See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/256718532

Experimental investigation on the thermal

conductivity and shear viscosity of
viscoelastic-fluid-based nanofluids

Article in International Journal of Heat and Mass Transfer · May 2012

DOI: 10.1016/j.ijheatmasstransfer.2012.02.052

CITATIONS READS

38 76

5 authors, including:

Feng-Chen Li Wenwu Zhou

Harbin Institute of Technology Iowa State University
179 PUBLICATIONS 1,159 CITATIONS 10 PUBLICATIONS 74 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Complex fluid flow and heat transfer View project

Electrokinetic Phenomena in viscoelastic fluids View project

All content following this page was uploaded by Feng-Chen Li on 09 October 2015.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
This article appeared in a journal published by Elsevier. The attached
copy is furnished to the author for internal non-commercial research
and education use, including for instruction at the authors institution
and sharing with colleagues.
Other uses, including reproduction and distribution, or selling or
licensing copies, or posting to personal, institutional or third party
websites are prohibited.
In most cases authors are permitted to post their version of the
article (e.g. in Word or Tex form) to their personal website or
institutional repository. Authors requiring further information
regarding Elsevier’s archiving and manuscript policies are
encouraged to visit:
http://www.elsevier.com/copyright
 Author's personal copy

Thermochimica Acta 556 (2013) 47–53

Contents lists available at SciVerse ScienceDirect

Thermochimica Acta
journal homepage: www.elsevier.com/locate/tca

Experimental study on the characteristics of thermal conductivity and shear

viscosity of viscoelastic-fluid-based nanofluids containing multiwalled carbon
nanotubes
Feng-Chen Li ∗ , Juan-Cheng Yang, Wen-Wu Zhou, Yu-Rong He, Yi-Min Huang, Bao-Cheng Jiang
School of Energy Science and Engineering, Harbin Institute of Technology, Harbin 150001, China

a r t i c l e i n f o a b s t r a c t

Article history: In order to obtain a novel thermo-fluid with both turbulent drag reducing and heat transfer enhance-
Received 17 October 2012 ment (compared with drag-reduced flow) abilities, we have prepared a viscoelastic-fluid-based nanofluid
Received in revised form 14 January 2013 (VFBN) using viscoelastic aqueous solution of cetyltrimethyl ammonium chloride/sodium salicylate as
Accepted 21 January 2013
base fluid and multiwalled carbon nanotubes (MWCNTs) as nanoparticles. The thermal conductivity and
Available online 8 February 2013
shear viscosity of the prepared VFBN with various particle volume fractions, temperatures and con-
centrations of the base fluid were then experimentally investigated. The results show that thermal
Keywords:
conductivities of the tested VFBNs are significantly higher than that of the corresponding base fluid
Viscoelastic-fluid-based nanofluids
Thermal conductivity
and increase with increasing particle volume fraction and fluid temperature, demonstrating potentials in
Viscosity heat transfer enhancement. A modified Li–Qu–Feng model (Y.H. Li, W. Qu, J.C. Feng, Chinese Phys. Lett. 25
Carbon nanotubes (2008) 3319–3322), which includes the effect of liquid layering, particle clustering, particle shape factor,
Surfactant solution Brownian motion and viscosity of base fluid, is proposed in the present paper to predict thermal con-
ductivity of VFBNs containing MWCNTs. The results predicted by this modified Li–Qu–Feng model show
excellent agreements with the measured data. The VFBN with MWCNTs shows a non-Newtonian fluid
behavior in its shear viscosity, and its shear viscosity increases with the increase of particle volume frac-
tion and decrease of temperature. It is expectable that the prepared VFBNs may also have drag-reducing
ability in turbulent flows.
© 2013 Elsevier B.V. All rights reserved.

1. Introduction with anomalously high thermal conductivity as reported by Kim

et al. [6]) have greatly improved thermal conductivity as com-
Nanofluids have attracted numerous researchers in the field of pared with its base fluid and can be applied to heat transfer
heat transfer because of their unique characteristics of substantial system. Choi et al. [7] reported the largest magnitude of effec-
high thermal conductivity and convective heat transfer efficiency. tive thermal conductivity enhancement, about 160%, when 1.0 vol%
Since Choi [1] created the new term nanofluids, which represents (volume fraction) CNTs was added in synthetic poly oil. Xie et al.
the mixtures of nanomaterials and traditional heat transfer fluids [8] obtained 19.6%, 12.7% and 7.0% thermal conductivity augmen-
(liquids), lots of studies have been performed on the thermal per- tation for 1.0 vol% CNTs dispersed in decene, ethylene glycol and
formance of nanofluids, see, e.g., Özerinç et al. [2], Das and Choi [3] distilled water, respectively. These studies suggest that nanofluids
and Kleinstreuer and Feng [4]. It has been shown that adding small containing CNTs are excellent flow media and might be the next
amount of nanoparticles, like Al2 O3 , Cu and CuO into conventional generation of heat transfer fluids from the viewpoint of thermal
heat transfer media (water, ethylene glycol, oil and so forth) can performance. However, if considering from the viewpoint of hydro-
significantly increase the thermal conductivity of the base fluids. dynamics, the addition of nanoparticles in the conventional fluid
Carbon nanotubes (CNTs) are fibrous tubes, and were initially will aggravate the pressure drop in a pumping flow system due to
observed by Iijima [5] two decades ago. For their unusual structure the increased viscosity of nanofluids, e.g., as stated by Teng et al.
and remarkable mechanical and electrical properties, CNTs have [9]. This defect constrains the application area of nanofluids in heat
been paid great attentions by related researchers. Recent studies transfer system.
indicate that nanofluids containing multiwalled CNTs (MWCNTs On the other hand, there is another group of liquid flow
media, namely turbulent drag-reducing fluids, firstly reported by
Toms in 1949 [10]: when adding a tiny amount of polymers
∗ Corresponding author. Tel.: +86 451 86403254; fax: +86 451 86403254. or some kind of surfactants into the Newtonian fluids (liquids),
E-mail address: lifch@hit.edu.cn (F.-C. Li). the frictional drag of turbulent flow can be greatly reduced.

