates higher combustion performance of graphene based nanofuels. They studied
the linear burning rates of nitro-methane in presence of aluminium, silicon oxide and
graphene sheets and reported highest increase in burning rate with graphene sheets. The
increased thermal conductivity. Pure aluminium particles are used as additive in solid rocket
difficult as it spontaneously gets oxidized in presence of atmosphere and catches fire. As the
heat of combustion of carbon and aluminium is 31.1 kJ/g and 32.8 kJ/g respectively,
engine. Kerosene is an aviation fuel and is also used as a fuel in semi cryogenic engine.
Moreover the use of graphene in fuel will not contribute to any particulate matter of plume
9
Through many experimental studies on thermal conductivity of graphene based nanofluids
are available in the literature; the studies on convective heat transfer are limited. Very few
researchers have worked on kerosene as base fluid. Literature review also indicates large
disparity among the researchers on the percentage enhancement in thermal conductivity with
respect to its temperature dependency even for the same nanofluid. The disagreement
between these reported values may be attributed to the different techniques used for realizing
nanoparticles and the method of preparing the nanofluid itself as discussed by the authors
[21]. Thus a systematic study with the standardization of experimental method from
nanofluid preparation till measurement is essential to evaluate the true potential of nanofluid
in heat transfer applications. In the best knowledge of the authors, no study has been carried
out so far using kerosene-graphene nanofluid. Thus a systematic study is presented in this
paper for kerosene- GNP nanofluid for its potential application in thrust chamber cooling of
rocket engine.
nanofluid using various surfactants for stabilization and determination of its thermo-physical
and turbulent flow convective heat transfer properties at different particle mass fraction. Total
heat transfer performance of these nanofluids is evaluated by measuring the pressure drop in
the test section during each test. The study also aims to determine the potential use of
2. Experimental Techniques
The particle size of the nanofluids is measured using Malvern Zetasizer, which is used by
many researchers in their study [2, 21]. Zetasizer works based on the principle of Dynamic
Light Scattering (DLS) and the particles are illuminated with a laser source. The fluctuations
10
in the intensity of laser light were detected, recorded and used for particle size determination.
In the present study, Zetasizer is used to compare and assess the stability of nanofluids by
measuring the hydrodynamic diameter (d) distribution of GNPs with time. The hydrodynamic
diameter (d) is the diameter of an equivalent sphere that has the same translational diffusion
coefficient as the particle. Zetasizer estimate the hydrodynamic diameter (d=KT/3πµD) using
diffusion coefficient.
In the present work, both thermal conductivity and dynamic viscosity of nanofluids are
measured under static condition. Transient hot wire (THW) method for thermal conductivity
measurement of nanofluids was successfully used by many researchers [12-15, 17, 21]. In
current study thermal conductivity is measured using thermal property analyser named KD2-
Pro, with a specified accuracy of ±5%. It works on the principle of transient hot wire method.
The nanofluid sample is kept in a constant temperature bath for the thermal conductivity
The dynamic viscosity is measured using Brookfield digital viscometer (model: LVDV-II
pro) equipped with constant temperature bath [21, 22]. In this technique, a controller drives a
spindle that is immersed in nanofluid. The resultant viscous drag generated due to the spindle
rotation is measured during the experiment and the fluid viscosity is determined. The rotation
of spindle for viscosity measurements was maintained at 100 rpm, as it provides the
maximum range of 3.04 cP for viscosity measurements. The accuracy specified for the
equipment is 1% of the maximum measurement range ie; 0.03 cP for the current
measurement setup. The measurement procedure detailed by the authors [21] is followed in
the current study to determine the viscosity of different nanofluids under investigation.
11
2.3. Convective heat transfer setup
The schematic of the closed loop convective heat transfer experimental setup is shown in Fig.
1. The experimental setup consists of a test section, resistively heated by a DC power source,
nanofluid ultrasonication unit, variable flow gear pump, turbine flow meter, heat exchanger, a
differential pressure indicator and data acquisition system. Turbine type flow meter used in
the experiments measures the flow rate with an accuracy of ±0.25%. A variable high current
DC power source of 20 kW (1000A, 20V) is used for resistive heating of test section. Water
cooled plate heat exchanger with cooling capacity of 25 kW is used to cool the fluid after the
test section. The differential pressure indicator with a range of 300 mbar is connected to the
test section for pressure drop measurement. A 30 channel data acquisition system is used to
record surface temperatures as well as inlet and outlet fluid temperature during the test. Test
section is made up of 1.2 m long stainless steel tube having 9.5 mm outer diameter with 0.9
mm thickness. Ten T-type thermocouples are welded on to the test section for measuring the
surface temperatures and two T-type thermocouples each of which are positioned at the inlet
and outlet of the test section respectively for measuring the corresponding fluid temperatures.
