Applied_Rheology_Vol_20_4:Applied_Rheology_ 08.10.

10 11:44 Seite 236

Viscosity Measurements on Colloidal Dispersions (Nanofluids)

for Heat Transfer Applications

David C. Venerus1*, Jacopo Buongiorno2, Rebecca Christianson3, Jessica

Townsend3, In Cheol Bang4,5, Gang Chen2, Sung Jae Chung16, Minking Chyu16,
Haisheng Chen6, Yulong Ding6, Frank Dubois7, Grzegorz Dzido8, Denis Funf-
schilling9, Quentin Galand7, Jinwei Gao2, Haiping Hong10, Mark Horton10, Lin-
wen Hu2, Carlo S. Iorio7, Andrzej B. Jarzebski8, Yiran Jiang1, Stephan Kabelac11,
Mark A Kedzierski12, Chongyoup Kim13, Ji-Hyun Kim4, Sukwon Kim13, Thomas
McKrell2, Rui Ni9, John Philip14, Naveen Prabhat2, Pengxiang Song15, Stefan Van
Vaerenbergh7, Dongsheng Wen15, Sanjeeva Witharana6, Xiao-Zheng Zhao9,
Sheng-Qi Zhou9

1 Illinois Institute of Technology, 10 W. 33rd St., Chicago, IL 60616, USA

2 Massachusetts Institute of Technology (MIT), 77 Massachusetts Avenue,
Cambridge, MA 02139, USA
3 Olin College of Engineering, Olin Way, Needham, MA 02492, USA
4 Ulsan National Institute of Science and Technology, School of Energy Engineering,

San 194 Banyeon-ri, Eonyang-eup, Ulju-gun, Ulsan Metropolitan City, Republic of Korea
5 Tokyo Institute of Technology, 2-12-1 Ookayama, Meguro-ku, Tokyo 152-8550, Japan
6 University of Leeds, Clarendon Road, Leeds LS2 9JT, UK
7 Universit Libre de Bruxelles, Chimie-Physique E.P. CP 165/62 Avenue P.Heger,

1050, Bruxelles, Belgium

8 Silesian University of Technology, Department of Chemical and Processing Engineering,

ul. M. Strzody 7, 44-100 Gliwice, Poland

9 Chinese University of Hong Kong, Department of Physics, G6, North Block,

Science Center, Shatin NT, Hong Kong, China

10 South Dakota School of Mines and Technology, 501 E Saint Joseph Street,

Rapid City, SD 57702, USA

11 Helmut-Schmidt University Hamburg, Institute for Thermodynamics,

22039, Hamburg, Germany

12 National Institute of Standards and Technology (NIST), MS 863, Gaithersburg, MD 20899, USA
13 Korea University, Anam-dong, Sungbuk-ku, Seoul 136-713, Republic of Korea
14 Indira Gandhi Centre for Atomic Research, SMARTS, NDED,

Metallurgy and Materials Group, Kalpakkam-603102, India

15 Queen Mary University of London, School of Engineering and Materials Science,

Mile End Road, London, E1 4NS, UK

16 University of Pittsburgh, Dept. Mechanical Engineering and Materials Science,

648 Benedum Hall, 3700 OHara Street, Pittsburgh, PA 15261, USA

* Email: [email protected]
Fax: x1.312.567.8874

Received: 2.11.2009, Final version: 15.12.2009

Abstract:
This article reports viscosity data on a series of colloidal dispersions collected as part of the International Nanoflu-
id Property Benchmark Exercise (INPBE). Data are reported for seven different fluids that include dispersions of
metal-oxide nanoparticles in water, and in synthetic oil. These fluids, which are also referred to as nanofluids,
are currently being researched for their potential to function as heat transfer fluids. In a recently published paper
from the INPBE study, thermal conductivity data from more than 30 laboratories around the world were report-
ed and analyzed. Here, we examine the influence of particle shape and concentration on the viscosity of these
same nanofluids and compare data to predictions from classical theories on suspension rheology.