0040-6031/$ – see front matter © 2013 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.tca.2013.01.023
 Author's personal copy

48 F.-C. Li et al. / Thermochimica Acta 556 (2013) 47–53

The drag-reducing polymer or surfactant solution is usually vis- Table 1

Property of MWCNTs used in the present experiment.
coelastic fluid. For cationic surfactants, under proper conditions
of surfactant/counterion chemical structures, ratios, concentra- MWCNT1 MWCNT2 MWCNT3
tions and temperature, rod-like micelles can be formed in their 3
Density (g/cm ) ∼2.1 ∼2.1 ∼2.1
aqueous solution. Exerted by a proper shear stress, the rod-like Thermal conductivity (W/(m K)) ∼3000 ∼3000 ∼3000
micelles form a network structure, and this microstructure imparts Specific heat (J/(kg K)) 600 600 600
viscoelasticity to the solution, which is responsible for the occur- Length (␮m) 10–30 10–30 0.5–2
Outer diameter (nm) 8 10–20 8
rence of turbulent drag reduction (DR) [11]. Li et al. [12,13]
Inner diameter (nm) 2 5 2
investigated the turbulent DR mechanism and heat transfer char-
acteristics in a heated turbulent flow of drag-reducing aqueous
solution of cetyltrimethyl ammonium chloride/sodium salicylate by chemical vapor deposition method (Chengdu Organic Chemi-
(CTAC/NaSal). It was showed that heat transfer reduction (HTR) cals Co. Ltd., China). Their properties are shown in Table 1. As the
always occurs together with DR at the same order. Several attempts chemical affinity between MWCNTs and polar solutions is weak, the
to enhance heat transfer of the turbulent drag-reducing flow by MWCNTs cannot be directly dispersed into aqueous solutions. By
additives have been reported in the literature so far, e.g., by Li et al. acid treatment, the hydrophilic functional groups can be introduced
[14] and Qi et al. [15]. Nevertheless, the problems associated with to the MWCNTs surface [8], similar to attachment of carboxyl func-
HTR are yet solved satisfactorily. tional group to the tubes, which is the situation of the MWCNTs
It is straightforward to conjecture that the combination of used in the present experiment.
advantages of the abovementioned nanofluids and turbulent drag- A two-step (preparing base fluid at first, and then nanofluid)
reducing solutions might produce a novel thermohydrodynamic method is employed to prepare the viscoelastic-fluid-based MWC-
fluid. We have named this kind of novel fluid as viscoelastic-fluid- NTs nanofluids. Since CTAC/NaSal aqueous solution has shown an
based nanofluid (VFBN) in a recently published paper [16], which excellent turbulent drag reducing ability in our previous exper-
can circumvent the drawbacks and strengthen the advantages of imental studies [11–13], we still choose it as base fluid herein.
both nanofluids and viscoelastic fluid. In 2010, Liu and Liao [17] Proper amount of CTAC and NaSal powder in 1:1 weight ratio was
firstly reported their experimental studies on flow and heat trans- added into the distilled water. After mixing for about 8 h, a stable
fer characteristics of drag-reducing fluid with addition of carbon and well-dissolved aqueous solution of CTAC/NaSal was obtained.
nanotubes. The fluid they used was just a kind of VFBN. Their The prepared solution was put still for one night to release air bub-
experimental results testified the similar turbulent drag reducing bles in the fluid. The eventual base fluid consisted of CTAC/NaSal
effects between flows of viscoelastic fluids and VFBN. Both fluids aqueous solution and dispersant TNDIS (Chengdu Organic Chem-
can greatly reduce the frictional resistance in their turbulent flows. icals Co. Ltd., China). The addition of dispersant was to help the
In addition, VFBN can also significantly improve the heat transfer dispersion of nanoparticles in the base fluid and the adding amount
efficiency. was dependent on the MWCNTs volume fraction. To prepare VFBNs,
We have experimentally investigated the thermal conductivity proper amount of MWCNTs (corresponding to a planned volume
and shear viscosity of VFBN with aqueous solution of CTAC/NaSal as fraction) was added into the base fluid (at a certain concentration).
the base fluid and containing Cu nanoparticles [16], and the charac- After strongly mixing for 1 h, a stable VFBN with suspensions of
teristics of turbulent flow resistance and heat transfer of such VFBN nanoparticles was obtained. In the present study, the mass con-
in a round tube [18]. It was obtained that the thermal conductiv- centrations of CTAC/NaSal aqueous solution were 0, 50, 200 and
ity of this VFBN increases with increasing particle volume fraction 1000 ppm (ppm – part per million), respectively. And each base
and temperature (a maximum 29% increase of thermal conduc- fluid was mixed with 0.05, 0.2, 1.0, 1.6 and 3.5 vol% of MWCNTs,
tivity for the tested VFBNs was obtained), and this VFBN behaves respectively.
non-Newtonian characteristics in shear viscosity [16]; a synergetic
effect of both viscoelasticity and nanoparticles expected for VFBN 2.2. Measurement of thermal conductivity
flow shows up, i.e. either that DR and heat transfer enhancement
co-occur compared with water based nanofluid flow, or that DR Transient hot-wire method (THWM) was employed to measure
and HTR co-occur but with HTR rate being smaller than DR rate the thermal conductivities of the prepared VFBNs in the present
(for viscoelastic base fluid only, HTR rate is larger than DR rate) experiment. THWM, which can avoid natural convection success-
[18]. fully and has excellent precision and fast response, is one of the
The present study is one of the continuous efforts paid for convenient measurement techniques for thermal conductivities of
exploring novel working fluids used in turbulent flows in order gas and liquid.
to obtain significant turbulent DR meanwhile enhance heat trans- Fig. 1 illustrates the measured thermal conductivities for dis-
fer for drag-reduced turbulent flows. Stable VFBNs with MWCNTs tilled water at different temperatures by means of the THWM
as nanoparticles and aqueous solution of CTAC/NaSal as the base system (XIA-TC-THW-L01, Xi’an Xiatech Electronic Technology Co.
fluid were produced for experimental studies. The thermal con- Ltd., China). The measurement error was estimated to be within 1%.
ductivities and shear viscosities of different VFBNs (with different Hence, the precision and reliability of the applied THWM system
base fluid concentrations and different volume fractions of MWCNT were validated.
nanoparticles) were then investigated experimentally. The exper-
imental results of thermophysical properties for these prepared 2.3. Measurement of shear viscosity
VFBNs are detailed in this paper.
Shear viscosity is one of the most important physical prop-
erties of fluids for a thermohydrodynamic application system,
2. Experimental procedures which reflects the inherent resistance of a fluid to flow and deter-
mines the pumping power of heat transfer system. On the other
2.1. Preparation of VFBNs hand, based on the variation trend of shear viscosity versus shear
rate, the fluid nature, i.e. either Newtonian or non-Newtonian,
Distilled water, CTAC, NaSal and MWCNTs were utilized in the is defined. Compared with the large amount of investigations on
present experiments to prepare VFBNs. The CNTs were produced thermal performance, only a few available studies have covered
 Author's personal copy