The accuracy of T-type thermocouple used in the setup is ±0.5⁰C. Experiments are carried out
in turbulent flow regime. The entrance length (X) for fully developed flow [29] in turbulent
X ≥ 10Di (1)
Di is the inner diameter of the test section. The minimum entrance length estimated for the
current setup is 0.077 m. The tube diameter in the setup is constant everywhere and before
the test section, a length of 0.2 m is provided for the flow to become hydrodynamically fully
developed. An entrance length of 0.5 m, (X /Di = 65) in the test section is taken before
making the measurements for computing heat transfer coefficient, to ensure fully developed
thermal and flow profiles. The test section is insulated with glass wool to ensure negligible
12
heat loss to the ambient. Kerosene used in the present study is Rocket grade kerosene,
without aromatic contents. Property of Rocket grade kerosene is determined and reported by
Deepak et al [30]. Thermal properties of kerosene and GNP at 30⁰C used in the current study
Nanofluid of desired particle mass concentration and quantity is prepared using oleylamine as
surfactant. Thermal conductivity and viscosity of nanofluid is measured using KD2 pro and
temperatures from 20 ⁰C to 70 ⁰C for each weight percentage and particle size. Samples are
kept at a constant temperature during the measurement using a constant temperature bath.
Nanofluid is filled in the reservoir and is continuously ultrasonicated with the help of
ultrasonic probe, flow cell and peristaltic pump. A variable speed gear pump delivers the
desired volume flow of nanofluid to the test section. The test section is resistively heated with
DC power source. The heated nanofluid after the test section gets cooled near room
temperature using water cooled plate heat exchanger and is emptied into the reservoir. The
process is continued till the attainment of steady state condition. The inlet fluid temperature
to the test section is maintained constant during the test. The experiments are carried out for
different Reynolds number by varying the flow rate and each test is repeated for four times to
ensure repeatability.
After each test, nanofluid is flushed out from the test setup using nitrogen gas. Subsequently,
A series of convective heat transfer experiments are carried out to investigate heat transfer
characteristics of nanofluid at turbulent flow regime. Steady state fluid temperature, surface
13
temperature, current (I) and voltage (V) from DC power source and flow rate are measured
during the experiment for calculating heat transfer coefficient and Nusselt number at various
particle mass concentrations and Reynolds numbers. Measured thermal conductivity and
viscosity at various temperatures are curve-fitted to estimate the values at any intermediate
temperature point. The effective density and specific heat of nanofluid are estimated based on
Where, Tin , Tout are fluid temperature at inlet and outlet of the test section respectively.
Qe = V I (5)
As the value of specific heat of nanofluid is calculated by mixing rule, the correctness of
specific heat is ensured by comparing heat absorbed by the nanofluid, Qf in test section to the
electrical heat supplied, Qe to test section. In all the experiments (Qe - Qf) is found to be
positive and the value is less than 5.9% and this indicated the effectiveness of the insulation
provided on the test section. Reynolds number and Prandtl number of nanofluid is determined
Local Heat transfer coefficient, hnf = Qf/ (πDi L (Ts -Tf)) (8)
14
where, Ts and Tf is the local surface and fluid temperature respectively.
Where, Δp is the measured pressure dro and .unf = Ṽnf/( πDi2) is nanofluid velocity at test
section.
The merit number is calculated using heat transfer coefficient and measured pressure drop
and is defined as
An uncertainty analysis of the experimental data is carried out and the maximum uncertainty
at various GNP mass concentrations and sizes. Nanofluids are prepared and stabilized using
the stability of these nanofluids with time. Measurement of thermo-physical properties along
with convective heat transfer experiments are carried out to investigate the utility of these
nanofluids as a heat transfer fluid. Table 3 presents matrix of experiments carried out in the
present study. Detailed discussions related to the results obtained from the set of experiments
15
Uniform dispersion of nanoparticle in the base fluid is an important requirement for any
colloidal system to be effectively used for heat transfer application. This has been analysed in
detail in the present work wherein nanofluids are prepared using a two-step technique. Three
catalogue sizes of Graphene Nano Platelets (GNP) 750, 500 and 300m2/g, procured from M/s
XG Sciences, USA are used in the present investigation. Kerosene used in the present study
is rocket grade kerosene used as a fuel in semi cryogenic engine. Initially, water and kerosene
based nanofluids were prepared with different ultrasonication time. The stability of
ultrasonication time, GNPs were found to settle immediately (within few minutes) in
kerosene. However, water based nanofluids were visually stable for few hours. As kerosene is
The stability of any colloidal system depends on the balance of repulsive and attractive forces
existing between the particles. The repulsive force must be larger as compared to the
attractive forces between the particles for stable nanofluids. The force between the particles
depends on the ionic concentration near the particles and it can be changed by either
electrostatic or steric stabilization techniques as reported by many researchers [2, 31- 33] to
stabilize nanofluids.
In the present work various surfactants are chosen based on its affinity towards the fluid and
were used to stabilize the GNP-kerosene nanofluid. Initially, based on our previous research
experience [21] Oleic acid was tried for stabilizing 0.05 wt% kerosene - GNP nanofluid but it
could not stabilize the nanofluids and GNPs started settling after few hours. At higher mass
concentrations above 0.1wt%, settling started after few minutes itself. Subsequently many
other surfactants were tried and behavior and stability of nanofluids are visually observed.