Appl. Rheol. 20 (2010) 44582 DOI: 10.3933/ApplRheol-20-44582

Applied Rheology
44582 -1 Volume 20 Issue 4
Applied_Rheology_Vol_20_4:Applied_Rheology_ 08.10.10 11:44 Seite 237

Zusammenfassung:
In diesem Beitrag werden experimentelle Daten zur Viskositt verschiedener kolloidaler Dispersionen vorge-
stellt, welche im Rahmen einer internationalen vergleichenden Ringmessung (International Nanofluid Proper-
ty Benchmark Exercise INPBE) gewonnen wurden. Es werden hierbei sieben unterschiedliche Fluide betrachtet,
unter anderem Wasser beziehungsweise synthetisches Oel mit suspendierten metalloxidischen Nanopartikeln.
Diese auch Nanofluide genannten Dispersionen werden derzeit intensiv aufgrund ihres moeglichen Potenzials
als effizientes Wrmetrger-Fluid untersucht. Vor Kurzem wurden in einem weiteren Beitrag der INPBE-Studie
Daten zur Wrmeleitfhigkeit von Nanofluiden verffentlicht, wie sie in den ber 30 teilnehmenden For-
schungslabors weltweit gemessen wurden. In diesem Beitrag wird der Einfluss der Partikel-Gestalt und der Par-
tikel-Konzentration auf die Viskositt der gleichen Nanofluide untersucht und mit Korrelationen aus der klassi-
schen Rheologie von Suspensionen verglichen.

Rsum:
Cet article prsente des donnes de viscosit sur des sries de dispersions collodales qui font partie du Inter-
national Nanofluid Property Benchmark Exercise (INPBE). Les donnes correspondent sept fluides diffrents
incluant des dispersions de nano particules doxyde mtalliques dans de leau et dans de lhuile synthtique. Ces
fluides, galement connus comme nano fluides font actuellement lobjet dtudes pour leur potentiel com-
me des fluides pour le transfert de chaleur. Dans un article rcemment publi dans le cadre dun tude INPBE,
des donnes de conductivit thermique de plus de 30 laboratoires diffrents ont t prsentes et analyses.
Dans ce travail, on tudie linfluence de la forme et de la concentration des particules sur la viscosit des nano
fluides qui en sont issus. Ces donnes sont compares aux prdictions thoriques des modles classiques sur la
rhologie de suspensions.

Sommario:
Questo articolo riporta dati sperimentali sulla viscosita di una serie di sistemi di nano-colloidi (nanofluidi) nel
contesto del programma internazionale INPBE (International Nanofluid Property Benchmark Exercise). I dati
riguardano sette diversi sistemi, comprendenti particelle di ossidi metallici in acqua ed olio sintetico. Questi flui-
di sono attualmente soggetti di interesse nella comunita scientifica per la loro possible applicazione come fluid
termici. I dati di conducibilita termica prodotti da piu di 30 laboratori internazionali sono stati riportati e discus-
si in un articolo su INPBE di recente pubblicazione. Invece qui esaminiamo linfluenza della forma e concentra-
zione delle nanoparticelle sulla viscosita degli stessi campioni, e confrontiamo i dati con i risultati delle classi-
che teorie reologiche per sospensioni colloidali.

Key words: nanofluids, colloidal dispersion, viscosity, thermal conductivity

1 INTRODUCTION further intensified by early studies reporting

anomalously large relative to classical model
Suspensions of nanometer-sized (1 100 nm), sol-
id particles, also known as colloids or colloidal dis- predictions enhancement of thermal conduc-
persions, are of great technological importance tivity. Several review papers on nanofluid heat
and can be found in applications ranging from transfer have been published [14].
inks to pharmaceuticals. For over a decade, col- Transport properties (e.g. electrical or ther-
loidal dispersions containing metallic, or metal- mal conductivity) of heterogeneous systems are
oxide, nanoparticles, sometimes referred to as often estimated using effective medium theory.
nanofluids, have been the focus of intensive For example, the effective thermal conductivity
research because of their potential as heat trans- k of a suspension composed of solid particles
fer fluids. This research has been driven by the with thermal conductivity kp at volume fraction
possibility of having fluids with enhanced ther- f ( 1) dispersed in a liquid of thermal conductiv-
mal conductivity, relative to the suspending liq- ity kf ( kp) can be written as
uid (basefluid), while mitigating problems asso-
ciated with erosion, sedimentation and clogging
that plague suspensions of larger particles. Inter-
est in nanofluids as heat transfer fluids has been (1)