F.-C. Li et al. / Thermochimica Acta 556 (2013) 47–53 49

Fig. 1. Verification of the applied THWM system for the measurement of thermal
conductivity.

viscosity of nanofluids around room temperatures, e.g., Mahbubul

et al. [19], Chen et al. [20] and Phuoc et al. [21]. Chen et al. [20]
conducted experimental research on volume fraction and tempera-
ture effects on viscosity of distilled water based MWCNT nanofluids.
They reported that viscosity increases with particle volume fraction
increase when the particle volume fraction is larger than 0.4%, and
increases significantly with temperature after 55 ◦ C. Phuoc et al.
[21] measured the influence of particle volume fraction on viscos-
ity for MWCNTs dispersed in distilled water and found a similar
conclusion available in the literature that viscosity of test fluids
increases accordingly with increasing particle loadings.
Since our prepared VFBNs have base fluid with viscoelasti-
city, non-Newtonian fluid properties are naturally expected for
such fluids. It is also the main purpose to conduct the present
study, i.e. exploring new flow media with viscoelasticity as well as
nanoparticles suspension. A stress-controlled rheometer (Kinexus
Pro, Malvern instruments, UK) with concentric rotational cylin-
Fig. 2. Typical SEM photos of the prepared VFBNs containing suspensions of MWC-
ders was employed to measure the shear-rate-dependent shear NTs. (a) 200 ppm aqueous CTAC/NaSal solution based nanofluids with 0.1 vol% of
viscosity of our prepared VFBNs. The available torque of the used MWCNT1 and (b) 200 ppm aqueous CTAC/NaSal solution based nanofluids with
rheometer ranges from 0.05 ␮N m to 200 mN m, indicating an out- 0.1 vol% of MWCNT3.
standing precision of shear viscosity measurement.
is essentially uniform without big entanglements; the lengths of
3. Results and discussion MWCNT1 are indeed longer than that of MWCNT3. Macroscopi-
cally, the prepared VFBNs containing MWCNTs are all uniform black
The microstructures formed by MWCNTs in the VFBNs were at colored liquids and can keep stable for at least one month (time
first visualized. Then thermal conductivities and shear viscosities window of our present experiments) without obvious depositions.
were measured for VFBNs with different concentrations of the base In order to gain more concrete information about the size and
fluid and different volume fractions of MWCNTs at various temper- shape of MWCNTs in the prepared VFBNs, we diluted the MWC-
atures. The results are illustrated from the abovementioned aspects NTs suspensions and took SEM pictures with finer resolution (by
in the following sections, respectively. Quanta 200FEG, FEI Company, USA). Fig. 3(a) depicts several fibrous
twisted MWCNTs for MWCNT1, which clearly shows that the length
3.1. SEM images of suspensions of MWCNT1 is around 20 ␮m. Fig. 3(b) demonstrates an exam-
ple of the visualized diameter of MWCNT1 in the prepared VFBNs.
Fig. 2 shows the typical scanning electron microscope (SEM) Using the scale bar, it is speculated that the diameter of MWCNT
images (by S-4700, HITACHI, Japan) of VFBNs containing MWCNTs. structure formed in the suspension is about 200 nm. After taking
From these pictures, we can observe the morphologies of MWCNTs into account the diameter (8 nm) of original MWCNT1 and thick-
in the prepared VFBNs. Fig. 2(a) represents the 200 ppm aqueous ness (20 nm) of sprouting gold, we finally obtain the thickness of
CTAC/NaSal solution based nanofluids with 0.1 vol% of MWCNT1, wrapping layer, which is about 76 nm. Shenogin et al. [22] utiliz-
and Fig. 2(b) shows the VFBN for the same base fluid added with ing molecular dynamics simulations obtained a series of equivalent
same volume fraction of MWCNT3. It can be seen that the nano- layers between the MWCNTs and polymer, which are in accordance
tubes in both visualized VFBNs are entangled and interweaved with with ours partially. But, in the review of Han and Fina [23], it is
