Professional Documents
Culture Documents
n
l
_
k
e
k
c;x
k
e
B
2; x
_
k
c; x
k
e
_
4
k
e
k
c;y
2k
e
_
1 B
2; x
__
k
c; y
k
e
_
_
0 2
where n and n/l are the volume fraction of the nanoparticles and the
complex nanoparticles, respectively. Other parameters in the above
equation are represented by the following relations:
l
abc
a tb tc t
(3)
B
2; y
B
2; z
1 B
2; x
2
(4)
B
2; x
abc
2
Z
N
0
du
_
u a
2
_
_
u a
2
__
u b
2
_
_
u c
2
_
r (5)
k
c; j
k
1
_
1 B
2; j
_
k
1
B
2 ;j
k
2
_
1 B
2; j
_
lk
2
k
1
_
1 B
2; j
_
k
1
B
2; j
k
2
B
2; j
lk
2
k
1
(6)
In these expressions, it is assumed that the complex nanoparticle is
composed of an elliptical nanoparticle of thermal conductivity k
2
with half radii of (a,b,c) and an elliptical shell of thermal conduc-
tivity k
1
with thickness t. Based on Maxwell theory and average
polarization theory he considered the interface effect between the
solid particles and the base uid. The author concluded that the
interfacial shell about the nanotube or nanoparticle has a major
contribution in thermal conductivity of the nanouid. As the
thickness of this interfacial shell becomes larger, the thermal
conductivity of the nanouid increases.
Xue [31] (hereafter we named Xue-2005) developed a new
model for the effective thermal conductivity of CNTs-based
composites relying on Maxwell theory considering the space
distribution and axial effects of the CNTs. Considering the very large
axial ratio and the space distribution of the CNTs, Xue built a new
model for the effective thermal conductivity of CNTs-based
composites. Based on Maxwell theory, two formulae of calculating
the effective thermal conductivity of CNTs-based composites were
generated. The model shows that the axial ratio and the space
distribution of the CNTs can signicantly affect the effective
thermal conductivity of CNTs-based composites so that the
dispersion of a very small amount of CNTs can result in a remark-
able enhancement in the effective thermal conductivity of the
composites.
Xue assumed that k
e
is the effective thermal conductivity of the
CNTs-based composites, k
c
is the thermal conductivity of the CNTs
and k
m
is the thermal conductivity of the main media. By consid-
ering two kinds of distribution functions, the following relations for
the effective thermal conductivities were developed:
k
e
k
m
1 f 4f =p
k
c
=k
m
p
arctg
_
p=4
k
c
=k
m
p
_
1 f 4f =p
k
m
=k
c
p
arctg
_
p=4
k
c
=k
m
p
_ (7)
where f is the total volume fraction of the CNT and
k
e
k
m
1 f 2f
k
c
k
c
k
m
ln
k
c
k
m
2k
m
1 f 2f
k
m
k
c
k
m
ln
k
c
k
m
2k
m
(8)
Xue concluded that the distribution state and large ratio of the CNTs
play an important role in the thermal conductivity of nanouids.
However, the large discrepancy between the calculated values
using eqs. (7) and (8) and the experimental measurements of the
thermal conductivity of CNTs-oil composite was noticeable.
Recently, Xue [47] (hereafter we name Xue-2006) proposed
a novel approach for predicting the effective thermal conductivity
for carbon nanotube composites by incorporating the thermal
resistance and the average polarization theory in his previous
works [25,31]. In this new method, the combined effect of carbon
nanotube length, diameter, concentration, interface and matrix on
the thermal conductivity of the carbon nanotube composite have
been considered simultaneously:
91 f
k
e
k
m
2k
e
k
m
f
"
k
e
k
c
33
k
e
0:14
d
L
_
k
c
33
k
e
_
4
k
e
k
c
11
2k
e
1
2
_
k
c
11
k
e
_
#
0 9
where R
k
is thermal resistance, L and d are length and diameter of
carbon nanotube, f is the carbon nanotube volume fraction, k
c
and
k
m
are the thermal conductivities of carbon nanotube and the
matrix, respectively. Also, the transverse and longitudinal equiva-
lent thermal conductivities k
c
11
and k
c
33
are obtained from the
following equations:
k
c
11
k
c
1
2R
k
kc
d
; k
c
33
k
c
1
2R
k
kc
L
(10)
It has been shown that the effective thermal conductivity increases
rapidly with increasing nanotube length, but the effective thermal
conductivity changes very little when the diameter changes. Also, it
is shown that the effective thermal conductivity increases rapidly
with increasing thermal conductivity of CNTs, whereas the effective
thermal conductivity decreases rapidly with increasing thermal
conductivity of the matrix. It was also found that the large interface
thermal resistance between the CNTs and the base uid causes
a signicant degradation in the thermal conductivity. This thermal
resistance can be attributed to the weak atomic bonding at the
particle-matrix interface [49]. It should be mentioned that this
model is valid in the whole range of carbon nanotube
concentration.
M.M. Papari et al. / International Journal of Thermal Sciences 50 (2011) 44e52 46
2.2. A renovated HamiltoneCrosser model
Yu and Choi [26] developed a method to modify the Maxwell
equation [50] for spherical particle-in-liquid suspensions with each
spherical particle surrounded by an ordered liquid layer. They
reported that the ordered liquid layer plays an important role in
enhancing the thermal properties of nanouids. Choi et al. [16] cited
that multi-walled carbon nanotubes enhanced the thermal
conductivity of the base uid by as much as 160% at a concentration
of 1 vol%; the highest enhancement reported for a nanouid.