Table 4 reports the outcome of the stability study using various surfactants. The kerosene-
16
GNP nanofluids prepared using Oleylamine, Oleic acid and Tween-20 surfactants are shown
in Figure 2. It can be clearly seen that the nanofluid prepared using Oleic acid and Tween-20
are not stable. Oleylamine provides good colloidal stability of GNP in kerosene and is found
nanofluid using a syringe and the nanofluid is continuously ultrasonicated during the addition
process. The surfactant to particle mass ratio was varied from 0.3 to 1.2 in steps. The stability
of nanofluid was very poor for the mass ratio of less than 0.6. An optimum quantity of
particle mass ratio of 0.6 was found to yield the maximum visual stability.
ratio of 0.6 are made. Different ultrasonication time of 40 min for 0.005 wt% to 3 hrs for
concentration of the GNPs. Nanofluids at lower mass fraction up to 0.05wt% were very
stable (> 4-12 months) as compared to those with higher particle mass concentration of
0.1wt% and 0.2wt% nanofluids for which the stability was found to be around 20-30 days.
Nanofluids at 0.05 wt % are prepared for 750 and 500 SSA GNPs. Particle sizing of the
nanofluid are measured with time (days) to determine the stability of nanofluid. The particle
size distribution for the samples is measured using Malvern Zetasizer. All the size
measurement is performed at 25⁰C with a scattering angle of 173 ⁰. Each sample is tested
Nanofluid from the top portion in the bottle is used as the sample for testing in first day, 30th
day, and at 60th day after preparation and particle size measurement is conducted. Figures 3
and 4 represents the particle size distribution based on intensity measurement from DLS
17
technique for 750 and 500 SSA kerosene-GNP nanofluid respectively. The size (d, nm) given
can be seen from the Figs 3 and 4 that intensity curve of 750 SSA nanofluid remains nearly
unchanged even after 60 days of preparation. The mean Z-average value measured for 750
SSA GNP nanofluid is found to vary only from 184.6 to 192.4 in 60 days. Z-average value is
a harmonic intensity averaged particle diameter reported by the measurement system and it is
very sensitive to small changes in the sample, viz particle agglomeration or settling. A
constant value of Z-average seen in the present study indicates highly stable nanofluid. The
particle intensity plot for 500 SSA GNPs nanofluid is slightly changed with time. The mean
Z-average value for 500 SSA nanofluids varies from 442 to 463 in 60 days. The maximum
dispersion in Z average particle size among the three DLS measurements was within 1.2% for
750SSA and 0.9% for 500SSA nanofluid. It can be noticed from these figures that though
both the nanofluids are stable up to 60 days, stability of 750 SSA nanofluid is higher
Heat transfer performance of any heat transfer system depends upon the thermal conductivity
of coolant medium. The increased viscosity due to addition of nanoparticles in the base fluid
increases the pressure requirement of the heat transfer system. Detailed measurements of
thermal conductivity and viscosity are made at 0.005wt%, 0.02wt%, 0.05wt%, 0.1wt%, and
0.2wt% GNPs in kerosene for 750, 500,300 SSA graphene sheets and the results are
presented here.
measurements, the instrument is calibrated with DI water and glycerine. The sample is kept
18
at a constant temperature of 30⁰C ± 0.25⁰C using a constant temperature bath. To reduce the
effect of convection current outside the sample tube, a special holder is made and the
constant temperature bath is kept switched off during the measurement process. For each
mass fraction, six measurements are taken and all of them fall within ±0.02 W/mK. Mean
mass concentrations for three different sizes of 750, 500 and 300 SSA is shown in Fig.5. The
ratio of thermal conductivity of nanofluid to base fluid increases with particle loading for all
sizes of GNPs. The enhancement ratio of thermal conductivity of nanofluid to base fluid is
found to be higher for the higher SSA particles as compared to the lower SSA for a range of
particle concentrations. The enhancement ratio at 0.2 wt% concentration for 750, 500, 300
SSA GNPs nanofluid is found to be 1.23, 1.17 and 1.13 respectively. The effect of particle
size on thermal conductivity can be attributed mainly to two reasons, viz, specific surface
area (SSA) and Brownian motion as discussed by Deepak et al [21]. The higher surface area
per unit volume of GNPs will result in increased interaction between the solid particles to the
base fluid and will also enhance the possibility of Brownian motion due to lower mass of
individual particles. The frequent interactions results in higher thermal conductivity and also
explain the reason for the higher enhancement in thermal conductivity for 750 SSA GNPs
Deepak et al [21] reported 7.4% increases in thermal conductivity for 0.05vol% (equivalent
to 0.25 wt%) kerosene – alumina nanofluid. In the current investigation thermal conductivity
enhancement is 23.2% for 0.2 wt% GNP-kerosene nanofluid. The higher enhancement in
nanofluid can be explained based on percolation model [26, 34]. According to the model,
19
long chain present in CNT works as interconnecting thermal network which acts as a heat
conducting path. Gupta et al [25] reported similar networks for GNPs. Moreover the thermal
conductivity of graphene is higher as compared to CNT and is expected to have the enhanced
thermal conductivity for GNP based nanofluids. Graphene, a two-dimensional structure with
very high aspect ratio has largest surface area as compared to CNT or other spherical
nanoparticles. Due to the large interface area between the particle and fluid, it has lowest
contact resistance that result in increased thermal conductivity. In addition, due to the shape
of the GNPs as discussed by Hamilton and Crosser [9], higher effective thermal conductivity
of nanofluids is expected.