ambient fibers. Since the samples prepared for the SEM observa- reported that the thermal contact resistance of MWCNTs suspen-
tions cannot be made extremely thin (which are actually overlaps sion corresponds to a polymer layer with thickness of only 10 nm,
of liquid layers), it is hardly to acquire more detailed information which is much smaller than 76 nm. This significant difference is
about the nanotubes within the resolution limitation of the present probably caused by high viscosity of dispersants and great specific
SEM images, except for the facts that the dispersion of MWCNTs surface of MWCNTs we used.
 Author's personal copy

50 F.-C. Li et al. / Thermochimica Acta 556 (2013) 47–53

Fig. 4. Comparisons of thermal conductivities between the viscoelastic CTAC/NaSal

solution and distilled water.

Fig. 5. Ratios of thermal conductivitiesof VFBNs with MWCNT1 nanoparticles (ke )

to that of base fluid (kf ) as a function of volume fraction.

Fig. 3. Typical SEM pictures of diluted MWCNT1. (a) The morphology of twisted Choi et al. [7], while slightly higher than that of Xie et al. [8]. This
MWCNTs. (b) A branch structure of MWCNT. discrepancy may be caused by the differences in CNTs’ specifica-
tion and dispersion situation. There are four concentrations of the
base fluid, which are 0 ppm (distilled water), 50 ppm, 200 ppm and
3.2. Thermal conductivities of VFBNs containing MWCNTs 1000 ppm aqueous CTAC/NaSal solution, respectively. Each original
aqueous solution is used as the base fluid to prepare a category of
At first, the thermal conductivities of the viscoelastic base flu- nanofluids, which contain 0.05 vol%, 0.2 vol%, 1.0 vol%, 1.6 vol% and
ids of the prepared VFBNs, i.e. aqueous CTAC/NaSal solutions, as 3.5 vol% of MWCNTs, respectively. On one hand, as shown in Fig. 5,
a function of the fluid temperature are plotted in Fig. 4. It shows the four categories of VFBNs all show a similar enhancement trend
that both amplitude and variation trend with temperature of the in the thermal conductivities with the increase of volume fraction
thermal conductivities have little differences between CTAC/NaSal of MWCNTs.
solutions and distilled water. Therefore, we can make sure one On the other hand, the VFBNs with denser aqueous CTAC/NaSal
important point that the enhancement of thermal conductivity of solution as base fluids show slightly higher thermal conductivities
VFBN, as shown in the following section, is not due to the addition than those with base fluids of lower concentrations. This phe-
of CTAC and NaSal into water. nomenon can be explained by the changes of surface tension of
The thermal conductivities of VFBNs with MWCNTs (MWCNT1) different concentrated base fluids, which contain the additions
particles at 299.15 K as a function of particle volume fractions are of both dispersant and CTAC/NaSal. Fig. 6 depicts the variations
shown in Fig. 5. Overall, the thermal conductivities of the above in surface tension (measured by BZY-1 surface tension meter,
nanofluids increase monotonously with the particle volume frac- Shanghai HengPing Instruments Factory, China) versus concen-
tions, and are significantly higher than that of the corresponding tration of aqueous CTAC/NaSal solution. It shows that, with the
base fluids. It indicates that flow of the prepared VFBNs should increase of solution concentration, the surface tensions of base flu-
have better convective heat transfer performance as compared ids in two categories decrease rapidly: from 42 mN/m for distilled
with its viscoelastic base fluid. For the case of 1000 ppm aque- water (containing dispersant TNDIS) to 32 mN/m for 1000 ppm
ous CTAC/NaSal base fluid, the thermal conductivities are increased CTAC/NaSal solution. It is well known that the interfacial resistance
by 16.9% and 30.6% at 1.6 vol% and 3.5 vol% nanoparticle volume between CNTs and matrix primarily depends on surface wettability.
fractions, respectively, which are much smaller than the results of And the surface wettability is determined by the value of surface
 Author's personal copy