Applied Rheology
Volume 20 Issue 4 44582 -2
Applied_Rheology_Vol_20_4:Applied_Rheology_ 08.10.10 11:44 Seite 238

where [k] is defined as follows: [k] = limf0{(k/kf ratio p, [h] p [7]. The linear dependence of h0 on
- 1)/f}. Over 100 years ago, Maxwell [5] derived particle volume fraction in Equation 2 typically
an expression for the thermal (electrical) con- holds for f 0.03 [7].
ductivity of a dilute system of spherical particles In addition to thermal conductivity, it is also
that yielded [k] = 3. Models for the effective ther- evident that viscosity has a significant impact on
mal conductivity of heterogeneous systems have the overall performance of a heat transfer fluid.
been formulated to account for the effects of Clearly, pumping a fluid with increased viscosity
non-spherical particle shape, particle-particle through a heat exchanger requires an increase in
interactions, and interfacial resistance between pumping energy, thereby reducing the overall ben-
the particle and continuous phases [14]. efit of a higher thermal conductivity fluid. In light
There have been numerous experimental of this, there have been a number of studies report-
studies reporting thermal conductivity data on ing the viscosity of nanofluids [8-14]. These studies
nanofluids. Much of the reported data show both considered nanofluids composed of metal-oxide
quantitative and qualitative differences from the (Al2O3, CuO, TiO2, Fe2O3) and metal (Cu), nanometer-
effective medium theory prediction in Equation 1. sized, spherical particles in water, ethylene glycol,
For example, in many early studies, the measured or hydrocarbon oil at concentrations f 0.05. For
levels of thermal conductivity enhancement for all of these nanofluids, Newtonian behavior was
dilute nanofluids (f 0.01) were significantly larg- observed, and the dependence of relative viscosity
er than the classical prediction given in Equation h/hf was approximately linear in particle concen-
1 with [k] = 3. Several investigators have reported tration f. However, the reported intrinsic viscosity
a dependence of thermal conductivity enhance- [h] for these nanofluids was in the range 4-16, which
ment on particle size and on temperature, neither is substantially larger than the theoretical value of
of which is consistent with Equation 1. In addition 2.5. Prasher et al., [9] and Garg et al., [12] proposed
to generating a great deal of optimism about the criteria for the overall effectiveness of nanofluids
potential of nanofluids for use as heat transfer flu- as heat transfer fluids suggesting that [h] should be
ids, early experimental results also sparked theo- 4 5 times smaller than [k] appearing in Equation
retical interest to identify the mechanism(s) 1. The effect of changes in specific heat due to the
responsible for the observed departures from presence of nanoparticles on the effectiveness of
effective medium theory predictions. Further dis- nanofluids as heat transfer fluids has also been
cussion of both experimental and theoretical investigated [15].
research on nanofluids can be found elsewhere At the first scientific conference centered on
[14]. At present, the study of nanofluids as heat nanofluids (Nanofluids: Fundamentals and Appli-
transfer fluids remains both an active and contro- cations, September 16-20, 2007, Copper Mountain,
versial field of research. Colorado), it was decided to launch an Interna-
It is well known that the addition of solid tional Nanofluid Property Benchmark Exercise
particles to a liquid can significantly alter its rhe- (INPBE), to resolve the inconsistencies in the data-
ological behavior. In shear flows, the apparent base and help advance the debate on nanofluid
viscosity of a fluid h is ratio of the shear stress to properties. Thermal conductivity data on eight dif-
the shear rate: h = s/g . The zero-shear rate vis- ferent nanofluids from this study, involving over
cosity is defined as: h0 = limg 0 {h}. The zero-shear 30 laboratories around the world, are contained in
rate viscosity h0 of a dilute suspension with liq- a recently published paper [16]. A subset of INPBE
uid phase viscosity hf can be expressed as follows: participants also collected viscosity data on the
same set of nanofluids. These data are reported in
this paper. The methodology and samples used for
(2) the INPBE are described in Section 2. The viscosity
data are presented and discussed in Section 3. The
where the intrinsic viscosity [h] of the suspension findings of this study are summarized in Section 4.
is defined as: [h] = limf0 {(h0/hf - 1)/f}. The well-
known prediction of Einstein [6] for a suspension 2 INPBE METHODOLOGY
of spherical particles gives [h] = 2.5, while for Eight different nanofluids were distributed in four
dilute suspensions of particles with large aspect sets to participating INPBE laboratories. To mini-

Applied Rheology
44582 -3 Volume 20 Issue 4
Applied_Rheology_Vol_20_4:Applied_Rheology_ 08.10.10 11:44 Seite 239

