Heat Mass Transfer (2016) 52:1693–1706

DOI 10.1007/s00231-015-1686-0

ORIGINAL

An empirical investigation on thermal characteristics

and pressure drop of Ag‑oil nanofluid in concentric annular tube
A. A. Abbasian Arani1 · H. Aberoumand2 · S. Aberoumand2 · A. Jafari Moghaddam3 ·
M. Dastanian4

Received: 30 May 2014 / Accepted: 17 September 2015 / Published online: 7 October 2015
© Springer-Verlag Berlin Heidelberg 2015

Abstract In this work an experimental study on Silver- k Thermal conductivity (W/m K)

oil nanofluid was carried out in order to present the laminar cp Specific heat (J/kg K)
convective heat transfer coefficient and friction factor in a D Round tube diameter (m)
concentric annulus with constant heat flux boundary con- Dh Concentric annular tube hydraulic diameter (m)
dition. Silver-oil nanofluid prepared by Electrical Explo- h Convective heat transfer coefficient (W/m2 K)
sion of Wire technique with no nanoparticles agglomera- h̄ Mean convective heat transfer coefficient (W/m2 K)
tion during nanofluid preparation process and experiments. L Length of the tube (m)
The average sizes of particles were 20 nm. Nanofluids Nu Nusselt number
with various particle Volume fractions of 0.011, 0.044 and P Tube cross section perimeter (m)
0.171 vol% were employed. The nanofluid flowing between Pe Peclet number
the tubes is heated by an electrical heating coil wrapped q″ Heat flux (W/m2)
around it. The effects of different parameters such as flow T Temperature (K)
Reynolds number, tube diameter ratio and nanofluid parti- Re Reynolds number
cle concentration on heat transfer coefficient are studied. X Distance from entrance of tube (m)
Results show that, heat transfer coefficient increased by P Pressure drop along the test section (Pa)
using nanofluid instead of pure oil. Maximum enhance-
ment of heat transfer coefficient occurs in 0.171 vol%. In Greek symbols
addition the results showed that, there are slight increases µ Dynamic viscosity (Pa.s)
in pressure drop of nanofluid by increasing the nanoparticle ρ Density (kg/m3)
concentration of nanofluid in compared to pure oil. ψ Weight concentration

List of symbols Subscripts

f Darcy–Weisbach friction factor bf Base fluid
nf Nanofluid
I Inlet
* H. Aberoumand m Mean fluid bulk temperature
h.aberoumand@gmail.com s Surface
1
Mechanical Engineering Department, University of Kashan,
Kashan, Iran
2
Department of Mechanical Engineering, Islamic Azad 1 Introduction
University, Takestan Branch, Takestan, Iran
3
Aerospace Engineering Department, K.N. Toosi University Miniaturization of heat transfer systems on the one hand
of Technology, Tehran, Iran and increase in heat generation of new equipments on the
4
Mechanical Engineering Department, University of Sistan & other hand, necessitates the needs to increase the heat trans-
Baluchestan, Zahedan, Iran fer in a short time with a high rate. Many studies have been

13
1694 Heat Mass Transfer (2016) 52:1693–1706

done to increase the heat transfer from equipments which 70 °C. Kim et al. [9] used two different nanofluids in their
among them, increasing heated surfaces (fins), vibration of experiments in the laminar and turbulent flow regimes. In
heated surfaces, injection or suction of fluid and etc. can be laminar flow regime γ-Al2O3–water nanofluids had a heat
noted [1, 2]. These methods can barely afford heat transfer transfer enhancement around 14 % whereas the amorphous
demands in process including electronic chips, laser sys- carbonic nanofluid showed enhancement around 7 %. In
tems and equipment with high energy consumption. There- turbulent flow regime, γ -Al2O3–water nanofluids had an
fore, there is an urgent need for new and innovative ideas to increase around 20 % and the amorphous carbonic water
increase the heat transfer rate. Nanofluid technology offers nanofluid showed no considerable enhancement.
high potential for the development of cooling systems with Mayer et al. [10] investigated the convective heat trans-
high performance, in small size and economical considera- fer and pressure drop of multi-walled carbon nanotubes
tion. Nanofluids are engineered by suspending nanoparti- flowing through a straight horizontal tube. They reported a
cles with average sizes below than 100 nm in traditional decrease in heat transfer coefficient when compared with
heat transfer fluids such as water, oil, and ethylene glycol. water. Numerical analysis of laminar flow heat transfer of
A very small amount of guest nanoparticles, when dis- Al2O3-ethylene glycol and Al2O3-water nanofluids in tube
persed uniformly and suspended stably in host fluids, can has been done by Palm et al. [11] and Roy et al. [12]. They
provide dramatic improvements in the thermal properties of observed an increase in wall shear stress with volume frac-
host fluids. tion and Reynolds number. Theoretical analysis for tur-
Lee et al. [3] suspended CuO and Al2O3 nanoparticles bulent flow has been presented by Sarma et al. [13]. They
with two different base fluids, water and ethylene glycol compared the Nusselt number predicted from theory with
(EG), and obtained four combinations of nanofluids, CuO that calculated from their experimental data of Al2O3-water
in water, CuO in EG, Al2O3 in water and Al2O3 in EG. nanofluid for 0.5 % volume fraction. Recent developments
Their experimental results showed that nanofluids have in heat transfer enhancements with nanofluids are summa-
substantially higher thermal conductivities than the base rized in review articles by Kakaç and Pramuanjaroenkij
fluids. For example they reported that suspension of 4 % [14] and Wang and Majumdar [15].
volume fraction of 35 nm CuO particles in ethylene gly- Razi et al. [16] investigated pressure drop and ther-
col shows 20 % increase in thermal conductivity. Duang- mal characteristics of CuO–oil nanofluid in laminar flow
thongsuk and Wongwises [4] carried out an investigation regime in flattened tubes with the heights of 11.5 mm (as
on comparison of the effects of measured and computed round tube), 9.6, 8.3, 7.5 and 6.3 mm under constant heat
thermophysical properties of nanofluids on heat transfer flux boundary condition. Their results showed that the heat
performance. Their results showed that thermophysical transfer performance is improved as the tube profile is flat-
properties predicted by theoretical models had almost the tened. Flattening the tube profile resulted in pressure drop
same accuracy as the experimental measured properties for increasing. Also they showed that, Nanofluids have better
the Nusselt number calculation. heat transfer characteristics when they flow in flattened
One of the other advantages of using nanofluid is higher tubes rather than in the round tube. Compared to pure oil
heat transfer coefficient of nanofluids in comparison with flow, Maximum heat transfer enhancement of 16.8, 20.5
the base fluid. Xuan and Li [5] demonstrated the enhance- and 26.4 % is obtained for nanofluid flow with 2 Vol% con-
ment up to 35 % for the turbulent forced convective heat centration inside the round tube and flattened tubes with
transfer coefficient of Cu-water nanofluid with nanoparticle internal heights of 8.3 and 6.3 mm, respectively.
volume fraction equal to 2.5 %. According to their report, Because of the lack of experimental works on annu-
the enhancement is lower for lower nanoparticles volume lar tubes and oil as base fluid in comparison with other
fractions. Lai et al. [6] reported enhancement of Nusselt reported studies in literates, this experimental study is
number about 8 % for Al2O3-water nanofluid with nanopar- done. In addition, since the stability of prepared nanofluids
ticle volume fraction of 1 % and size of 20 nm when the have undeniable and essential role in the results of experi-
flow regime was laminar. Saha and Langille [7] conducted mental researches, the novel one step technique known as
a series of experiments with water to study the effect of Electrical Explosion of Wire (E.E.W) is applied for prepa-
full length and short length strip tape inserts on heat trans- ration of utilized nanofluids. The aim of this study is empir-
fer enhancements. They observed no significant decrease ical investigation of heat transfer and pressure drop in con-
in heat transfer coefficient compared to full length tapes. centric annular tube at low concentration in laminar flow.
However, the pressure drop is quite more. Yang et al. [8] In this paper we intended to experimentally investi-
did experiments with graphite-water nanofluids under gate the effectiveness of passive enhancement technique
laminar flow regime. For a 2.5 Vol% they experienced an by mainly focusing on studying the additional nanopar-
increase in heat transfer coefficient of 22 % over the base ticles effect in the laminar flow regime. The heat transfer
fluid at a temperature of 50 °C and 15 % at a temperature of and pressure drop as well as heat transfer coefficient and