F.-C. Li et al. / Thermochimica Acta 556 (2013) 47–53 51

Fig. 8. Temperature-dependent thermal conductivities of VFBNs with 200 ppm

Fig. 6. Concentration dependence of surface tension for different base fluids. CTAC/NaSal solution as the base fluid and 1.6 vol% CNTs with different diameters
and lengths.

tension. Namely, a material with a smaller surface tension owns

nice surface wettability, representing lower interfacial resistance For nanoparticle dependency, firstly, the thermal conductivities of
value. The surface tension of 1000 ppm aqueous CTAC/NaSal solu- VFBN containing MWCNT1 (8 nm outer diameter) are gently higher
tion (CTAC itself is surfactant) is smaller than other three categories’ than that with MWCNT2 (10–20 nm outer diameter). This result is
solutions. Thus the interfacial thermal resistance (the thermal resis- in good agreement with the conclusion of Fujiii et al. [24]. It is said
tance between nanoparticles and base fluids) is lower in 1000 ppm that the thermal conductivity of MWCNTs increases as its diameter
CTAC/NaSal solution based nanofluid and the corresponding ther- decreases. Secondly, at room temperature, there is no obvious dif-
mal conductivity is higher. ference in thermal conductivities between VFBNs with MWCNT1
Fig. 7 shows the temperature-dependent thermal conductivi- and with MWCNT3. Han and Fina [23] and Chiu et al. [25] stated
ties of VFBNs with 1000 ppm CTAC/NaSal solution as base fluid at 0, that the thermal conductivity of CNTs remains unchanged when
0.05, 1.0, 3.5 vol% volume fraction of MWCNTs (MWCNT1), respec- the tube length is much longer than the mean free path of CNT,
tively. Generally, the thermal conductivities nonlinearly increase which is in accordance with our experimental data. However, this
with the increase of particle volume fraction and temperature. Sim- balance is broken beyond 310 K. The thermal conductivity of VFBN
ilar results have also been reported in many previous studies. It with MWCNT3 becomes larger than that with MWCNT1. The rea-
means that the thermal conductivity of the base fluid can be sig- son of this phenomenon probably lies in the difference in intensity
nificantly improved by adding some nanoparticles. And when the of the Brownian motions at elevated temperature. The CNT with a
temperature rises, the intensity of Brownian motions augments shorter length can move faster than that with a longer tail which
greatly, which leads to a corresponding enhancement of thermal may more or less entangle with each other. Thereby, the conductive
conductivity also. heat transfer condition in MWCNT3 solution becomes better than
Fig. 8 shows the temperature-dependent and nanoparticle- MWCNT1 solution.
dependent thermal conductivities for VFBNs with 200 ppm Since the Maxwell equation was proposed [26] more than
CTAC/NaSal solution as the base fluid and 1.6 vol% MWCNT1, a century ago, many improved models for the effective ther-
MWCNT2 and MWCNT3 as nanoparticles, respectively. For tem- mal conductivity of nanofluids have been proposed. Fig. 9 shows
perature dependency, the same as that shown in Fig. 7, the thermal comparisons between the presently measured thermal conduc-
conductivities of test fluids increase with the temperature increase. tivities of VFBNs containing MWCNTs (MWCNT1) and predictions
by five different models, i.e. Maxwell model, Hamilton–Crosser
model [27], Jang–Choi model [28], Li–Qu–Feng model [29] and the

Fig. 7. Effect of temperature on thermal conductivities of VFBNs at different volume Fig. 9. Comparisons between experimental results and predictions by theoretical
fractions of MWCNTs (MWCNT1). models for different kinds of VFBNs at 299.15 K.
 Author's personal copy

52 F.-C. Li et al. / Thermochimica Acta 556 (2013) 47–53

Table 2
Five models for thermal conductivity evaluated in the present paper.