Because nanotubes are non-spherical, the renovated Maxwell
model [50] cannot be used to predict the effective thermal
conductivity of nanotube-based nanouids. In contrast, Hamilton
and Crosser [34] showed that when the particle-to-liquid conduc-
tivity ratio of a suspension is above 100, the particle shape has
a substantial effect on the effective thermal conductivity of the
suspension. For nanotube-in-liquid suspensions, the ratio of the
thermal conductivities of the two phases would be of the order of
10,000. Although the renovated Maxwell model is limited to
suspensions with spherical particles, an important feature of the
HamiltoneCrosser model [34] is that it can predict the effective
thermal conductivity of suspensions owing non-spherical particles.
However, the HamiltoneCrosser model does not include the effect
of the nanolayer. Yu and Choi, in 2004, extended the Hamil-
toneCrosser model [26] toellipsoidal particle-in-liquid suspensions
with each ellipsoidal particle surrounded by an interfacial layer.
To calculate the effective thermal conductivity of the complex
ellipsoid-in-liquid suspension containing monosized anisotropic
ellipsoids of the equivalent thermal conductivity k
pj
(j a, b and c)
and the equivalent volume concentration f
e
, it is assumed that the
thermal conductivity of the base liquid, k
1
is isotropic. Based on the
average theory and this assumption, by rewriting and averaging the
HamiltoneCrosser equation [26] along the semi-axis directions of
the complex ellipsoids, one obtains the effective thermal conduc-
tivity k
e
of the suspension of monosized ellipsoidal particles of
semiaxes a, b, and c. That is
k
e
_
1
nf
e
A
1 f
e
A
_
k
1
(11)
n 3J
a
(12)
J
2et
_
1 e
2
t
1=6
et
1 e
2
t
p
arcsinet
(13)
et
1
b
2
t
a
2
t
s
1
c
2
t
a
2
t
s
(14)
A
1
3
X
j a;b;c
k
pj
k
1
k
pj
n 1k
1
(15)
f
e
_
a
2
t
__
b
2
t
_
_
c
2
t
_
r
abc
f (16)
k
pj
_
1
k
p
k
s
k
p
rdj;0dj; t k
s
rdj; 0dj;tr
_
k
s
(17)
r
_
a
2
t
__
b
2
t
_
_
c
2
t
_
r
abc
(18)
d
j;n
_
a
2
n
__
b
2
n
_
_
c
2
n
_
r
2
Z
N
0
dw
_
j
2
n w
_
_
a
2
n w
__
b
2
n w
_
_
c
2
n w
_
r
(19)
where f is the volume concentration of the solid ellipsoids without
surrounding layers, f
e
is the volume concentration of the complex
ellipsoids, a is an empirical parameter and J is the sphericity
dened as the ratio of the surface area of a sphere to a volume equal
to that of the particle, n is the empirical shape factor, k
p
and k
s
are
the thermal conductivities of the solid ellipsoid and its surrounding
layer. However, this model fails to predict the nonlinear behavior of
the nanouid thermal conductivity.
3. Conventional articial neural networks
Although articial neural networks (ANNs) is widely utilized to
extract management knowledge from acquired data, but neural
information processing models mainly assume that the data are
compatible and the learning data for training a neural network are
sufcient. For cases in which the data are insufcient, it is impos-
sible to recognize a nonlinear system. In other words, there exist
a non-negligible error between the real function and the estimated
function from a trained neural network [51]. Moreover using basic
neural networks with small data points (actual points), it is difcult
to guarantee that a good predictive model will be obtained in the
complete actual domain. In most cases, only good predictions are
achieved in regions in the vicinity of actual points contained in the
training dataset [52]. However, a sufcient training data set is
dened as a data set which provides enough information to
a learning method to obtain stable training accuracy.
The ANNs supplies a non-linear function mapping of a set of
input variables into the corresponding network output variables,
without the requirement of having to specify the actual mathe-
matics form of the relation between the input and output variables.
The multilayer percepetron (MLP) is a type of feed forward neural
network that has been used commonly for the approximate func-
tions [53]. The ANN technique has been applied successfully in
various elds of modeling and prediction in many engineering
systems, mathematics, medicine, economics, metrology and many
others. It has become increasingly popular in during last decade.