A bench mark study carried out by the researchers [35], of various organizations have used
Nan’s model [35] to compare their finding and showed a good agreement in the measured
knf / kbf = (3 + φ [2β11 (1-L11 ) + β33 (1-L33 )])/(3 – φ (2β11 L11 + 2β33 L33)) (13)
Where, Lii and φ are the geometrical factor and the volume fraction of particles respectively.
βii is given as
Where, kp is in-plane thermal conductivity [22] of nanoparticle. As the aspect ratio of GNPs
Zheng and Hong [36] reported a model by modifying the original Yamada-Ota model [37]
with the inclusion of effect of thermal boundary layer (TBR). The model is represented as;
In Fig. 6 the comparison of thermal conductivity measurements from the current study with
those calculated based on models by Naan et al. [36] and the Zheng and Hong model [37] for
20
three sizes of GNPs nanofluid are presented in detail. It is to be noted that the TBR value play
a significant role in the modified Yamda and Ota model [38] which is unknown for the
current nanofluid under investigation. Therefore TBR value for the present nanofluids are
estimated using the measured thermal conductivity and is found to be 10 -7 m2 KW-1, for 750
and 500 SSA nanofluid and 0.7x10 -7 m2 KW-1 for 300SSA nanofluid. The value of TBR
estimated is comparable to the value used by other researchers [6, 39]. In the current work,
Le is taken as 1000nm for 750 GNP nanofluids. In the case of 500 and 300 GNP, Le is
calculated in proportionate to its specific surface are per unit volume. Le/d is taken as 400 for
Figure 7 shows the ratio of thermal conductivity of nanofluid to the base fluid with change in
temperature for various GNP sizes and weight percentage. The ratio of thermal conductivity
for 750 GNPs particle size increases from 1.23 to 1.30 over 20⁰C to 70⁰C temperature range.
It can be noted that for kerosene-alumina nanofluid as reported by Deepak et al. [21], thermal
conductivity ratio was increases from 1.12 to 1.25 (> 2 times) for 13nm particle size in the
25⁰C to 65⁰C temperature range. Such a large enhancement in thermal conductivity with
temperature is not noticed for the current nanofluid. The present observation of lesser
temperature dependence is very similar to the observation made by other researchers [11, 18,
and 19]. It is to be noted that the significant effect of temperature on thermal conductivity
reported by previous researchers is mainly for metallic and ceramic nanofluids and is mainly
higher temperature. It is expected that at high temperature, the random motion of nanoparticle
along with the decrease in base fluid viscosity causes increased micro convection and results
in enhanced thermal conductivity. The CNT based nanofluid does not show temperature
dependency [33] due to the network structure as discussed earlier. GNPs used in the current
study has minimum 2.5 nm thickness sheet with varying length from 100nm to 2000nm. The
21
smaller GNPs would participate in Brownian motion and result in enhanced thermal
conductivity with temperature. However the longer GNPs will not produce Brownian motion
and the thermal conductivity increase in such case is due to percolation theory [25, 34].
Percolation does not depend upon temperature, hence thermal conductivity of these GNPs
In the present study the ratio of thermal conductivity increases from 1.23 to 1.30 for 750
GNPs, 1.14 to 1.18 for 500 GNPs and 1.11 to 1.14 for 300 GNPs particle size over the 20⁰C
is higher for 750 GNPs nanofluid as compared to the others. This can be attributed to the
different sized GNP particles present in the nanofluids. The smaller sized particle undergoes
random movement and thereby contributes to the Brownian motion, whereas with bigger
sized particle, network like chain structure, as postulated in percolation model could be
dominant. Gupta et al [25] also observed the similar hybrid behaviour of GNPs nanofluid
and reported that the thermal conductivity of Graphene nanofluids lies in between CNT and
metallic nanofluids.
The consistency in the measured thermal conductivity of the nanofluid with time indicates the
periodically and presented in Fig. 8. It can be seen that the thermal conductivity of 0.1wt%
and 0.2wt% nanofluids indicates marginal drop over time for initial 20 days and remain
lower concentration of 0.005 wt%, 0.02 wt% and 0.05wt% does not show any drop with time
and remain almost constant. This trend from thermal conductivity indicates very high
stability for the current kerosene- GNP nanofluids and is considered to be a prime
requirement of such fluid to be used for heat transfer application in rocket engine. This
22
observation is in line with the observation made using particle size distribution presented in
different temperatures. Nanofluids of particle mass fraction 0.005%, 0.02%, 0.05%, 0.1% and
0.2% are tested for the dynamic viscosity for three particle sizes along with pure kerosene.