Models Equations
kp +2kf +2(kp −kf )ϕ
Maxwell [26] ke = k
kp +2kf −(kp −kf )ϕ f
kp +(n−1)kf +(n−1)(kp −kf )ϕ
Hamilton–Crosser [27] ke = kp +(n−1)kf −(kp −kf )ϕ
kf
Jang–Choi model [28] ke = (cos2 )ˇkp ϕ + (1 − ϕ)kf
kpn +2kf +2ϕeff (kpn −kf ) ϕ cp
 KB T
Li–Qu–Feng model [29] ke = k + eff
kpn +2kf −ϕeff (kpn −kf ) f
kpn +(n−1)kf +(n−1)ϕeff (kpn −kf )
2

3rc 
ϕeff cp KB T
Modified Li–Qu–Feng model ke = kpn +(n−1)kf −ϕeff (kpn −kf )
kf + 2 3rc 

modified Li–Qu–Feng model that is proposed presently as

introduced later, respectively. Table 2 gives the detailed informa-
tion of these models, where ke , kp , kf , and kpn are the thermal
conductivity of nanofluids, CNTs, base fluids and effective thermal
conductivity of CNTs [29], respectively; ϕ is the volume fraction Fig. 10. Viscosity as a function of shear rate for MWCNTs (MWCNT1) nanofluid.
of CNTs; n is the empirical shape factor (for CNTs, n = 6); ˇ is the
Kapitza resistance;  is the angle between a given direction and
a fiber axis, ϕeff is the effective volume fraction of nanoparticles fluid of the prepared VFBNs, aqueous solution of CTAC/NaSal, is a
considering the nanolayer of nanoparticles [29]; KB is the Boltz- typical viscoelastic fluid. It is thus natural to expect that VFBNs
mann constant KB = 1.3806505 × 10−23 J/K; rc is the average size of also behave non-Newtonian characteristics. The measured shear-
CNTs;  is the viscosity of the nanofluid. As mentioned in Fig. 4 rate-dependent shear viscosity profiles for 1000 ppm CTAC/NaSal
the thermal conductivities of the base fluids are very close to each aqueous solution (with dispersant TNDIS) and VFBNs are shown in
other, so, we neglect the tiny difference and keep kf = 0.61 as a con- Fig. 10. It can be clearly seen that the shear viscosity of all the test
stant at 299.15 K. Note that the Jang–Choi model shown in Table 2 fluids decreases dramatically at first, then gradually decrease to a
is actually a simplified expression which overlooks the Brownian relatively stable value 0.00202 Pa s beyond 50 s−1 shear rate. These
motions of CNTs. The parameters ˇ and cos2  are chosen to be results show a typical non-Newtonian behavior. This shear thinning
0.006 and 1/3, respectively [30]. From Fig. 9, one can clearly see effect can be interpreted by possible realignment of bundled struc-
a significant overall underestimation of the thermal conductivi- ture in the direction of the shearing force, leading to smaller viscous
ties of suspensions by Maxwell model. And the Hamilton–Crosser drag. Besides, the shear viscosities of VFBNs increase obviously as
model shows a comparative agreement with experimental result the particle volume fraction increases.
at low nanoparticle volume fraction, while exhibits great discrep- Although the elastic features of the prepared VFBNs, such
ancy at larger volume fraction. The Jang–Choi model matches as relaxation time and extensional viscosity have not yet been
with the data pretty well through the whole measured range of obtained directly due to the lack of necessary apparatus, it is still
nanoparticle volume fraction. The maximum deviation is within expectable that this kind of VFBN with suspensions of MWCNTs has
4%. These divergences can be interpreted by the fact that both drag-reducing ability in its turbulent flow based on the following
the Maxwell and Hamilton–Crosser models are initially derived three viewpoints: (1) the shear viscosity profiles for the prepared
for particles larger than micrometer in diameter and no particle VFBNs show similar non-Newtonian shear-thinning features com-
size, Kapitza resistance, particle clustering have been taken into pared with that of the viscoelastic base fluid; (2) slightly reverse
account. The Jang–Choi model [28], however, was derived from rotation behavior occurred after stopping to rotate one prepared
molecular dynamics theory and the effects of contact resistance VFBN (with 1000 ppm CTAC/NaSal solution as the base fluid and
and particle aggregation on thermal conductivity of nanofluids containing 1.0 vol% MWCNT1 as an example) in a beaker, showing
were all exhibited. The Li–Qu–Feng model [29], which considers the viscoelasticity qualitatively; (3) we do have obtained signifi-
the effect of Brownian motion, liquid layering around nanoparti- cant DR in the previous experimental study on turbulent flow of
cles, and clustering together, shows a higher result than Maxwell VFBN with CTAC/NaSal solution as the viscoelastic base fluid and
model, but lower than Hamilton–Crosser model, and underesti- Cu nanoparticles as suspensions [18], and it is reasonable to con-
mates the experiment results. The reason can be explained by that, jecture that the two kinds of VFBNs containing MWCNTs and Cu
for the special shape of CNTs, the shape factor must be consid- nanoparticles, respectively, should have qualitatively similar rhe-
ered in the static part of thermal conductivity model, while in the ological properties. Both the turbulent drag-reducing ability and
original Li–Qu–Feng model, no shape factor was included. In con- heat transfer enhancement ability of the prepared VFBNs contain-
sideration of the shape factor of CNTs, the modified Li–Qu–Feng ing MWCNTs will be testified in the future experiments.
model is then proposed in the present paper and shown in Table 2.
We can see from Fig. 9 that the predicted thermal conductivities 4. Conclusions
of VFBN containing CNTs by the modified Li–Qu–Feng model also
show excellent agreements with the measured data. The modi- VFBNs with viscoelastic aqueous solution of CTAC/NaSal as the
fied Li–Qu–Feng model is more suitable for nanofluids with special base fluid and MWCNTs as nanoparticles have been prepared and
shaped other than spherical nanoparticles. their morphologies, thermal conductivities and shear viscosities
have been studied experimentally. The following conclusions are
3.3. Shear viscosities of VFBNs containing MWCNTs drawn from the results.