Figure 1:
Viscosity versus shear rate
for spherical Al2O3 nanopar-
ticles in oil: filled symbols
f = 0.01 (S1S3); open sym-
bols f = 0.03 for (S1S4).
Different symbols indicate
data from different labora-
tories. Dashed line is aver-
age for oil base fluid (S1S7).
mize spurious effects due to nanofluid prepara-
Table 1:
tion and handling, all organizations were given Characteristics of nanofluid
identical samples from these sets, and were asked and base fluid samples.
to adhere to the same sample handling protocol.
The exercise was semi-blind, as only minimal
information about the samples was given to the
participants at the time of sample shipment. The of deformation applied to the sample. In one cat-
data were then collected and posted, on the INPBE egory, that includes gravity driven capillary, con-
website (http://mit.edu/nse/nanofluids/bench- centric cylinder and immersion viscometers,
mark/index.html). Additional details of the INPBE sample deformation is non-homogeneous and
methodology and list of participants can be found the (average) rate of deformation is usually not
in the paper containing the thermal conductivity known. In the other category are viscometers
data [16] and at the INPBE website. that apply a controlled and (approximately) uni-
Table 1 gives the main characteristics of the form deformation to the sample such as cone-
nanofluids used in the INPBE study; the sample plate viscometers. Only two of the 10 samples
names in the first column of the table indicate considered in this study showed non-Newtonian
Set#Sample# as designated in the previous behavior (i.e., viscosity h was dependent on shear
paper [16]. Samples S1S1-S1S7 were provided by rate g ), so that data from both types of viscome-
Sasol. Sample S1S2 is simply de-ionized H2O and ters are included for the eight Newtonian fluids.
is the base fluid for S1S1; sample S1S7 is a mixture For the two samples displaying non-Newtonian
of poly-alpha olefin (PAO) oil (SpectraSyn-10 by behavior, only data from cone-plate viscometers
Exxon Oil) and 5%wt. dispersant (Solsperse will be presented. Finally, all data reported here
21000 by Lubrizol Chemical) and is the base flu- were collected at room temperature, which was
id for samples S1S3-S1S6. The nanofluids desig- in the range 20-26 C. We recognize that changes
nated as S1S1 and S1S3-S1S6 contain alumina in viscosity of roughly 10% are possible over this
(Al2O3) nanoparticles with the shapes and sizes range of temperatures.
indicated in Table 1. Sample S3S1 was supplied by
W. R. Grace & Co. and is a colloidal dispersion of 3 RESULTS AND DISCUSSION
silica (SiO2) nanoparticles dispersed in de-ionized First we examine the viscosity versus shear rate
H2O stabilized by the addition of small amounts behavior of several fluids listed in Table 1. Data
of Na2SO4 (pH = 9). Finally, samples in Set 4 were for samples S1S3 and S1S4 from four different lab-
supplied by Dr. Jorge Gustavo Gutierrez of the oratories are shown in Figure 1. The dashed line
University of Puerto Rico Mayaguez (UPRM). in Figure 1 represents the average viscosity of the
Sample S4S2 is a solution of de-ionized H2O (75% base fluid, sample S1S7. From this figure, it is clear
wt) and tetramethylammonium hydroxide that these nanofluids are Newtonian over the
[(CH3)4NOH] stabilizer (25% wt). Sample S4S1 con- range of shear rates considered. This is expected
tains Mn-Zn ferrite (Mn1/2-Zn1/2-Fe2O4) nanoparti- given that these fluids contain spherical particles
cles dispersed in a based fluid S4S2. To summa- at relatively modest volume fractions (f = 0.01,
rize, this study involves seven nanofluids and 0.03); this is consistent with previous results on
three base fluids. Note that one of the eight
nanofluids that were part of the INPBE study, a
dilute colloidal gold dispersion, is excluded from
this paper because its rheology was indistin-
guishable from that of water. Additional details
concerning these samples and their characteri-
zation can be found elsewhere [16].
Viscosity data on each of the fluids listed in
Table 1 were reported by approximately 8-10 dif-
ferent laboratories. All reported data were
obtained using well-established methods with
commercial viscometers that can be broadly sep-
arated into two categories according to the type

Applied Rheology
Volume 20 Issue 4 44582 -4
Applied_Rheology_Vol_20_4:Applied_Rheology_ 08.10.10 11:44 Seite 240

Figure 2 (left above):

Viscosity versus shear rate
for spherical SiO2 nanoparti-
cles in water:
f = 0.32 (S3S1). Different
symbols indicate data from
different laboratories. Inset
shows relative viscosity ver-
sus dimensionless shear rate
for one set of data.