13
Heat Mass Transfer (2016) 52:1693–1706 1695

friction factor were studied experimentally for the fluid sensors and electronic controlling system. The underliquid
flow and heat transfer in tubes in the Reynolds number explosion gives us the fully dispersed and stabilized metal-
range from 50 to 200, three sample of nanoparticles weight lic nanofluids in arbitrary liquid solution. In fact, the pro-
fraction of Ag in oil as base fluid under the uniform wall duction, dispersion and stabilization is done simultaneously
heat flux boundary condition. [17, 18]. Figure 1 shows a schematic of E.E.W method.
In fact, the opportunity of profit from both velocity As mentioned before, utilized Nanofluids, is made of
profile distortion induced by the wall curvature and of the E.E.W method. PNC1K device is an electrical explosion by
heat transfer coefficient enhancement due to adding Ag placing electrodes in liquid media. Nanopowder produc-
nanoparticles to oil could provide an attractive answer for tion and distribution carried out simultaneously (Fig. 2). It
great demand in practical application. The investigation is necessary to mention that in this process, liquid phase
was started by preparing the nanofluid. Thermal conductiv- consisting of deionized water (DW), oil, glycerin, alcohol,
ity and dynamics viscosity were measured before using the acetone, ethylene glycol, and hydrogen peroxide (H2O2)
nanofluid in experimental setup. The temperature as well could be used. There is only one restriction on the use of
as pressure drop across the tube were measured during the liquid explosives and it is non-liquid electrolyte. It is neces-
experimental test for steady state condition with constant sary to explain that, another special feature of this system is
heat flux condition. The average friction factor and ther- the possibility of adding a surfactant to the liquid. So, the
mal heat transfer coefficient were then evaluated. In each nanofluid produced with this method, save primary distribu-
section a complete discussion about obtaining results were tion for a long time. Among all of the existing methods in
presented. the production of metal nanoparticles, electrical explosion
method is the most economical and industrial method [19].

2 Nanofluid preparation

There are two methods for nanofluids preparation: one step

method and two step method. Two steps method is dispers-
ing the nanoparticles in base fluids after prepare the dry
nanoparticles individually. The one-step method is to dis-
perse the nanoparticles in the base fluids at the same time
with the synthesis of the nanoparticles.
In comparison with the two-step method, the one-step
method has many benefits, such as less contamination of
the particle, smaller particle size and high stability. Not
only the one step method is expensive, but also the volume
fraction of nanofluids is lower than the two step method.
Moreover, almost all of the nanotechnology companies Fig. 1  Schematic of E.E.W method
do not have a capacity to product nanofluids in large scale
in quantities. Electrical Explosion Wire (E.E.W) as a one
step method has so many differences relative to other one
step methods. In this Method by applying extra high elec-
tric voltage and current, the primary bulk wire is converted
into the nanoparticles via pulse explosive process. It is
worth to state that the whole process is one continuously
and that is the main advantage of this technique. Among all
physical methods developed for nanoparticle production,
the Plasma technology is the most economic method. By
this technique any wired conductive material can be trans-
formed to the nanoparticles. Low cost, high efficiency and
environmental friendly are other advantages of this tech-
nology. As mentioned before, the explosion of wire can be
applied in gas or liquid media. Pure metal, oxides and also
nitride can be obtained in electrical explosion of wire in gas
media via controlling the chamber atmosphere. The ratio of
the gas mixture in the chamber can be controlled by precise Fig. 2  PNC1K device

13
1696 Heat Mass Transfer (2016) 52:1693–1706

Table 1  Characteristic operation of PNC1K for this study

Model PNC1K

Output voltage 0.5–1 kV

Input power 1P 220VAC 500 W
Shot period 1–5 s
Max. diameter wire 0.25 mm
Exploding length wire 1–5 mm
Output wire Ag
Particle size, average 15 nm