Ratio of dynamic viscosity of nanofluids to the base fluid at various fluid temperatures and
particle concentrations for 300, 500 and 750 SSA particle size nanofluids is presented in Fig.
10. It is observed that the dynamic viscosity of nanofluid is higher as compared to the base
fluid and the enhancement in viscosity increases with particle loading and specific surface
area (SSA). The increase in viscosity with particle concentration is due to the formation of
agglomeration and clustered structure, at higher concentration, and could result in increased
internal shear stress for the nanofluid. The increase in dynamic viscosity for higher SSA
GNPs is large as compared to the lower SSA GNPs nanofluids. Higher surface area led to
higher interfacial resistance with the fluid layer and result in increased viscosity for 750 SSA
GNPs nanofluid as compared to others. The ratio of dynamic viscosity of nanofluid to the
base fluid is 1.076, 1.068 and 1.06 for 750, 500 and 300 GNP nanofluid respectively at 0.2wt
% and 30 ⁰C.
dynamic viscosity of 0.2 wt% nanofluid at various temperatures ranging from 20⁰C to 70⁰C.
Dynamic viscosity of nanofluid for three sizes of GNP nanofluids along with pure kerosene is
plotted in Fig. 11. The viscosity decreases rapidly with rise in temperature and display an
asymptotic behaviour. For the base fluid, as the temperature increases, due to thermal
expansion viscous shear stress goes down resulting in lower viscosity at high temperature.
23
The decrease in nanofluid viscosity with increase in temperature is expected due to the
weakening of the interparticle and intermolecular adhesion forces and similar trends have
also been observed in almost all type of nanofluids by many researchers [21, 32].
behaviour of nanofluids, the net benefit of nanofluid as heat transfer fluids is evaluated
through the heat transfer coefficients. The convective heat transfer performance of any fluid
depends upon various physical and thermo-physical properties. Though the enhanced thermal
conductivity of nanofluid will result in reduced resistance to thermal diffusion in the laminar
sub layer, increased viscosity of nanofluid could increase the thickness of sub layer and in
The test setup and experimental procedure used in current study is validated with pure
kerosene and is also used for the determination of convective heat transfer performance of
closed loop experimental test rig is carried out for 0.005wt%, 0.05wt% and 0.2wt% particle
transfer is experimentally studied for 750, 500 and 300 SSA GNPs nanofluids. Thermo-
physical properties at average film temperature ((Ts+Tf)/2) is used for the estimation of heat
transfer coefficient from the experimental data. Each experiment is repeated four times and
the mean values are used for the subsequent analysis. The maximum dispersion in the
calculated heat transfer coefficients among the experiments for any particular case is within
±2.7%.
24
Figure 11 shows the heat transfer coefficient for 750, 500 and 300 SSA GNPs nanofluids
derived from the experimental data at various particle concentrations and Reynolds numbers.
It can be clearly seen that kerosene- GNP nanofluids show enhanced heat transfer properties
as compared to the pure kerosene at all Reynolds numbers. As expected, the heat transfer
particle concentrations. Higher increase in heat transfer coefficient is noticed for higher SSA
nanofluids at higher concentration. This observation is similar to many other researchers [30-
32, 40] for their nanofluids. Highest increase in heat transfer coefficient is noticed for 0.2wt%
Ding et al. [4] proposed the reason for higher heat transfer coefficient observed using
nanofluid due to the increase in effective thermal conductivity (k) of nanofluid and the
decreases in thickness of the thermal boundary layer (δ); together it results in augmentation in
heat transfer coefficient (h=k/δ). The hypothesis of thinning of boundary layer was developed
based on the assumption of chaotic motion of nanoparticles in the nanofluid, which tends to
disturb the boundary layer, resulting in thinning of the same. This trend represented by Ding
et al [4] is also discussed by Aravind et al [41] and in authors opinion; it is true in the present
investigation too. The large enhancement observed in convective heat transfer is due to
discussed by Aravind et al [41]. The possible reasons for boundary layer thickness reduction
include particle migration in nanofluids due to shear action, viscosity gradient, and Brownian
motion as discussed by Phillips et al [42] and Ding et al [43]. Further as suggested by Xuan
and Li [44], the increase in effective thermal conductivity due to dynamic condition could be
another major factor for significant enhancement in convective heat transfer coefficient. It
has to be recalled that the thermal conductivity measurements carried out in the present study
25
is at static conditions and significant effect of shear could exist under the flowing condition
which would improve the thermal conductivity of nanofluid during convective experiments.