The advantage of ANN compared to conceptual models are its high
speed, simplicity and large capacity which reduce engineering
attempt. The ANNs are discussed in details in the literature, and
therefore only a few prominent features of it are given here to
describe the general nature of the network. MLP neural networks
consist of multiple layers of simple activation units called neurons
that are arranged in such a way that each neuron in one layer is
connected with each neuron in the next by weighted connections
(see Fig. 1). Neurons are arranged in layers that make up the global
architecture. MLP networks are comprised of one input layer, at
least one hidden layer and an output layer. The number of neurons
in the input layer is dened by the problem to be solved. The input
layer receives the data. The output layer delivers the response
corresponding to the property values. The hidden layer processes
and organizes the information received from the input layer and
delivers it to the output layer. The number of neurons in the hidden
layer, which to some extent play the role of intermediate variables,
may be considered as an which might be either output from other
M.M. Papari et al. / International Journal of Thermal Sciences 50 (2011) 44e52 47
neurons or input from external sources. Each neuron thus has
a series of weighted inputs, w
ij
which might be either output from
other neurons or input from external sources. Each neuron calcu-
lates a sum of the weighted inputs and transforms it by the
following transfer function:
n
j
1
1 exp x
(20)
where n
j
is the output of the j-th neuron and x is given by:
x
X
n
i 1
w
ij
p
i
b
j
(21)
where w
ij
represents the weights applied to the connections from
the i-th neurons in the previous layer to the j-th neurons, p
i
is the
output from the i-th neuron and b is a bias term. MLP networks
operate in the supervised learning mode. In the training algorithm,
commonly back propagation error called BP, training data are given
to the networks and the network iteratively adjusts connection
weights w and biases b (starting from initial random values) until
the network predicted values match the actual values satisfactorily.
In the BP training algorithm, this adjustment is carried out by
comparing the actual value t
ij
and the predicted values a
ij
of the
network by means of calculation of the total sumof the square error
(SSE) for the n data of the training dataset,
SSE
X
n
i 1
_
t
ij
a
ij
_
2
(22)
3.1. Detailed steps in the ANN modeling procedure
In this work, the actual data set was provided by the Refs. [14,21]
and [54e57]. The number of actual data points is 43 in total and
presented in the sequence of Figs. 4e9.
According to available data, there are two input variables cor-
responding to thermal conductivity of uid (k
f
), volume fraction of
CNTs (f) and thermal conductivity ratio (k
e
/k
f
) of carbon nanotube
(k
e
/k
f
) as output variable. As explained previously, the available
experimental data points (43 data points) do not provide sufcient
information for training robust prediction neural network. There-
fore, the research of Li et al. [58] was referred to, and the mega-
trend-diffusion and estimation of domain range techniques were
used to create virtual data (adding a number of virtual data points)
to construct the predicted model. Because the purpose of this
research is to apply the procedure proposed by Li et al. to construct
a prediction model, the details of the derivation of the methodology
are omitted here. In order to ll all the data gaps and estimate the
data trend, the mega-trend-diffusion method assumes that the data
selected are located within a certain range and have possibility
indexes determined by a common membership function (MF). The
numbers of data which are smaller or greater than the average of
the data are considered the skewness of the MF, which is used to
estimate the trend of the population. When the data collected are
insufcient to build a reliable predicting model, the procedure
systematically generates some virtual data between the estimated
lower and upper extremes (L and U) of the range to ll the data
gaps. The computation of L and U mainly employs the statistical
diffusion techniques provided in the procedure. The procedure
further calculates the MF values for each virtual data as its impor-
tance level.
The Li et al. [57] showed that if the number of actual data
(experimental data) chosen for generation of virtual data increases
the average error rate of trained net decreases. So in this study rst
all available actual data were used for generating a virtual data then
the ANN was trained using generated data. Finally the actual data
were used for testing the ANN. There are a total of four steps in the
procedure which was implemented with 43 actual data:
Step 1 Select all actual data as the training data of ANN.
Step 2 Use eqs. (23) and (24) to calculate the estimated lower
bound (L) and upper bound (U),
L U
set
Sk
L
2 bs
2
x
=N
L
ln4L
q
(23)
U U
set
Sk
U
2 bs
2
x
=N
U
ln4U
q
: (24)
In these equations:
h
set
bs
2
x
=n (25)
U
set
min max=2; (26)
where bs
2
x
P
n
i1
x
i
x
2
=n 1 is the data set variance; n is the
dataset size; and U
set
, the core of the dataset so that
Sk
L
N
L
=N
L
N
U
and Sk
U
N
U
=N
L
N
U
(27)
The skewness (Sk) characterizes the degree of asymmetry of
a distribution around dataset core. Sk
U
indicates a distribution
fraction with an asymmetric tail extending greater than the core
value. Sk
L
indicates a distribution fraction with an asymmetric tail
extending less than core value. N
L
and N
U
represent the number of
data smaller and greater than U
set
, respectively.
Since a common function to diffuse the data set was utilized, the
values of the membership function for these two data range limits,
the left end L and the right end U, approach zero. Consequently, in
eqs. (23), and (24), a very small number, 10
20
was set for 4(L) and
4(U) to be the value of the membership function, respectively. The
lower (L) and upper (U) limits are calculated using equations (23)
and (24) for each variable (i.e. thermal conductivity of uid (k
f
),
volume fraction of CNTs (f) as inputs and thermal conductivity ratio
(k
e
/k
f
) of carbon nanotube (k
e
/k
f
) as output variable) of the ANN.
The results are listed in Table 1.
Step 3 generate a virtual data between L and U for each variable.
Step 4 for each variable, the corresponding value of the member-
ship function (MF) using eq. (28) is calculated. The thermal
Fig. 1. Topology of conventional P-S-1 MLP.
M.M. Papari et al. / International Journal of Thermal Sciences 50 (2011) 44e52 48
conductivity ratio is the output for the ANN, and there is no
need to calculate its MF value. The numbers randomly
selected for the variables and the corresponding value of MF
are summarized in Table 2.