Fig. 4  Pure oil and Ag-oil nanofluid at 0.011 Vol% prepared by

E.E.W technique

that it could provide easily demanded flows for perform the

experiments, a nanofluid reservoir, a gate valve, a non return
valve, test section, RTD PT100 temperature sensors, MPX-
V5004DP pressure sensor, data analyzer USB 4716 and a
three way valve. Test section has built of two copper tubes.
Inner tube has 1500 mm length and its inner diameter is
6.35 or 12.7 and 1.0 mm thickness. Outer tube has 1500 mm
length and 25.4 mm inner diameter and 1.3 mm thickness.
Fig. 3  TEM image of Ag nanoparticles An electrical wire coil wrapped around the outer copper
tube, which links to AC power supply. Then the outer cop-
per tube was covered by the k-flux insulator. Two calibrated
One of the greatest advantages of this method is the temperature RTD PT100 sensors are installed in entrance
ability to produce metal and oxide nanopowders and metal and exit the test section for measurement inlet temperature
oxide nanofluids of a wide range of metal nanoparticles. and outlet temperature of fluids. Nine RTD PT100 sensors
In fact, any metal that can be made of thin wire; it will be were employed to measure the wall temperature of the test
possible to produce metal nanoparticles. For made applied section and all of them welded optionally at 3, 7, 15, 25, 45,
Ag-oil nanofluid in this study, PNC1K device and thin wire 100, 120, 130, 150 cm of axial distance. Every one of the
were ready according to Table 1. PT100 sensors have been calibrated by portable program-
Ag–H2O, Ag–Glycerin, Ag–oil, Cu–H2O, Cu–oil, mable USB 4716 and their accuracy were found 0.1 °C.
Molybdene–H2O, Au–H2O, Brass–H2O, Stainless Steel– To measure the pressure drop MPX-V5004DP sensor with
H2O, Nickel Curium–H2O, Al–H2O and Al–oil nanofluids an uncertainty of ±13 Pa (±0.01 cmHg) is used and all of
could be obtained from PNC1K device more stabilized. the sensors and pump connected to USB 4716. A 1 l ves-
From the Fig. 3, it is crystal clear that the Ag nanoparticles sel and stopwatch accurate to 0.001 s is used to measure the
are dispersed well and the mean diameter of the Ag nan- flow rate. Heat exchanger unit settle in after the test section.
oparticles is around 20 nm. Pure oil and Ag-oil nanofluid Fluid turn upside down to the pump from the fluid reservoir,
at 0.011 Vol% (0.022 Vol%) prepared by E.E.W technique then it pumps to the test section and bulk temperatures and
have shown in Fig. 4. wall temperatures measure with sensors. The experimental
data for oil and nanofluids are listed in Table 2.

3 Experimental apparatus
4 Data analysis
In order to study on the convective heat transfer and pres-
sure drop in the concentric annulus tube, the accurate 4.1 Data collection
experimental setup designed and assembled (Fig. 5). The
experimental setup chiefly includes a flow loop, including To calculate the convective heat transfer coefficient and
a pump (CALPEDA, made in Italy, with three-speed motor) other heat transfer characteristics, all of the rheological

13
Heat Mass Transfer (2016) 52:1693–1706 1697

Fig. 5  Schematic of experimental setup

properties of both pure oil and nanofluid must be deter- x is the distance from tube inlet. For calculate the convec-
mined. To determine the specific heat capacity (cp) in dif- tive heat transfer coefficient:
ferent temperature of nanofluids and pure oil, a differen-
q′′
tial scanning calorimeter (DSC F3 Maia, manufactured h(x) = (5)
by NETZSCH-Germany) was used. SVM3000 devise was (Ts (x) − Tm (x))
used to determine the density in different temperatures and
h(x)Dh
volume fractions of nanofluid. The thermal conductivity Nu(x) = (6)
and viscosity of nanofluid and pure oil were measured by k
KD2 thermal properties analyzer and Brookfield viscom- where Dh is hydrolic diameter and k is the thermal conduc-
eter (DV-II + Pro Programmable Viscometer), respectively. tivity. All of the properties measured in average tempera-
As Fig. 6 with consider the energy balance on a differen- ture as following:
tial control volume of the fluid inside the tube:  
Tb,in + Tb,o
dTb q′′ P P
Tm = (7)
2
= = h(Ts − Tb ) (1)
dx ṁcp ṁcp
where ṁ and P are mass flow rate and perimeter of the tube,
Dh = Do − Di (8)
respectively. Ts and Tb are the surface temperature and bulk where, Do and Di are outer diameter and inner diameter
temperature respectively. cp is specific heat. For the heat respectively. Reynolds and Prandtl number are defined as
flux on outer surface as boundary condition, the convective follow:
heat transfer is easy to calculate. Due to q′′ is independent
4ṁ
to x: Re = (9)
πDh µ
q′′ = qconv (P · L) (2)
µcp
Also we have: Pr = (10)
k
qconv = ṁcp (Tmo − Tmi ) (3) Also the mean heat transfer coefficient and mean Nusselt
So by integrate of x = 0 and constant heat flux: number are calculate as Eqs. 11 and (12), respectively.
q′′ Px L
1
Tm (x) = Tmi (x) + (4) h̄ = h(x)dx
ṁcp L (11)
0

13
1698 Heat Mass Transfer (2016) 52:1693–1706

Table 2  Experimental data for No. ϕ (%) H (W/m2K) Nu Re Pr Q (m3/s)