Since the motivation of preparing VFBN is to combine the (a) The base fluid of aqueous solution of CTAC/NaSal shows no
enhanced thermal performance of nanofluid and the turbulent drag essential difference in its thermal conductivity as compared
reducing ability of viscoelastic fluid, after measuring the thermal with water.
conductivity of our prepared VFBNs, it is necessary to study the (b) Overall, the thermal conductivities of the prepared VFBNs
rheological properties before exploring the characteristics of tur- are enhanced with increasing temperature and nanoparti-
bulent flow and convective heat transfer of such fluids. The base cle volume fraction, showing a potential ability of enhancing
 Author's personal copy

F.-C. Li et al. / Thermochimica Acta 556 (2013) 47–53 53

convective heat transfer. The measured thermal conductivities [8] H. Xie, H. Lee, W. Youn, M. Choi, Nanofluids containing multiwalled carbon
of VFBNs are in good agreement with the predictions using nanotubes and their enhanced thermal conductivities, J. Appl. Phys. 94 (2003)
4967–4971.
Jang–Choi model [28]. [9] T.P. Teng, Y.H. Hung, C.S. Jwo, C.C. Chen, L.Y. Jeng, Pressure drop of TiO2
(c) A new model considering the shape factor of CNTs is proposed in nanofluid in circular pipes, Particuology 9 (2011) 486–491.
the present paper by modifying the Li–Qu–Feng model [29]. The [10] B.A. Toms, Some observation on the flow of linear polymer solutions through
straight tubes at large Reynolds number, in: Proceedings of the First Interna-
predicted thermal conductivities by this model show excellent tional Congress of Rheology, vol. 2, North Holland, Amsterdam, 1949, p. 135.
agreements with the measured data. [11] F.C. Li, Y. Kawaguchi, T. Segawa, K. Hishida, Reynolds-number dependence
(d) The VFBNs with larger mass concentration of CTAC/NaSal show of turbulence structures in a drag-reducing surfactant solution channel flow
investigated by PIV, Phys. Fluids 17 (2005), 075104.
higher thermal conductivities than that with lower concentra-
[12] F.C. Li, Y. Kawaguchi, K. Hishida, Investigation on the characteristics of turbu-
tions of CTAC/NaSal. This can be explained by the decrease of lence transport for momentum and heat in a drag-reducing surfactant solution
surface tensions of the base fluids. flow, Phys. Fluids 16 (2004) 3281–3295.
[13] F.C. Li, Y. Kawaguchi, K. Hishida, Structural analysis of turbulent transport in a
(e) The thermal conductivity of suspensions with MWCNTs
heated drag-reducing channel flow with surfactant additives, Int. J. Heat Mass
increases as the CNT diameter decreases. At room temperature, Transfer 48 (2005) 965–973.
the thermal conductivities of test fluids with MWCNT1 (long) [14] P.W. Li, Y. Kawaguchi, H. Daisaka, A. Yabe, K. Hishida, M. Maeda, Heat
and with MWCNT3 (short) are similar, while after 310 K the transfer enhancement to the drag-reducing flow of surfactant solution in
two-dimensional channel with mesh-screen inserts at the inlet, ASME J. Heat
thermal conductivity of suspension with MWCNT3 is larger than Transfer 123 (2001) 617–623.
that with MWCNT1. [15] Y.Y. Qi, Y. Kawaguchi, R.N. Christensen, J.L. Zakin, Enhancing heat transfer ability
(f) The measured results for shear viscosity show that the pre- of drag reducing surfactant solutions with static mixers and honeycombs, Int.
J. Heat Mass Transfer 46 (2003) 5161–5173.
pared VFBNs, similar to its viscoelastic base fluid, behave [16] J.C. Yang, F.C. Li, W.W. Zhou, Y.R. He, B.C. Jiang, Experimental investigation
non-Newtonian features: shear-thinning in its shear viscosity. on the thermal conductivity and shear viscosity of viscoelastic-fluid-based
It is expectable that the prepared VFBNs may also have drag- nanofluids, Int. J. Heat Mass Transfer 55 (2012) 3160–3166.