MF
(
NvL
Uset L
N
v
< U
set
UNv
UUset
N
v
> U
set
(28)
where N
v
is a virtual number.
Because we repeat this procedure 90 times for each variable, this
gives rise to 90 virtual data for the training the MLP network. Table
2 contains the data generated for each variable.
The conventional topology of MLP can be shown as a P-S-1
network appeared in Fig. 1. In this topology, P represent the two
input variables corresponding to the thermal conductivity of the
uid (k
f
) and volume fraction of CNTs (f); S is the neuron in hidden
layer and 1 is the output variable corresponding to the thermal
conductivity ratio (k
e
/k
f
).
In sum, the hybrid model that was employed in this work
comprises both the mega-trend diffusion technique along with the
neural network. This hybrid model comprises an input layer of four
neurons corresponding to the thermal conductivity of the uid (k
f
),
the volume fraction of CNTs (f) and their corresponding values of MF,
plus a variable number of neurons inthe hiddenlayer anda neuronin
the output layer corresponding to the thermal conductivity ratio
(k
e
/k
f
). This topology is illustrated in Fig. 2. The output layer neuron
has a linear transfer function, while the hidden layer neurons have
a sigmoid transfer function. The neural network was trained
according to scaled conjugate gradient back propagation (train scg)
algorithm available in the neural network toolbox of MATLAB.
The 90 virtual data obtained in Step 4 were assigned to the
training of MLP, while the actual data were utilized to test the
model.
In the rst step of the training procedure, all data points were
scaled to the range of [1, 1]. For the training step, the number of
neurons in the hidden layer plays an important role in the network
optimization. Therefore, in order to optimize the network, to ach-
ieve generalization of the model and avoid over-tting, we started
with 1 neuron in the hidden layer and gradually increased the
number neurons until no signicant improvement in he perfor-
mance of the net was observed. For this study, the mean square
error (MSE) was chosen as a measure of the performance of the
networks. The network model with eight neurons in the hidden
layer, maximum epoch 300, momentum constant 0.8 and learning
rate 0.01. The MSE for this conguration was 0.14%. After deter-
mining the optimal topology of the network, the actual data and
the corresponding MF values are used as testing data for the trained
MLP to calculate the absolute average error. The average absolute
deviation (AAD) is dened as:
AAD
P
N
i 1
j
_
X
exp:
i
X
calc:
i
_.
X
exp
i
j 100
N
(29)
where N is the number of data.
4. Results and discussion
As already mentioned, in this study articial neural network
(ANN) scheme has been employed to compute the effective thermal
conductivity of nanouids. The selected nanouids were the multi-
walled CNTs suspended in oil (a-oln), decene (DE), distilled water
(DW), ethylene glycol (EG) and also single-walled CNTs suspended in
epoxy and poly methylmethacrylate (PMMA). Several improved
models cited in the previous sections [25,26,31,47] capable of calcu-
lating the therrmal conductivity of CNTs-based composites were
selected to be compared against our results. In these models, the size
of the particles and interface between CNTs and uid (interface
effect), and the concentration of CNTs are considered. It should be
remembered that most of the old theoretical models [40,41] depend
only on the thermal conductivity of particles, the thermal conduc-
tivity of the base liquid and the particle volume fraction.
In the present calculations, the thermal conductivity of CNTs is
taken as 600e3000 W/mK, and the thermal conductivity of oil,
decene, distilled water, ethylene glycol, epoxy and poly methyl-
methacrylate are taken as 0.1448, 0.14, 0.60, 0.25, 0.198 and 0.21 W/
mK, respectively.
As already stated, Xue-2003 [25] developed a new model for
predicting the effective thermal conductivity of nanouids. This
model is based on the average polarization theory and takes into
account the effect of the interface between the solid particles and
the base uid. Xue-2003 claims that the model predictions are in
good agreement with the actual data for nanotube-in-oil nano-
uids and exhibits nonlinearity with the nanotube loadings.
We selected the following parameters to predict the thermal
conductivity of the aforementioned nanouids: a 50,000 nm,
b c 12.5 nm, k
l
0.1448 W/mK, k
p
2000 W/mK, interfacial
layer thickness 3 nm, k
s
5 W/mK, and depolarization factor
along a axis 0.0062. We performed the calculations of eqs.
(2)e(6) using the above parameters for MWCNTS in oil suspension
Table 2
The random numbers selected between L and U for each factor.
k
f
f k
e
/k
f
N
v
0.1455 0.1993 0.2513 0.2713 0.1295 0.3333 1.2867 1.0767 1.0337
MF 0.0238 0.2580 0.4841 0.1088 0.0519 0.1336 0.3675 0.983 0.0432
Fig. 2. The topology of 2P-S-1 MLP for the small data set training.
Table 1
The lower and upper limits of virtual data for each variable.
Limit k
f
f k
e
/k
f
L 0.14 0.0 1
U 0.6 4.99 2.56
M.M. Papari et al. / International Journal of Thermal Sciences 50 (2011) 44e52 49
and found that two of the parameters are incorrect for applications
to nanouids. We corrected the semiaxis a to 25,000 nm, and the
depolarization factor along the axis to 0.0000018. By using the two
corrected parameters, the predicted values of the thermal
conductivity of CNT/oil mixture were much higher than those
reported in Ref. [25]. Fig. 3 illustrates the predicted thermal
conductivity of CNT/oil nanouid using Xue-2003 original param-
eters and our corrected parameters as well as the experimental
ones [16]. As Fig. 3 conrms, the calculated thermal conductivity
from the corrected parameters disagrees with the experimental
data. It should be mentioned that Yu and Choi [26] have already
pointed out that the aforementioned parameters were incorrect.