Figure 12 shows the ratio of convective heat transfer coefficient of nanofluid to base fluid at
various wt% and Reynolds number range. Higher enhancement is noticed for higher SSA
nanofluid for all the particle mass fractions. The higher performance of lower sized particle
(higher specific surface area) was observed by many researchers in the past [30-32, 40]. The
maximum enhancement noticed is 49% for 750 SSA nanofluid and 38% for 500 SSA
Though the current investigation show enhanced heat transfer coefficient of the kerosene-
GNP nanofluid, it is expected that the increased viscosity of nanofluids could also result in
higher pressure drop. Higher pressure drop may limit nanofluid applicability in thrust
chamber cooling as it may necessitate a bigger sized pump to push the fluid though the
The friction factor for kerosene- GNP nanofluid at various particle mass concentration and
sizes is computed from the measured pressure drop. It is to be noted that the increase in
viscosity of the current nanofluids are not very significant. Figure 13 shows the variation of
friction factor calculated in the present study along with the correlation developed using the
measurements at various Reynolds number. The correlation for friction factor is obtained in
present study by curve fitting all the data points and is expressed as;
f = 0.26Re-0.22 (16)
It can be noted that the maximum dispersion in the friction factor from the above correlation
is within 11%. The above correlation is very similar to Blasius correlation (f = 0.314 Re -0.25),
and shows that a homogeneous fluid correlation can be used to accurately predict pressure
26
drop characteristics of the nanofluids as used in the current study. Though the nanofluid
friction factor obtained from the present study does not significantly differ from the pure
fluid, a clear trend of higher friction factor is observed for higher SSA and for higher
0.2wt% 750 SSA kerosene-GNP nanofluid. It is to recall that the trend observed in friction
factor is very similar to the trend observed in heat transfer coefficient as discussed in section
3.3.
The pressure drop and friction factor depends upon the Reynolds number and GNP
concentration. As expected, similar to the pure fluid, the friction factor decreases with
increase in Reynolds number. Table 5 shows the increase in friction factor at various GNP
sizes and concentrations for Reynolds number range of 6700-28000. It can be noted that the
higher enhancement in friction factor is noticed for 750 GNP nanofluids and it increases with
particle concentration. At higher particle concentration, the density of fluid changes and
results in higher viscous drag thereby increasing the friction factor. The dynamic viscosity of
750 SSA nanofluid is also noticed higher compared to other particle size as discussed in
section 3.2.3. The maximum enhancement in friction factor noticed in the present study is
The total heat transfer performance of kerosene- GNP nanofluid is determined by Merit
number, comparing the heat transfer augmentation of nanofluid to the pressure penalty due to
increased density and the viscosity of nanofluid. Merit number is used by Deepak et al [30]
and Yu et al. [45] for evaluating the total heat transfer performance of kerosene- Al2O3 and
water- SiC nanofluid respectively. It is defined as the ratio of heat transfer enhancement to
the increase in pressure drop and the expression is given in Eq. (12).
27
Figures 14 show the heat transfer performance of 750 and 500 SSA nanofluid based on merit
number with respect to Reynolds number for 0.05% and 0.2% GNP mass concentrations. For
nanofluid with particle mass fraction of 0.05%, Merit number is found to be lower than one
and indicates lower performance of these nanofluids as a coolant medium. Merit Number is
found to be above unity for 0.2% mass concentration nanofluid for both sizes of GNPs.
Figure 15 presents the heat transfer coefficient of nanofluid with respect to the measured
pressure drop. A solid line shown in figure represents the heat transfer coefficient of pure
kerosene with respect to pressure drop. It is to be noted that the data point lies left to the solid
line indicate higher heat transfer performance of nanofluid compared to pure kerosene for an
identical value of pressure drop. Higher heat transfer coefficient with minimum pressure drop
is observed for 0.2wt% nanofluid for all sizes of GNPs. The study clearly suggests the total
heat transfer performance of 0.2wt% GNP-kerosene nanofluids to be higher than that of pure
kerosene and can be used as a better coolant in regenerative passage of semi-cryo thrust
chamber.
4. Conclusion
The present study aims at syntheses, characterization and assessing the potential use of
measurements are made to determine the effect of particle mass concentration, size and
1. An optimum oleylamine to particle mass ratio of 0.6 is determined after a series of trials
28
2. Stability of nanofluids is accessed by measuring thermal conductivity along with particle
size and time. Both thermal conductivity and particle size measurements show good
particle loading and sizes. The highest thermal conductivity enhancement ratio of 1.3,
1.18 and 1.14 is noticed for 750, 500 and 300 SSA for 0.2wt% kerosene-GNP nanofluid
respectively. Similarly, the increase in viscosity ratio is limited to 1.076 for 750 SSA at
4. Thermal conductivity enhancement from 23% at 20⁰C to 29% at 70⁰C is noticed for
0.2wt% 750SSA nanofluid. Not so significant effect of fluid temperature noticed for
sheet in the GNPs. Brownian motion in smaller size sheets and thermal conduction in
longer chain structure plays important role for enhanced thermal conductivity of
kerosene-GNP nanofluids.
regime is also carried out at various particle loading and sizes. Experimental data
pure kerosene and the maximum enhancement ratio observed is 1.49, 1.38 for 750 and
500 SSA GNPs respectively at 0.2% mass concentration nanofluid. Higher enhancement
6. Based on the measured pressure drop during the convective heat transfer experiments a
correlation for friction factor is developed, which lies within ±10% for all the sizes and
29
7. Detailed analysis of convective heat transfer coefficient and measured pressure data
shows Merit number to be above unity for nanofluids with particle mass concentration
above 0.2wt% for 750 and 500 SSA kerosene-GNP nanofluids. Merit number above
unity indicates potential use of these nanofluids as a suitable heat transfer fluid in
Acknowledgement
The authors acknowledge the help of Shri. Jophy Peter, Thermal Engineering Division,
30
References:
ANL/MSD/CP-84938.