[17] Z.H. Liu, L. Liao, Forced convective flow and heat transfer characteristics of
reducing ability in turbulent flows. aqueous drag-reducing fluid with carbon nanotubes added, Int. J. Therm. Sci.
49 (2010) 2331–2338.
Acknowledgments [18] J.C. Yang, F.C. Li, W.W. Zhou, H.P. Xu, Y.R. He, Y.M. Huang, B.C. Jiang, Experimen-
tal study on the characteristics of heat transfer and flow resistance in turbulent
pipe flows of viscoelastic-fluid-based Cu nanofluid, Int. J. Heat Mass Transfer,
This work was sponsored by National Natural Science Foun- under review.
dation of China (51076036), Foundation for Innovative Research [19] I.M. Mahbubul, R. Saidur, M.A. Amalina, Latest developments on the viscosity
of nanofluids, Int. J. Heat Mass Transfer 55 (2012) 874–885.
Groups of the National Natural Science Foundation of China
[20] L. Chen, H. Xie, Y. Li, W. Yu, Nanofluids containing carbon nanotubes treated by
(51121004), Specialized Research Fund for the Doctoral Pro- mechanochemical reaction, Thermochim. Acta 477 (2008) 21–24.
gram of Higher Education of China (20112302110020) and [21] T.X. Phuoc, M. Massoudi, T.H. Chen, Viscosity and thermal conductivity of
nanofluids containing multi-walled carbon nanotubes stabilized by chitosan,
the Fundamental Research Funds for the Central Universities
Int. J. Therm. Sci. 50 (2011) 12–18.
(HIT.BRET1.2010008). [22] S. Shenogin, L.P. Xue, R. Ozisik, P. Keblinski, Role of thermal boundary resis-
tance on the heat flow in carbon-nanotube composites, J. Appl. Phys. 95 (2004)
8136–8144.
References
[23] Z.D. Han, A. Fina, Thermal conductivity of carbon nanotubes and their polymer
[1] S.U.S. Choi, Enhancing thermal conductivity of fluids with nanoparticles, in: D.A. nanocomposites: a review, Prog. Polym. Sci. 36 (2011) 914–944.
Singer, H.P. Wang (Eds.), Developments and Applications of Non-Newtonian [24] M. Fujiii, X. Zhang, H. Xie, Measuring the thermal conductivity of a single carbon
Flows, FED-vol. 231/MD-vol. 66, ASME, New York, 1995, pp. 99–105. nanotube, Phys. Rev. Lett. 95 (2005) 065502.
[2] S. Özerinç, S. Kakaç, A.G. Yazıcıoğlu, Enhanced thermal conductivity of nanoflu- [25] H.Y. Chiu, V.V. Deshpande, Ballistic phonon thermal transport in multiwalled
ids: a state-of-the-art review, Microfluid Nanofluid 8 (2010) 145–170. carbon nanotubes, Phys. Rev. Lett. 95 (2005) 2261011.
[3] S.K. Das, S.U.S. Choi, A review of heat transfer in nanofluids, Adv. Heat Transfer [26] J.C. Maxwell, A Treatise on Electricity and Magnetism, 2nd ed., Oxford Univer-
41 (2009) 81–197. sity Press, Cambridge, UK, 1904, p. 435.
[4] C. Kleinstreuer, Y. Feng, Experimental and theoretical studies of nanofluid ther- [27] R.L. Hamilton, O.K. Crosser, Thermal conductivity of heterogeneous two com-
mal conductivity enhancement: a review, Nanoscale Res. Lett. 6 (2011) 229. ponent systems, Ind. Eng. Chem. Fundam. 1 (1962) 187–191.
[5] S. Iijima, Helical microtubules of graphitic carbon, Nature 354 (1991) 56–58. [28] S.P. Jang, S.U.S. Choi, Role of Brownian motion in the enhanced thermal con-
[6] P. Kim, L. Shi, A. Majumdar, P.L. McEuen, Thermal transport measurements of ductivity of nanofluids, Appl. Phys. Lett. 84 (2004) 4316–4318.
individual multiwalled nanotubes, Phys. Rev. Lett. 87 (2001) 215502. [29] Y. Li, W. Qu, J.C. Feng, Temperature dependence of thermal conductivity of
[7] S.U.S. Choi, Z.G. Zhang, W. Yu, F.E. Lockwood, E.A. Grulke, Anomalous thermal nanofluids, Chin. Phys. Lett. 25 (2008) 3319–3322.
conductivity enhancement in nanotube suspensions, Appl. Phys. Lett. 79 (2001) [30] Y. Hwang, J.K. Lee, Y.M. Lee, Stability and thermal conductivity characteristics
2252–2254. of nanofluids, Thermochim. Acta 455 (2007) 70–74.

View publication stats