Then, we decided to check further the validity of Xue-2003 model.
We extended the model to other nanouids including CNTs/DE and
CNTs/DW. The required parameters for these latter nanouids were
tted to the experimental data reported in refs. [17,56]. Surpris-
ingly, the calculated thermal conductivities were again found to be
much higher than the experimental ones [17,56]. Shown for
comparison, Fig. 3 illustrates the calculated values and the exper-
imental observations of the thermal conductivity of CNT/DE and
CNT/DWcomposites. This comparison reinforces the discovery that
the Xue-2003 model is inadequate to predict the thermal conduc-
tivity of nanouids containing CNTs and a conventional uid.
Therefore, the validity and accuracy of Xue-2003 must be investi-
gated further. Because of this limitation, we removed the Xue-2003
model [25] from our next comparisons. We continued our studies
by applying other three remaining models [26,31,47] to the uids
under consideration. To apply the Yu and Choi model [26] to all
carbon nanotubes-based composites, the following parameters
were tted with the experimental values:
a 25; 000 nm; b c 12:5 nm; k
p
2000 W=mK;
t 54 nm:
where a for oil and epoxy were taken as 1.55 and for the rest of
nanouids a 1. In addition, we chose k
s
100k
1
for CNTs/oil,
k
s
10k
1
for CNT/DW and CNT/epoxy also k
s
k
1
for CNT/DE, CNT/
EG and CNT/PMMA. The limitation inherent to this model, as
indicated by Yu and Choi [26], is that it fails to represents the
nonlinearity of the thermal conductivity with nanotube
concentration.
Furthermore, in order to calculate the thermal conductivity of
the aforementioned nanouids by Xue-2005 model [31], the
following parameters were selected: k
c
600 W/mK for CNT/oil,
CNT/EGand CNT/DW, k
c
3000 W/mK for CNT/DE and CNT/PMMA,
k
c
1000 W/mK for CNT/epoxy. Besides, we found that the best
results for the calculated thermal conductivity of CNT/oil, CNT/DW
and CNT/epoxy could be obtained by eq. (7) and for other remaining
nanouids by eq. (8).
Finally, the parameters for the calculation of the thermal
conductivities of the present nanouids required by Xue-2006
model [47] have these magnitudes: d 25 nm, L 50,000 nm for
CNT/oil and other nanouids have d 15 nm, L 30,000 nm. In this
model k
c
3000 W/mK, the thermal resistance (R
k
) is chosen as:
Fig. 3. Comparison of Xue-2003 model [25] predictions with actual data.
1
1.5
2
2.5
3
0 0.2 0.4 0.6 0.8 1 1.2
CNT volume fraction %
T
h
e
r
m
a
l
c
o
n
d
u
c
t
i
v
i
t
y
r
a
t
i
o
Choi et al (2001)
Xue (2005) model
Xue (2006) model
Yu and Choi model
+ Neural network
Fig. 4. Comparison of the thermal conductivity ratio of MWCNTs-oil suspensions
obtained from the Diffusion-Trend-Neural network with experiment and some models.
1.0
1.1
1.2
1.3
1.4
0 0.2 0.4 0.6 0.8 1 1.2
CNT volume fraction %
T
h
e
r
m
a
l
c
o
n
d
u
c
t
i
v
i
t
y
r
a
t
i
o
Xie et al (2003)
Xue (2005) model
Xue (2006) model
Yu and Choi model
+ Neural network
Fig. 5. Comparison of the thermal conductivity ratio of the MWCNTs-DE suspensions
obtained from the Diffusion-Trend-Neural network with experiment and with some
models.
1
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
0 0.2 0.4 0.6 0.8 1 1.2
CNT volume Fraction %
T
h
e
r
m
a
l
c
o
n
d
u
c
t
i
v
i
t
y
r
a
t
i
o
Xie et al (2003)
Wen D S (2004)
Xue (2005) model
Xue (2006) model
Yu and Choi model
+ Neural network
Fig. 6. Comparison of the thermal conductivity ratio of MWCNTs-DW suspensions
obtained from the Diffusion-Trend-Neural network with experiment and with some
models.
M.M. Papari et al. / International Journal of Thermal Sciences 50 (2011) 44e52 50
30 10
8
, 15.9 10
7
, 13.9 10
7
, 20 10
-8
, 25 10
-8
and 20 10
-7
Km
2
W
-1
for CNT/oil, CNT/DE, CNT/EG, CNT/DW, CNT/epoxy and
CNT/PMMA, respectively.
We carried out articial neural network calculations, explained
in the previous section, for all uids and compared the results
obtained against those calculated by the above-mentioned models.
For verication of the validity of our calculations and the models,
the predicted values of the thermal conductivities have been
compared with the experimental measurements [16,17,56].