oil and nanofluids
1 0 72.88 10.44 21.62 394.6 0.000010
2 0 78.41 11.23 32.03 405.9 0.000015
3 0 84.00 12.04 42.13 411.5 0.000020
4 0 86.03 12.33 52.22 414.9 0.000025
5 0 91.02 13.04 68.4 418.1 0.000033
6 0 92.43 13.24 78.50 419.5 0.000038
7 0 97.34 13.94 98.7 421.3 0.000048
8 0 98.58 14.12 118.9 422.5 0.000058
9 0 100.7 14.43 139.1 423.4 0.000068
10 0.022 77.30 10.56 19.87 417.7 0.000010
11 0.022 83.24 11.48 29.08 428.1 0.000015
12 0.022 86.93 11.99 36.47 432.3 0.000019
13 0.022 89.98 12.40 43.86 435.0 0.000023
14 0.022 94.21 12.98 56.61 438.1 0.000029
15 0.022 95.80 13.22 62.35 439.0 0.000033
16 0.022 98.48 13.36 73.45 440.5 0.000039
17 0.022 98.74 13.39 79.00 441.3 0.000042
18 0.022 104.1 14.13 104.9 442.1 0.000056
19 0.022 105.8 14.40 117.8 443.1 0.000063
20 0.022 107.8 14.57 130.9 443.9 0.000070
21 0.022 109.0 14.90 136.4 444.5 0.000072
22 0.044 81.34 10.68 17.49 418.0 0.000010
23 0.044 87.52 11.50 25.61 428.1 0.000015
24 0.044 92.05 12.10 35.37 433.8 0.000021
25 0.044 94.92 12.48 43.52 436.6 0.000026
26 0.044 98.10 12.94 50.03 438.2 0.000030
27 0.044 98.80 12.97 56.54 439.3 0.000034
28 0.044 101.7 13.37 64.69 440.5 0.000039
29 0.044 103.7 13.63 69.58 441.0 0.000042
30 0.044 105.2 13.83 79.35 441.9 0.000048
31 0.044 105.8 13.90 85.87 442.4 0.000052
32 0.044 108.3 14.23 95.65 443.6 0.000058
33 0.044 110.8 14.56 107.1 443.6 0.000065
34 0.044 112.2 14.74 116.8 444.0 0.000071
35 0.044 113.2 14.88 128.2 444.4 0.000078
36 0.044 114.1 15.00 136.4 444.6 0.000083
37 0.176 82.88 10.62 12.06 417.7 0.000010
38 0.176 86.49 11.07 17.64 428.3 0.000016
39 0.176 88.78 11.38 23.25 433 0.000020
40 0.176 91.90 11.83 28.85 436.1 0.000025
41 0.176 94.05 12.03 34.46 438.1 0.000030
42 0.176 97.00 12.42 40.07 439.6 0.000035
43 0.176 99.03 12.67 45.68 440.7 0.000040
44 0.176 101.0 12.94 51.29 441.5 0.000045
45 0.176 103.1 13.17 56.9 442.2 0.000050
46 0.176 104.1 13.32 62.51 442.7 0.000055
47 0.176 107.5 13.75 73.73 443.6 0.000065
48 0.176 109.9 14.09 84.95 444.2 0.000075
49 0.176 112.1 14.30 96.18 444.7 0.000085
50 0.176 112.3 14.27 107.41 445.1 0.000095

13
Heat Mass Transfer (2016) 52:1693–1706 1699

Table 2  continued No. ϕ (%) H (W/m2K) Nu Re Pr Q (m3/s)

51 0.176 115.4 14.69 113.0 445.2 0.00010
52 0.176 116.9 14.89 124.2 445.4 0.00011
53 0.176 118.1 15.10 135.5 445.7 0.00012

 n  2  21
 ∂R
UR = UVi (16)
∂Vi
i=1

where UR and UVi are the uncertainties associated with the

parameter R and independent variables (Vi), respectively.
In addition, n is the number of the independent variables.
This method also has been used by other researchers such
as Fakoor Pakdaman et al. [21]. Hence, the existing inher-
Fig. 6  Energy balance on a differential control volume
ent errors in measuring parameters such as the inner tube
diameter, length of the test section, mass flow rate, head
Table 3  The maximum values of uncertainty relating to Nusselt of mercury manometer, slop of mercury manometer, inner
number, convective heat transfer coefficient, friction factor and pres- wall temperature and fluid average temperature are taken
sure drop
into account using the following equations.
Nu hm f ∆p
 2  2  2  21
±5.4 % ±4.3 % ±6.1 % ±3.1 % δ(Nunf )m Q̇m δL δ�T
= + + (17)
(Nunf )m Q̇m L �T

 2  1
δDi 2
 2 
δ(hnf )m δ�T 2 2
 
δ Q̇m δL
h̄Dh = + + +
N ū = (12) (hnf )m Q̇m Di L �T
k
Also the Nusselt Number calculated by Eq. 13: (18)

n
Nun  2  2  21
δh 2 δ(sin θ ) 2 δDi 2
      
N ū = δf δ Q̇ δL
n (13) = + + 5 + 2 +
1
f h (sin θ ) Di Q̇ L
(19)
The friction factor was calculated as follow:

�p  2  2  21
f =   2  δ(�p) δh δ(sin θ )
1
ρ v2 (14) = + (20)
D �p h (sin θ )

The maximum uncertainty corresponding to the above

Pe = Re. Pr (15) mentioned parameters are evaluated and presented in
Table 3.
The uncertainty in experimental study by Anoop et al.
4.2 Uncertainty [22], were evaluated and to be around 2.45 % for Nusselt
number in experimental investigation of diameter effect of
All parameters used in the experimental study for calculat- Al2O3-water Nanofluid. Their study was performed for alu-
ing the major heat transfer characteristics such as Nusselt mina–water with average particle sizes of 45 and 150 nm
number, convective heat transfer coefficient, friction factor and particle concentrations for 1, 2, 4 and 6 Vol% The
and pressure drop has uncertainties due to inherent errors. uncertainty in Nusselt number was within 3 % in another
In this work, Coleman and Steele uncertainty method [20] experimental work carried out by Wen and Ding [23], who
was employed for this purpose. focused on the entrance region under laminar flow regime,

13
1700 Heat Mass Transfer (2016) 52:1693–1706

for Al2O3-water nanofluid with various concentrations and Table 4  The values of Am and γm2 [28]
constant heat flux boundary conditions. In an experimental M Am γ2m
investigation, Fotukian and Nasr Esfahany [24] quoted 4 %
uncertainty in the measurement of Nusselt number on the 1 0.007630 25.67
turbulent flow regime of convective heat transfer and pres- 2 0.002053 83.86
sure drop for dilute Al2O3-water nanofluids inside a circular 3 0.000903 174.2
tube. Duangthongsuk and Wongwises [25, 26] conducted 4 0.000491 296.5
an experimental investigation on the TiO2–water nanoflu- 5 0.000307 450.9
ids flowing in a horizontal concentric tube under a turbu-
lent flow regime at low concentrations. They reported the
uncertainty of Nusselt number and pressure drop around 5
and 6 %, respectively. In an experimental work by He et al.
[27] in fluid flow and heat transfer behaviour of the TiO2-
water nanofluids flowing upwards in a straight pipe under
both laminar and turbulence/transition flow regime, the
uncertainty reported about 3 % for Nusselt number.
In the present work, the evaluated uncertainties asso-
ciated from the flow parameters are reported in Table 3,
where the deviation is within an acceptable range.