[4] Ding, Yulong, Hajar Alias, Dongsheng Wen, and Richard A. Williams. "Heat transfer
[5] Aravind, SS Jyothirmayee, Prathab Baskar, Tessy Theres Baby, R. Krishna Sabareesh,
based nanofluids." The Journal of Physical Chemistry C 115, no. 34 (2011): 16737-
16744.
[6] Harish, Sivasankaran, Kei Ishikawa, Erik Einarsson, Shinya Aikawa, Shohei Chiashi,
glycol with single-walled carbon nanotube inclusions." International Journal of heat and
[7] Horton, Mark, Haiping Hong, Chen Li, Bo Shi, G. P. Peterson, and Sungho Jin.
31
[8] Novoselov, Kostya S., Andre K. Geim, S. V. Morozov, D. Jiang, Y_ Zhang, S. V.
[10] Balandin, Alexander A., Suchismita Ghosh, Wenzhong Bao, Irene Calizo, Desalegne
Teweldebrhan, Feng Miao, and Chun Ning Lau. "Superior thermal conductivity of
[11] Yu, Wei, Huaqing Xie, and Wei Chen. "Experimental investigation on thermal
[12] Baby, Tessy Theres, and S. Ramaprabhu. "Investigation of thermal and electrical
(2010): 124308.
[13] Baby, Tessy Theres, and Sundara Ramaprabhu. "Enhanced convective heat transfer
using graphene dispersed nanofluids." Nanoscale research letters 6, no. 1 (2011): 1-9.
[14] Baby, Tessy Theres, and Sundara Ramaprabhu. "Synthesis and nanofluid application of
(2011): 9702-9709.
[15] Mehrali, Mohammad, Emad Sadeghinezhad, Sara Tahan Latibari, Salim Newaz Kazi,
Mehdi Mehrali, Mohd Nashrul Bin Mohd Zubir, and Hendrik Simon Cornelis
(2014): 1-12.
32
[16] S. K. Das, N. Putra, P. Thiesen, W. Roetzel, Temperature dependence of thermal
conductivity enhancement for nanofluids, Journal of Heat Transfer, 125 (2003): 567-
574.
[17] Ijam, Ali, A. Moradi Golsheikh, R. Saidur, and P. Ganesan. "A glycerol–water-based
nanofluid containing graphene oxide nanosheets." Journal of Materials Science 49, no.
17 (2014): 5934-5944.
[18] Wang, Fuxian, Lijuan Han, Zhengguo Zhang, Xiaoming Fang, Jingjing Shi, and Wenshi
[19] Liu, Jian, Fuxian Wang, Long Zhang, Xiaoming Fang, and Zhengguo Zhang.
2841-2849.
[21] Deepak Kumar Agarwal, Aravind Vaidyanathan, and S. Sunil Kumar. "Synthesis and
[22] Kole, Madhusree, and T. K. Dey. "Investigation of thermal conductivity, viscosity, and
33
[23] E. V.Timofeeva, A. N. Gavrilov, J.M. McCloskey, Y.v. Tolmachev, “Thermal
[24] Y. S. Ju, J. Kim, Ming-Tsung Hung, “Experimental study of heat conduction in aqueous
092403-1-6.
percolation and sheet dynamics heat conduction in poly dispersed graphene nanofluids.”
[28] Justin L. Sabourin , Daniel M. Dabbs , Richard A. Yetter , Frederick L. Dryer, Ilhan A.
[29] Incropera Frank P., DeWitt David P., “Fundamental of heat and mass transfer.” USA,
[30] Deepak Kumar Agarwal, Aravind Vaidyanathan, and S. Sunil Kumar. “Investigation on
[31] Wang, Xiang-Qi, and Arun S. Mujumdar. "Heat transfer characteristics of nanofluids: a
[34] Sastry, N N Venkata, Avijit Bhunia, T. Sundararajan, and Sarit K. Das. "Predicting the
study on the thermal conductivity of nanofluids.” Journal of Applied Physics 106 09312
(2009): 1-14.
[36] C. W. Nan, Z. Shi, Y. Lin, “A simple model for thermal conductivity of carbon
[37] Zheng, Yingsong, and Haiping Hong. "Modified model for effective thermal
[38] Yamada, Etsuro, and Terukazu Ota. "Effective thermal conductivity of dispersed
[39] Harish, Sivasankaran, Kei Ishikawa, Shohei Chiashi, Junichiro Shiomi, and Shigeo
dispersed nanofluids for heat transfer applications.” Journal of Applied Physics 112,
(2012): 124304.
(2013): 4199-4106.