Figs. 4e7 compare the thermal conductivity ratio (k
e
/k
f
) of
a multi-walled carbon nanotube (MWCNT) in oil, decene, distilled
water, and ethylene glycol, respectively, as a function of the volume
fraction of nanotubes using the articial neural network with those
values obtained by models and experiments. As it may be observed
in Figs. 4e7, the calculated thermal conductivities using the ANN
method agree well with both the experimental values [16,17,56]
and other models [26,31,47].
Further, Figs. 8 and 9 illustrate the comparison between the
calculated thermal conductivity of single-walled carbon nanotube
(SWCNT) in epoxy and PMMA uids using the ANN method and
models. Shown for comparison, the experimental values of thermal
conductivities of aforementioned composites [14,52,57e59] have
also been demonstrated.
In order that the validity of the results obtained are checked
further, the AAD of thermal conductivity ratios from the literature
values in the entire range of volume fraction for whole of nano-
uids has been tabulated in Table 3. As the numbers in Table 3
attest, the thermal conductivity ratio computed from the Mega-
Trend-Diffusion neural network method stands over other models.
From Table 3, it is immediately veriable that the DNN-MLP model
possesses a high ability to predict the thermal conductivity with an
absolute average error of 3.26%. There is a good agreement between
the predicted and the experimental values of the thermal
conductivity ratio (r 0.991). Herein, the standard deviation in the
relative errors was 2.3%, meaning that the dispersion around the
average was very small.
It should be mentioned that examining articial neural network
approach to predict other transport properties, say viscosity, of
nanouids remains for future work.
Acknowledgements
The authors thank computer facility provided by Shiraz
University of Technology and Shiraz University.
References
[1] C.P. Poole, F.J. Owens, Introduction to Nanothechnology. Wiley-Interscience,
Hoboken, New Jersey, 2003.
[2] S. Iijima, Helical microtubules of graphitic carbon, Nature 354 (1991) 56e58.
[3] R.E. Tuzun, D.W. Noid, B.G. Sumpter, R.C. Merkle, Dynamics of uid ow inside
carbon nanotubes, Nanotechnology 7 (1996) 241e246.
[4] R.E. Tuzun, D.W. Noid, B.G. Sumpter, R.C. Merkle, Dynamics of He/C60 ow
inside carbon nanotubes, Nanotechnology 8 (1997) 112e118.
[5] D. Qian, W.K. Liu, R.S. Ruoff, Mechanics of C60 in nanotubes, Journal of
Physical Chemistry B 105 (2001) 10753e10758.
[6] V.P. Sokhan, D. Nicholson, N. Quirke, Fluid ow in nanopores: accurate
boundary conditions for carbon nanotubes, Journal of Chemical Physics 117
(2002) 8531e8539.
[7] S. Supple, N. Quirke, Rapid imbibition of uids in carbon nanotubes, Physical
Review Letters 90 (2003) 214501e2145014.
[8] J. Li, L. Porter, S. Yip, Atomistic modeling of nite-temperature properties of
crystalline b-SiC II. Thermal conductivity and effects of point defects, Journal
of Nuclear Materials 255 (1998) 139e152.
[9] Q. Wang, J.K. Johnson, Molecular simulation of hydrogen adsorption in single-
walled carbon nanotubes and idealized carbon slit pores, Journal of Chemical
Physics 110 (1999) 577e586.
1
1.05
1.1
1.15
1.2
0 0.2 0.4 0.6 0.8 1 1.2
CNT volume Fraction %
T
h
e
r
m
a
l
c
o
n
d
u
c
t
i
v
i
t
y
r
a
t
i
o
Liu et al (2005)
Xie et al (2003)
Assael et al (2006)
Xue (2005) model
Xue (2006) model
Yu and Choi model
+ Neural network
Fig. 7. Comparison of the thermal conductivity ratio of the MWCNTs-EG suspensions
obtained from the Diffusion-Trend-Neural network with experiment and with some
models.
1
1.5
2
2.5
3
3.5
4
4.5
5
0 0.5 1 1.5 2 2.5
CNT volume Fraction %
T
h
e
r
m
a
l
c
o
n
d
u
c
t
i
v
i
t
y
r
a
t
i
o
Biercuk et al (2002)
Assael et al (2003)
Bryning et al (2005)
Xue (2005) model
Xue (2006) model
Yu and Choi model
+ Neural network
Fig. 8. Comparison of the thermal conductivity ratio of the SWCNTs-epoxy suspen-
sions obtained from the Diffusion-Trend-Neural network with experiment and with
some models.
1
1.2
1.4
1.6
1.8
2
2.2
2.4
2.6
0 2 4 6 8
CNT volume Fraction %
T
h
e
r
m
a
l
c
o
n
d
u
c
t
i
v
i
t
y
r
a
t
i
o
Du. F et al (2006)
Xue (2005) model
Xue (2006) model
Yu and Choi model
+ Neural network
Fig. 9. Comparison between actual data of SWCNTs-PMMA suspensions and the
calculated values from mentioned models and Diffusion-Trend-Neural network.
Table 3
The AAD% of the calculated effective thermal conductivity ratio using Xue-2005,
Xue-2006, Yu and Choi models, as well as ANN method from the experimental data.