5 Result and discussion

5.1 Validation check

In order to verify the dependability and correctness of the

experimental setup, initial experiments are performed with
oil as the working fluid. The experiments are conducted
Fig. 7  Comparison between experimental data and predicted theo-
within the Reynolds number of 140. Because of low Reyn- retical values from equation [28] in straight tube
olds number and high Prandtl number of oil, hydrodynami-
cally fully developed laminar flow and thermal entrance
region for flow, are assumed respectively, for theoreti- numbers at nine axial locations (3, 7, 15, 25, 45, 100, 120,
cal calculations (xt/D < 0.05 RePr). The values of Nusselt 130, 150 cm of axial distance), the average Nusselt numbers
numbers that are measured experimentally are compared are obtained using Eqs. 11, 12 and 13. Figure 7 shows the
with the values obtained by the following theoretical equa- experimental results of mean Nusselt number for pure base
tion presented in Eq. 21, [28]: oil in the circular tube at different Peclet number measured
−1  by Eq. 15 in comparison with those obtained by the theo-
∞

1  exp(−γm2 x ∗ ) µs −0.14 retical Eq. 21. This figure indicates that measurements agree

1
Nux = − . (21) well with the predictions of this equation. Table 5 shows the
Nu∞ 2 γm4 Am µm
m=1
values of experimental data and predicted theoretical values
In Eq. 21, Nu∞ = 4.364 and Nux is the local Nusselt number and relative errors between them.
at the distance of x from inlet of the test section, x* = (2x/D) Also, after measuring the pressure drop experimen-
RePr is a non-dimensional parameter, γm is the necessary tally, it compared with the pressure drop obtained from the
eigenvalue and Am is a constant value. The values of Am and Eq. 14, In which, ρ is measured at the average values of
γm2 are displayed in Table 4. For liquids, for instance oil, inlet and outlet temperatures and f = 64/Re. Figure 8 shows
where viscosity variation is dependable for its rheological the comparison between theoretical values for pressure
and thermal behavior, it is found that the correction coef- drop and experimental data along the circular tube versus
ficient, (μs/μm)−0.14, defined in Eq. 21 is often the best esti- Reynolds number. As it can be seen from Fig. 8, the experi-
mate. In this coefficient, μs is calculated at the surface tem- mental pressure drop data for pure oil and the theoretical
perature, while μm is evaluated at the bulk temperature. This predictions using Eq. 14 have a good agreement. By setting
solution is used for obtaining local Nusselt number of a fluid up an accurate and confident set up, the heat transfer and
flow with temperature varying viscosity inside round tube pressure drop characteristics of oil-based silver nanofluids
under constant heat flux condition. Having the local Nusselt flowing inside the concentric annular tubes are investigated

13
Heat Mass Transfer (2016) 52:1693–1706 1701

Table 5  the values of experimental data and predicted theoretical

values and relative errors between them
Num (experimental) Num (theoretical) Error (%)

11.10 11.74 5.76

11.53 11.89 3.12
12.10 12.20 0.83
13.17 12.80 2.80
13.30 13.50 1.50
13.26 13.82 4.22
13.90 14.01 0.79
14.41 14.31 0.69
14.98 14.53 3.00
15.00 14.69 2.10
15.10 14.89 1.30
15.26 15.02 1.50
15.53 15.20 2.10
15.75 15.50 1.60
Fig. 9  Variation of mean heat transfer coefficient versus Reynolds
number at different volume fraction in concentric annular tube

and the nanofluid with 0.011 % (0.022 Vol%), 0.044 %

(0.044 Vol%) and 0.171 % (0.176 Vol%) volume fraction
flow inside concentric annulus tube at constant heat fluxes
of 204 watt shown in Fig. 9. Based on the results, for cer-
tain Reynolds number, the convective heat transfer coeffi-
cient (h) of nanoparticles suspended in base oil is higher
than base oil. This enhancement noticeably is reliant on
the concentration of nanoparticles. For instance at Reyn-
olds number about 73.45, the heat transfer coefficients are
7.33, 13.7 and 17.32 % greater than those of the base fluid
when the nanoparticle concentration are 0.011 %, 0.36 %
and 0.171 Vol%, respectively. In point of fact, the nanofluid
with suspended nanoparticles enhances the thermal con-
ductivity of the mixture and a large energy exchange pro-
cess resulting from the anarchic movement of nanoparticle
decrease boundary layer thickness and delay in boundary
layer development as claimed by previous research works
Fig. 8  Comparison between theoretical values for pressure drop and
experimental value along the straight tube versus Reynolds number [23, 29]. It has seen that slightly enhancement in heat trans-
fer coefficient with increasing nanoparticle concentration.
A study on nanofluid’s thermal conductivity organized by
experimentally for laminar flow regime under constant Buongiorno et al. [30] reveals that the nanofluid’s ther-
heat flux boundary condition. It is important to know that, mal conductivity enhancement is noticeable only when the
obtained experimental data for heat transfer and pressure nanoparticle concentration is higher than 1 Vol%. There-
drop for each two specific cases of runs have not measured fore, higher Nusselt number for nanofluid with low nano-
under exactly the same Reynolds numbers. This is because particle concentration is due to the heat transfer coefficient
of the viscosity of oil-based nanofluid that is so dependent enhancement.
on fluid temperature and particle volume fraction. Figures 10 and 11 show the results for Ag-oil nanofluid
with concentrations of 0.011, 0.044 and 0.171 Vol% at vari-
5.2 Heat transfer characteristics ous Peclet number at 204 watt on outer tube and laminar
flow in concentric annulus tube. It shows the increments in
Variation of heat transfer coefficient related to Ag-oil Nusselt number and heat transfer coefficient by increasing
nanofluid as a function of Reynolds number for pure oil the Peclet number.