35
[42] Ronald J. Phillips, , Robert C. Armstrong, Robert A. Brown, Alan L. Graham, and
James R. Abbott. "A constitutive equation for concentrated suspensions that accounts
for shear‐ induced particle migration." Physics of Fluids A: Fluid Dynamics (1989-
[43] Ding, Yulong, Hajar Alias, Dongsheng Wen, and Richard A. Williams. "Heat transfer
[44] Y. Xuan, Q. Li, Investigation on convective heat transfer and flow features of
[45] Yu, W., France, David M., Smith, David S., Singh, D., Timofeeva, Elena V., Routbort,
Jules L., Heat transfer to a silicon carbide/water nanofluid, International Journal of Heat
36
List of figures
Figure 2 Kerosene- GNP nanofluid with (a) Oleylamine (b) Oleic acid (c) Tween-20
Figure 3 Particle size distribution of 750 SSA 0.05 wt% kerosene-GNP nanofluid with
time (days)
Figure 4 Particle size distribution of 500 SSA 0.05 wt% kerosene- GNP nanofluid with
time (days)
Figure 8 Thermal conductivity of 750 SSA kerosene- GNP nanofluid with time (days)
0.2wt %
Figure 12 Variation in ratio of heat transfer coefficient of nanofluid to base fluid with
Figure 14 Variation in Merit number with Reynolds number for various particle size and
concentrations of nanofluids
37
Figure 15 Variation in heat transfer coefficient with pressure drop for various particle
List of Tables
temperature
38
Table 1
39
Table 2
k 5
µ 4
Re 4.9
Pr 7.1
h 6.8
ΔP 5
Merit
8.5
Number
f 5.2
40
Table 3
Measurement
Weight % SSA (g/mm2) Temperature, ⁰C Re
parameters
Thermal 0.005, 0.02,0.05,
750, 500, 300 20, 30, 50, 70 --
conductivity 0.1, 0.2
0.005, 0.02,0.05,
Viscosity 750, 500, 300 20, 30, 50, 70 --
0.1, 0.2
Convective 6700-
0.005, 0.05, 0.2 750, 500, 300 ---
heat transfer 28000
Friction 6700-
0.005, 0.05, 0.2 750, 500, 300 ----
factor 28000
41
Table 4
42
Table 5
Concentration Percentage
GNP size
(wt%) increment
300 0.005 0-8
300 0.05 8-12
300 0.2 11-13
500 0.005 2-10
500 0.05 10-15
500 0.2 15-18
750 0.005 3-11
750 0.05 14-18
750 0.2 16-21
Highlights
43
Figure1
Fig. 1
Figure2
Fig. 2
Figure3
10
1st day
8
30th day
Intensity (%)
6 60th day
4 Z-Average (d,mm)
1st day 184.6
30th day 190.1
2 60th day 192.4
0
0 500 1000 1500 2000
Size (d, nm)
Figure4
14
1st day
12
30th day
10 60th day
Intensity (%)
8 Z-Average (d,mm)
1st day 442.2
30th day 456.8
6 60th day 463.0
0
0 500 1000 1500 2000
Size (d, nm)
Figure5
1.3
1.1
SSA 750
1 SSA 500
SSA 300
0.9
0 0.05 0.1 0.15 0.2 0.25
Φ (wt%)
Figure6
1.3
SSA 750
1.35
0.1350
0.1000
0 10 20 30 40 50 60 70 80 90 100
Number of Days
Figure9
1.08
1.04
SSA 750
1.02 SSA 500
SSA 300
1
0 0.05 0.1 0.15 0.2 0.25
Φ (wt%)
Figure10
1.8
Kerosene
1.6 SSA 750
0.8
0.6
10 30 50 70
Temperature ⁰C
Figure11
5000
4500
4000
3500 SSA 300, 0.2%
SSA 300, 0.05%
h (w/m2K)
3000 SSA 300, 0.005%
2500 SSA 500, 0.005%
2000 SSA 500, 0.05%
SSA 500, 0.2%
1500 SSA 750, 0.05%
1000 SSA 750, 0.005%
500 SSA 750, 0.2%
Pure Kerosene
0
0 10000 20000 30000 40000 50000
Re
Figure12
2
1.8
SSA 300, 0.2%
1.6
SSA 300, 0.05%
1.4
SSA 300, 0.005%
1.2
hnf/hbf
SSA 500, 0.2%
1
SSA 500, 0.05%
0.8
SSA 500, 0.005%
0.6 SSA 750, 0.2%
0.4 SSA750, 0.05%
0.2 SSA750, 0.005%
0
0 10000 20000 30000 40000 50000
Re
Figure13
0.05
Current Correlation
0.045 SSA 300 0.2%
SSA 500 0.2%
0.04
SSA 750 0.2%
Friction Factor
0.035 SSA 300 0.05%
SSA 500 0.05%
+11%
0.03 SSA 750 0.05%
SSA 300 0.005%
0.025
- 11%
SSA 500 0.005%
1.6
Merit Number
1.2
SSA 750 0.05%
1 SSA 300 0.2%
0.8 SSA 500 0.2%
SSA 750 0.2%
0.6
0.4
0 20000 40000
Re
Figure15
5000
4500