System Xue-2005 Xue-2006 Yu and Choi 2004 Neural network
MWCNT/Oil 4.75 7.89 4.52 2.79
MWCNT/DE 0.38 0.62 0.81 2.50
MWCNT/DW 7.22 4.89 5.22 3.64
MWCNT/EG 1.24 1.15 1.79 1.86
SWCNT/Epoxy 15.86 5.47 3.32 2.48
SWCNT/PMMA 15.74 16.90 15.29 6.31
Overall 7.53 6.15 5.16 3.26
M.M. Papari et al. / International Journal of Thermal Sciences 50 (2011) 44e52 51
[10] S. Berber, Y.-K. Kwon, D. Tomanek, Unusually high thermal conductivity of
carbon nanotubes, Physical Review Letters 84 (2000) 4613e4616.
[11] J. Che, T. Cagin, W.A. Goddard, Thermal conductivity of carbon nanotubes,
Nanotechnology 11 (2000) 65e69.
[12] M.A. Osman, D. Srivastava, Temperature dependence of the thermal conduc-
tivity of single-wall carbon nanotubes, Nanothecnology 12 (2001) 21e24.
[13] H. Rai-Tabar, Computational modeling of thermo-mechanical and transport
properties of carbon nanotubes, Physics Reports 390 (2004) 235e452.
[14] M.J. Assael, C.-F. Chen, I. Metaxa, W.A. Wakeham, Thermal conductivity of
suspensions of carbon nanotubes in water, International Journal of Thermo-
physics 25 (2004) 971e985.
[15] M.J. Assael, I.N. Metaxa, J. Arvanitidis, D. Christolos, C. Lioutas, Thermal
conductivity enhancement in aqueous suspensions of carbon multi-walled
and double-walled nanotubes in the presence of two different dispersants,
International Journal of Thermophysics 26 (2005) 647e664.
[16] S.U.S. Choi, Z.G. Zhang, W. Yu, F.E. Lockwood, E.A. Grulke, Anomalous thermal
conductivity enhancement in nanotube suspensions, Applied Physics Letters
79 (2001) 2252e2254.
[17] H. Xie, H. Lee, W. Youn, M. Choi, Nanouids containing multiwalled carbon
nanotubes and their enhanced thermal conductivities, Journal of Applied
Physics 94 (2003) 4967e4971.
[18] Y. Xuan, Q. Li, W. Hu, Aggregation structure and thermal conductivity of
nanouids, AIChE Journal 49 (2003) 1038e1043.
[19] S.P. Jang, S.U.S. Choi, Role of Brownian motion in the enhanced thermal
conductivity of nanouids, Applied Physics Letters 84 (2004) 4316e4318.
[20] P. Bhattacharya, S.K. Saha, A. Yadav, P.E. Phelan, R.S. Prasher, Brownian
dynamics simulation to determine the effective thermal conductivity of
nanouids, Journal of Applied Physics 95 (2004) 6492e6494.
[21] R. Prasher, P. Bhattacharya, P.E. Phelan, Thermal conductivity of nanoscale
colloidal solutions (nanouids), Physical Review Letters 94 (2005) 025901.
[22] W. Evans, J. Fish, P. Keblinski, Role of Brownian motion hydrodynamics on
nanouid thermal conductivity, Applied Physics Letters 88 (2006) 93116.
[23] D.H. Kumar, H.E. Patel, V.R.R. Kumar, T. Sundararajan, T. Pradeep, S.K. Das,
Model for heat conduction in nanouids, Physical Review Letters 93 (2004)
144301-1e144301-4.
[24] W. Yu, S.U.S. Choi, The role of interfacial layers in the enhanced thermal
conductivity of nanouids: a renovated Maxwell model, Journal of Nano-
particle Research 5 (1e2) (2003) 167e171.
[25] Q.-Z. Xue, Model for effective thermal conductivity of nanouids, Physics
Letters. A 307 (2003) 313e317.
[26] W. Yu, S.U.S. Choi, The role of interfacial layers in the enhanced thermal
conductivity of nanouids: a renovated HamiltoneCrosser model, Journal of
Nanoparticle Research 6 (4) (2004) 355e361.
[27] C.W. Nan, G. Liu, Y. Lin, M. Li, Interface effect on thermal conductivity of
carbon nanotube composites, Applied Physics Letters 85 (2004) 3549e3551.
[28] L. Xue, P. Keblinski, S.R. Phillpot, S.U.S. Choi, J.A. Eastman, Effect of liquid
layering at the liquidesolid interface on thermal transport, International
Journal of Heat and Mass Transfer 47 (2004) 4277e4284.
[29] K.C. Leong, C. Yang, S.M.S. Murshed, A model for the thermal conductivity of
nanouids-the effect of interfacial layer, Journal of Nanoparticle Research 8
(2006) 245e254.
[30] C. Sitprasert, P. Dechaumphai, V. Juntasaro, A thermal conductivity model for
nanouids including effect of the temperature-dependent interfacial layer,
Journal of Nanoparticle Research 11 (2009) 1465e1476.
[31] Q.-Z. Xue, Model for thermal conductivity of carbon nanotube-based
composites, Physica B 368 (2005) 302e307.
[32] P. Keblinski, S.R. Phillpot, S.U.S. Choi, J.A. Eastman, Mechanisms of heat ow in
suspensions of nano-sized particles (nanouids), International Journal of Heat
and Mass Transfer 45 (4) (2001) 855e863.
[33] N.R. Karthikeyan, J. Philip, B. Raj, Effect of clustering on the thermal conduc-
tivity of nanouids, Material Chemistry and Physics 109 (2008) 50e55.