13
1702 Heat Mass Transfer (2016) 52:1693–1706

Fig. 10  Variation of mean heat transfer coefficient versus Peclet Fig. 12  Variation of mean heat transfer coefficient versus Reynolds
number at different volume fraction in concentric annular tube number at different volume fraction in concentric annular tube at
Dh = 0.0127 m

and 0.171 Vol% respectively rather than using pure oil as

base fluid. It may be noted that the improvement in the heat
transfer coefficient is much higher compared to the increase
in thermal conductivity. This conclusion has been referred
by other researchers such as Anoop et al. [22] and Fara-
jollahi et al. [31]. Anoop et al. [22] reported that in their
experimental work for 45 nm based nanofluid the enhance-
ment in heat transfer coefficient is about 25 % whereas the
thermal conductivity for the same has increased only by
6 %. Similarly for 150 nm based nanofluids the heat trans-
fer coefficient increase is about 11 %, whereas the thermal
conductivity enhancement was around 4 % only.
As shown in Fig. 12, that it shows the variation of heat
transfer coefficient, changes with Reynolds number for dif-
ferent concentrations of Nanoparticles. It is clear that by
increasing the dimensionless diameter or reduce hydraulic
diameter, heat transfer coefficient increases with increas-
Fig. 11  Variation of mean Nusselt versus Peclet number at different
volume fraction in concentric annular tube ing Reynolds number due to increasing the fluid velocity.
More ever, by intensify the concentration of nanoparticles
because of the increase of chaos and movement of nano-
The enhancement of Nusselt number and heat transfer particles, also improving the thermal conductivity, heat
coefficient exhibit that the heat transfer performance of transfer coefficient of nanofluids increase than base fluid.
nanofluid is rely on various parameters such as Nanoparti- With decrease the Hydraulic diameter, due to increase
cles movement, Brownian motion and reduction in thermal heat power per volume, bulk temperature increases and
boundary layer thickness and possible slip condition at the heat transfer coefficient increases. As Shown in Fig. 12 in
walls are other possible reasons for enhancement of heat Reynolds number equal to 39.5, heat transfer coefficient for
transfer coefficient respectively, and the thermal conductiv- concentrations of 0.011, 0.044 and 0.171 %, increase by 8,
ity is not the only main factor. Base on the results, at cer- 17.4 and 24.2 % respectively compared to the base fluid.
tain Peclet number, for example 41,584, heat transfer coef- Due to the pressure drop for a concentric annulus tube with
ficient with adding nanoparticles increase approximately hydraulic diameter of 0.25 was large; the tests did not car-
4.3, 9.9 and 14.45 % for volume fraction of 0.011, 0.044 ried out for this hydraulic diameter.

13
Heat Mass Transfer (2016) 52:1693–1706 1703

It is well known that the enhancement by using of

nanofluids in heat transfer is also associated with signifi-
cant increase in pressure drop or pumping power. Various
thermal performance evaluation criteria, which take into
account both heat transfer enhancement and pressure drop
(pumping power), have been suggested for these enhanced
performance surfaces based on different design constraints
and objectives [32–35]. In fact the original version of this
equation is based on equal pumping power for the rough-
ened and smooth surfaces suggested by Webb and Eckert
[32] that has been most widely used [36–39]. The perfor-
mance ratios used by different researchers for the multitube
heat exchangers are:
qnf hnf
β= = (22)
qbf hbf
The ratio of heat transfer coefficients can, ℎnf/ℎbf, can
replaced by the ratio of the overall heat transfer coeffi-
cients, Unf/Ubf, when heat transfer between two fluids is Fig. 13  Ratio of convective heat transfer coefficient of nanofluid to
the base fluid versus Reynolds number
considered. The heat transfer coefficient ℎnf or Unf is cal-
culated at the nanofluids Reynolds number Renf. It is worth
to note that thermal performance given by above equation the convective heat transfer coefficient, may be formulated
has been suggested for enhanced surfaces when the basic approximately as h = kf/dt, in which kf is thermal conduc-
geometry of the exchanger is kept fixed and the mass flow tivity and dt is thermal boundary layer. Both an increase
rate is variable for fixed pumping power. in kf or/and a decrease in dt enhance the convective heat
Figure 13 shows the ratio of convective heat trans- transfer coefficient. As explained previously increase in
fer coefficient of Ag-oil nanofluid to the base fluid, ver- Reynolds number leads to a decrease in the boundary layer
sus Reynolds number at volume fractions of 0.011, 0.044 thickness. Also an addition of nanoparticles to base fluid
and 0.171 Vol% in concentric annulus tube with constant increases the thermal conductivity of nanofluid and the
heat power of 204 W in the laminar flow regime. Results enhancement increases with increasing particle concen-
indicated that the average heat transfer coefficient in vol- tration. The increase of the thermal conductivity should
ume fractions of 0.011, 0.044 and 0.171 Vol% increase increase the convective heat transfer coefficient. How-
approximately 6.2, 13.77 and 16.3 % respectively, which ever, the increase in particle concentration also increases
causes it to increase the Brownian motion and movement of the fluid viscosity, which should result in an increase in
nanoparticles, also increase irregularity nanoparticles and the boundary layer thickness hence a decrease in the con-
reduce the thickness of the boundary layer. vective heat transfer coefficient. As shown clearly in this
Buongiorno [40] in his investigation presented a relation study, addition of nanoparticles enhances the convective
between heat transfer enhancement and volume fraction heat transfer. These results demonstrate that the positive
and Reynolds. He proposed that the heat transfer increase effect of the thermal conductivity enhancement overcome
due to existence of nanofluid in turbulent flow regime the negative effect of the viscosity increase under the con-
relate to the decrease of viscosity and consequent thin- ditions of this investigation. It is worth to note that a large
ning of laminar sublayer. He said that the laminar sublayer number of researchers believe that particle size has a mar-
is a resistance against the heat exchange between the sur- ginal effect (compared to Reynolds number and volume
face and fluid and its thickness is proportional to the vis- fraction) on the convective heat transfer coefficient under
cosity of sublayer. Heat transfer increase by nanoparticles the conditions of this work. This concept come from the
in laminar sublayer is due to decrease in laminar sublayer reality that nanofluids containing larger nanoparticles have
viscosity. As a consequent, with increasing Reynolds num- a lower thermal conductivity and a higher viscosity, both
ber and nanofluid volume fraction, laminar sublayer gets of two effect should have led to a lower convective heat
thinner and convective heat transfer coefficient increases transfer coefficient. Wen and Ding [41] related this effect to
considerably. the particle migration mechanism. According to this idea,
The effects of volume fraction and nanofluid Reynolds large nanoparticles tend to migrate to the central part of
number may be explained by the macroscopic theory for the pipe, which could result to a particle depletion region
the forced convective heat transfer. Based on this theory with low viscosity at the near wall hence a decrease in the

13
1704 Heat Mass Transfer (2016) 52:1693–1706

boundary layer thickness. Also, small nanoparticles tend to

be uniformly distributed over the pipe cross-section due to
the Brownian motion. Hence, for a given average particle
concentration, the wall region could have a higher solids
concentration and hence a higher viscosity when the flow-
ing nanofluids contain smaller nanoparticles. The combina-
tion of the above two opposite cited effects could have been
consequence for the observed marginal effect of nanopar-
ticle size under the conditions of this investigation. It must
be cited that, the proposed particle migration mechanism is
a hypothesis; further experimental study is needed in order
to verification and confirmation of this conclusion.

5.3 Pressure drop results

It is essential to know that our nanofluid is practical to

utilize in the industries. There are two way for drawing
the graphs of friction factor or pressure drop. One of them Fig. 14  Variation of pressure drop versus Reynolds number at differ-
is measuring pressure drop while flow rate is fixed, which ent volume fraction in concentric annular tube
the experimental friction factor data shifts to lower Reyn-
olds numbers. This is due to the higher viscosity of the
Silver-oil nanofluids. Since Silver-oil nanofluids exhibit a
higher viscosity than pure oil. Another way is measure the
pressure drop by fixing Reynolds numbers. The friction
factor and pressure drop of Ag-oil nanofluild at differ-
ent volume fractions in concentric annular tube has been
investigated experimentally. Figure 14 shows the variation
of pressure drop versus Reynolds number at different vol-
ume fraction in concentric annular tube. It is found that
the pressure drop increases with increasing the Reynolds
number.
Figure 14 indicated that, there is non-significant increase
in pressure drop of nanofluid with increase the particle
concentration compared to the pure oil flow in concentric
annular tube. In addition, the results show that for the con-
centric annular tube, the maximum pressure drop increases
about 3 % at nanofluid with 0.171 Vol% As the tube pro-
file becomes annulus, the contact area between the flow
and tube surface increases and therefore it can increase the
Fig. 15  Ratio of friction factor of Ag-oil nanofluid to the base fluid
friction drag force considerably and consequently leads to
versus velocity
increase the pressure drop. Also, for annular tubes, the flow
pattern and velocity distribution will change. This can also
lead to walls shear stress increasing which results in pres- 6 Conclusion
sure drop enhancement. The maximum increasing in pres-
sure drop up to 53 % is obtained for the concentric annular Electrical Explosion Wire (E.E.W) as a one step method
tube with the diameter ratio of 0.25 and maximum Reyn- with special characteristics such as low cost, high effi-
olds number of 179.5. ciency and environmental friendly has been used for pro-
Figure 15 indicated the ratio of friction factor of Ag-oil ducing the Ag-oil nanofluids. The values of heat transfer
nanofluid to the base fluid versus velocity. Results show the coefficient of base fluid and Ag-oil nanofluid as a function
average ratio of (fnf/fbf) was about 1.008, 1.015 and 1.025 of Reynolds number at 0.011 % (0.022 Vol%), 0.044 %
for 0.011, 0.044 and 0.171 Vol% Therefore, there is no sig- (0.044 Vol%) and 0.171 % (0.176 Vol%) volume fraction
nificant increase in friction factor for nanofluids. for flow inside concentric annular tube at constant heat