[34] R.L. Hamilton, O.K. Crosser, Thermal conductivity of heterogeneous two-
component systems, Industrial and Engineering Chemistry Fundamentals 1
(1962) 187e191.
[35] C.H. Liu, S.S. Fan, Effects of chemical modications on the thermal conduc-
tivity of carbon nanotube, Applied Physics Letters 86 (12) (2005) 1e3.
[36] M.J. Assael, I.N. Metaxa, K. Kakosimos, D. Constantinou, Thermal conductivity
of nanouids e experimental and theoretical, International Journalof Ther-
mophysics 27 (2006) 999e1017.
[37] S. Shenogin, L.P. Xue, R. Ozisik, P. Keblinski, D.G. Cahill, Role of thermal
boundary resistance on the heat ow in carbon-nanotube composites, Journal
of Applied Physics 95 (2004) 8136e8144.
[38] J. Phillip, P.D. Shima, B. Raj, Enhancement of thermal conductivity in mag-
netitie based nanouid due to chainlike structures, Applied Physics Letters 91
(2007) 2031081e2031083.
[39] J. Phillip, P.D. Shima, B. Raj, Nanouid with tunable thermal properties,
Applied Physics Letters 92 (2008) 0431081e0431083.
[40] D.J. Jeffrey, Conduction through a random suspension of spheres, Proceedings
of Royal Society of London, Series A 335 (1973) 355e367.
[41] R.H. Davis, The effective thermal conductivity of a composite material with
spherical inclusions, International Journal of Thermophysics 7 (1986)
609e620.
[42] B.X. Wang, L.P. Zhou, X.F. Peng, A fractal model for predicting the effective
thermal conductivity of liquid with suspension of nanoparticles, International
Journal of Heat and Mass Transfer 46 (2003) 2665e2672.
[43] C.W. Nan, Z. Shi, Y. Lin, A simple model for thermal conductivity of carbon
nanotube-based composites, Chemical Physics Letters 375 (2003) 666e669.
[44] Y. Xuan, Z. Yao Lattice, Boltzmann model for nanouids, Heat and Mass
Transfer 41 (2005) 199e205.
[45] Q. Xue, W.M. Xu, A model of thermal conductivity of nanouids with inter-
facial shells, Material Chemistry and Physics 90 (2005) 298e301.
[46] J. Koo, C. Kleinstreuer, A new thermal conductivity model for nanouids,
Journal of Nanoparticle Research 6 (2004) 577e588.
[47] Q.Z. Xue, Model for the effective thermal conductivity of carbon nanotube
composites, Nanotechnology 17 (2006) 1655e1660.
[48] T. Ohara, D. Suzuki, Intermolecular energy transfer at a solid-liquid interface,
Microscale Thermophysics Engineering 4 (2000) 189e196.
[49] L. Xue, P. Keblinski, S.R. Philpot, S.U.S. Choi, J.A. Eastman, Two regimes of
thermal resistance at a liquidesolid interface, Journal of Chemical Physics 118
(2003) 337e339.
[50] J.C. Maxwell, A Treatise on Electricity and Magnetism, second ed. Oxford
University Press, Cambridge, UK, 1904, pp. 435e441.
[51] C.F. Huang, C. Moraga, A diffusion-neural-network for learning from small
samples, International Journal of Approximate Reasoning 35 (2004) 137e161.
[52] R. Lanouette, J. Thibault, J.L. Valade, Process modeling with neural networks
using small experimental datasets, Computer and Chemical Engineering 23
(1999) 1167e1176.
[53] P.P. Van der Smagt, Minimization methods for training feedforward neural
network, Neural Networks 7 (1) (1994) 1994.
[54] M.B. Bryning, M.F. Islam, J.M. Kikkawa, A.G. Yodh, Very low conductivity
threshold in bulk isotropic single-walled carbon nanotube-epoxy composites,
Advanced Materials 17 (9) (2005) 1186e1191.
[55] M.S. Liu, M.C.C. Lin, I.T. Huang, C.C. Wang, Enhancement of thermal conduc-
tivity with carbon nanotube for nanouids, International Communication in
Heat and Mass Transfer 32 (2005) 1202e1210.
[56] D.S. Wen, Y.L. Ding, Effective thermal conductivity of aqueous suspensions of
carbon nanotubes (carbon nanotube nanouids), Journal of Thermophysics
and Heat Transfer 18 (2004) 481e485.
[57] F. Du, C. Guthy, T. Kashiwagi, J.E. Fischer, K.I. Winey, An inltration method for
preparing single-wall nanotube/epoxy composites with improved thermal
conductivity, Journal of Polymer Science: Part B: Polymer Physics 44 (2006)
1513e1519.
[58] D.C. Li, C.S. Wu, T.I. Tsai, Y.S. Lina, Using mega-trend-diffusion and articial
samples in small data set learning for early exible manufacturing system
scheduling knowledge, Computers & Operations Research 34 (2007) 966e982.
[59] M.J. Biercuk, M.C. Radosavljevic, J.K. Hyun, A.T. Johnson, J.E. Fischer, Carbon
nanotube composites for thermal management, Applied Physics Letters 80
(2002) 2767e2769.
M.M. Papari et al. / International Journal of Thermal Sciences 50 (2011) 44e52 52