13
Heat Mass Transfer (2016) 52:1693–1706 1705

fluxes boundary condition in laminar flow regime were be longitudinal strip inserts under uniform wall heat flux. J Heat
measured. Based on this results, for all Reynolds number Transfer 124:421–432
12. Roy G, Nguyen CT, Lajoie PR (2004) Numerical investigation
and value of nanoparticle concentration under study the of laminar flow and heat transfer in a radial flow cooling system
convective heat transfer coefficient (h) of nanofluids were with the use of nanofluids. Superlattices Microstruct 35:497–511
higher than base fluid. For higher volume fraction higher 13. Sarma PK, Kedarnath Ch, Sharma KV, Sundar LS, Kishore PS,
heat transfer coefficient was obtained. As an example at Srinivas V (2010) Experimental study to predict momentum and
thermal diffusivities from convective heat transfer data of nano
Reynolds number about 73.45, the heat transfer coefficients fluid with Al2O3 dispersion. Int J Heat Technol 28:123–131
have 7.33, 13.7 and 17.32 % enhancement as compared to 14. Kaka CS, Pramuanjaroenkij A (2009) Review of convective heat
base fluid at nanoparticle concentration of 12, .36. % and transfer enhancement with nanofluids. Int J Heat Mass Transf
0.171 Vol%, respectively. In addition the values of heat 52:3187–3196
15. Liu C, Yang X, Yuan H, Zhou Z, Xiao D (2007) Preparation of
transfer coefficient for oil and Ag-oil nanofluid at various silver nanoparticle and its application to the determination of ct-
Peclet number with same cited above conditions were eval- DNA. Sensors 7:708–718
uated. These results showed that for all Peclet number and 16. Razi P, Akhavan-Behabadi MA, Saeedinia M (2011) Pressure
nanoparticles concentrations a noticeable increment in heat drop and thermal characteristics of CuO–base oil nanofluid lami-
nar flow in flattened tubes under constant heat flux. Int Commun
transfer coefficient was observed as compared with base Heat Mass Transf 38:964–971
fluid. As an example at Peclet number equal to 41,584, heat 17. Alqudami A, Annapoorni S, Shivaprasad Govind SM (2007) Ag–
transfer coefficient of nanofluids increase approximately Au alloy nanoparticles prepared by electro-exploding wire tech-
4.3, 9.9 and 14.45 % for volume fraction of 0.011, 0.044 nique. J Nanopart Res. doi:10.1007/s11051-007-9333-4
18. Tien DC, Liao CY, Huang JC, Tseng KH, Lung JK, Tsung
and 0.171 Vol% respectively rather than base fluid. This TT, Kao WS, Tsai TH, Cheng TW, Yu BS, Lin HM, Stobinski
investigation showed that there is non-significant increase L (2008) Novel technique for preparing a nano-silver water
in pressure drop of nanofluid with increase the particle con- suspension by the arc-discharge method. Rev Adv Mater Sci
centration compared to base fluid. 18:750–756
19. www.pnf-co.com
20. Coleman HW, Steele WG (1989) Experimental and uncertainty
analysis for engineers. Wiley, New York
21. Fakoor Pakdaman M, Akhavan-Behabadi MA, Razi P (2012)
References An experimental investigation on thermo-physical properties
and overall performance of MWCNT/heat transfer oil nanofluid
1. Bergles AE (1973) Recent development in convective heat trans- flow inside vertical helically coiled tubes. Exp Therm Fluid Sci
fer augmentation. Appl Mech Rev 26:675–682 40:103–111
2. Thome JR (2004) Engineering Data Book III, Wolverine Tube, 22. Anoop KB, Sundararajan T, Das SK (2009) Effect of particle
Inc size on the convective heat transfer in nanofluid in the develop-
3. Xuan Y, Li Q (2003) Investigation on convective heat transfer ing region. Int J Heat Mass Transf 52:2189–2195
and flow features of nanofluids. J Heat Transfer 125(1):151–155 23. Wen D, Ding Y (2004) Experimental investigation into convec-
4. Duangthongsuk W, Wongwises S (2010) Comparison of the tive heat transfer of nanofluids at the entrance region under lami-
effects of measured and computed thermophysical properties of nar flow conditions. Int J Heat Mass Transf 47:5181–5188
nanofluids on heat transfer performance. Exp Thermal Fluid Sci 24. Fotukian SM, Nasr Esfahany M (2010) Experimental inves-
34:616–624 tigation of turbulent convective heat transfer of dilute Al2O3-
5. Lee S, Choi SUS, Li S, Eastman JA (1999) Measuring thermal water nanofluid inside a circular tube. Int J Heat Fluid Flow
conductivity of fluids containing oxide nanoparticles. J Heat 31:606–612
Transfer 121:280–289 25. Duangthongsuk W, Wongwises S (2009) Heat transfer enhance-
6. Lai WY, Duculescu B, Phelan PE, Prasher RS (2006) Convective ment and pressure drop characteristics of TiO2–water nanofluid
heat transfer with nanofluids in a single 1.02-mm tube, Proceed- in a double-tube counter flow heat exchanger. Int J Heat Mass
ings of ASME International Mechanical Engineering Congress Transf 52:2059–2067
and Exposition (IMECE).pp 65-76 26. Duangthongsuk W, Wongwises S (2010) An experimental study
7. Kim D, Kwon Y, Cho Y, Li C, Cheong S, Hwang Y et al (2009) on the heat transfer performance and pressure drop of TiO2-
Convective heat transfer characteristics of nanofluids under lami- water nanofluids flowing under a turbulent flow regime. Int J
nar and turbulent flow conditions. Curr Appl Phys 9(2):119–123 Heat Mass Transf 53:334–344
8. Yang Y, Zhang Z, Grulke E, AndersonW WuG (2005) Heat trans- 27. He Y, Jin Y, Chen H, Ding Y, Cang D, Lu H (2007) Heat transfer
fer properties of nanoparticles-in-fluid dispersions (nanofluids) and flow behaviour of aqueous suspensions of TiO2 nanoparti-
in laminar flow. Int J Heat Mass Transf 48(6):1107–1116 cles (nanofluids) flowing upward through a vertical pipe. Int J
9. Palm SJ, Roy G, Nguyen CT (2004) Heat transfer enhancement Heat Mass Transf 50:2272–2281
in radial flow cooling system-using nanofluid. in: Proceeding of 28. Kays WM, Crawford ME, Weigand B (2005) Convective heat
the ICHMT International symposium on advance computational and mass transfer, Fourth ed., Mc Grow Hill, pp. 104–105 and
heat transfer, Norway, CHT-04, 110–121 pp. 330–335
10. Meyer JP, McKrell TJ, Grote K (2013) The influence of multi- 29. Chen H, Yang W, He Y, Ding Y, Zhang L, Tan C, Lapkin AA,
walled carbon nanotubes on single-phase heat transfer and Bavykin DV (2008) Heat transfer behavior of aqueous suspen-
pressure drop characteristics in the transitional flow regime of sions of titanate nanofluids. Powder Technol 183:63–72
smooth tubes. Int J Heat Mass Transf 58:597–609 30. Buongiorno J, Venerus CD, Prabhat N, McKrel lT, Townsend J,
11. Saha SK, Langille P (2002) Heat transfer and pressure drop Christianson R et al (2009) A benchmark study on the thermal
characteristics of laminar flow through a circular tube with conductivity of nanofluids. J Appl Phy 106:1–14

13
1706 Heat Mass Transfer (2016) 52:1693–1706

31. Farajollahi B, Etemad SGH, Hojjat M (2010) Heat transfer of 37. Thianpong C, Eiamsa-ard P, Eiamsa-ard S (2012) Heat trans-
nanofluids in a shell and tube heat exchanger. Int J Heat Mass fer and thermal performance characteristics of heat exchanger
Transf 53(1–3):12–17 tube fitted with perforated twisted-tapes. Heat Mass Transf
32. Webb RL, Eckert ERG (1972) Application of rough surfaces to 48(6):881–892
heat exchanger design. Int J Heat Mass Transf 15(9):1647–1658 38. D’Agaro P, Comini G (2008) Thermal-performance evaluation
33. Boer KW (1978) Payback of solar systems. Sol Energy of coolant passages with staggered arrays of pin fins. Heat Mass
20:225–232 Transf 44(7):815–825
34. Altfeld K, Leiner W, Fiebig M (1988) Second law optimiza- 39. Elsayed ML, Mesalhy O (2014) Studying the performance of
tion of flat-plate solar air heaters—part I: the concept of net solid/perforated pin-fin heat sinks using entropy generation mini-
exergy flow and themodeling of solar air heaters. Sol Energy mization. Heat Mass Trans 1–12
41(2):127–132 40. Buongiorno J (2006) Convective transport in nanofluids. J Heat
35. Cort´es A, Piacentini R (1990) Improvement of the efficiency of Transfer 128(3):240–250
a bare solar collector by means of turbulence promoters. Appl 41. Wen DS, Ding YL (2005) Effect on heat transfer of particle
Energy 36(4):253–261 migration in suspensions of nanoparticles flowing through min-
36. Wang L, Sunden B (2007) Experimental investigation of local ichannels. Microfluid Nanofluid 1(2):183–189
heat transfer in a square duct with various-shaped ribs. Heat
Mass Transf 43(8):759–766

13