Figure 3 (right above):

Viscosity versus shear rate
for rod-shaped Al2O3
nanoparticles in oil: filled
symbols f = 0.01 (S1S5);
open symbols f = 0.03 for
(S1S6). Different symbols
indicate data from different inset to Figure 2 shows a commonly used means
laboratories. Dashed line is
average for oil base fluid of rescaling viscosity and shear rate to demon-
(S1S7). strate the competition between convection
(flow) and diffusion (Brownian motion) in sus-
Figure 4 (below): pensions [7].
Viscosity of spherical Al2O3
nanoparticles in oil as mea- Viscosity versus shear rate data from four
sured by different laborato- different laboratories for the nanofluids com-
ries: filled symbols are for prised of rod-shaped Al2O3 particles in oil (S1S5
f = 0.03 (S1S4); open sym- and S1S6) are shown in Figure 3. The dashed line
bols are for base fluid (S1S7).
Solid and dashed lines indi- in Figure 3 represents the average viscosity of the
cate average values. base fluid (S1S7). From this figure, we see that the
lower concentration (f = 0.01) fluid is Newton-
ian, while the higher concentration (f = 0.03) flu-
id is non-Newtonian. Again, agreement between
similar systems [8 - 14]. It is also evident from Fig- the different laboratories is good. Comparison of
ure 1 that there is good agreement between the the data in Figures 1 and 3 for f = 0.03 shows, as
data reported by the different laboratories with expected, that the viscosity of the fluid with non-
variations of approximately 10%. Results similar spherical particles is more sensitive to shear com-
to those shown in Figure 1 were obtained for sam- pared to the fluid with spherical particles. For
ples S1S1, S1S7, S4S1 and S4S2; therefore only the non-spherical particles, non-Newtonian effects
average viscosity of these Newtonian fluids will are the result of flow-induced perturbation of
be presented later in this section. particle orientations from the random state that
Figure 2 shows the viscosity versus shear exists at equilibrium [7]. In this case, it is a com-
rate behavior of the concentrated colloidal silica petition between the tendency for flow to orient
dispersion sample S3S1. At low shear rates, these particles and thermally driven rotation of parti-
data appear to approach a constant viscosity h0 cles.
with a value of approximately 0.4 Pa-s, and a con- Viscosity data for the nanofluid comprised
stant viscosity at shear rates greater than 10 s-1. of spherical Al2O3 particles with f = 0.03 (S1S4)
At intermediate shear rates, between 0.1 - 10 s-1, and its oil base fluid (S1S7) as reported by differ-
pronounced shear thinning is observed. The scat- ent INPBE participants are shown in Figure 4. For
ter among the data from different laboratories is a given laboratory, the reported uncertainty was
larger in Figure 2 than Figure 1, and is probably typically around 5%, which is roughly the size
the result of the higher sensitivity to shear his- of the symbols in Figure 4. The data in this figure
tory of the more concentrated system, as give an idea of consistency of viscosity data
explained next. Shear thinning behavior, which reported by different laboratories using different
is typical for concentrated suspensions of spher- techniques, which is within approximately
ical particles, can be explained by the flow- 20%. Similar results were found for the other
induced perturbation of particle positions [7]. nanofluids used in this study.
The rate of particle displacement by flow is giv- The relative viscosity versus particle con-
en roughly by the shear rate g , and the rate of centration for nanofluids with spherical particles
particle movement by Brownian motion is and rod-shaped particles are shown in Figures 5
roughly kBT/hd3. Hence, when flow perturbs par- and 6 respectively. As expected, the data in both
ticle positions at a rate faster than they are recov- figures indicate a linear dependence of h/hf on f.
ered by thermal fluctuations, the viscosity of the The measured dependence [h], however, is
suspension becomes shear-rate dependent. The roughly ten times larger than predicted, with

Applied Rheology
44582 -5 Volume 20 Issue 4
Applied_Rheology_Vol_20_4:Applied_Rheology_ 08.10.10 11:44 Seite 241

Figure 5 (left above):

Relative viscosity h/hf versus
particle concentration f for
nanofluids with spherical
particles: filled circles for
Al2O3 particles in oil (S1S3
and S1S4) and open triangle
for Mn1/2Zn1/2Fe2O3 in water
(S4S1). Lines indicate h/hf = 1
+ [h]f with prediction [h] =
5/2 (dashed) and fit [h] =
23.4 (solid).

Figure 6 (right above):

Relative viscosity h/hf versus
particle concentration f for
nanofluids with rod-shaped
particles: filled circles for
Al2O3 particles in oil (S1S5
and S1S6) and open circle for
Al2O3 in water (S1S1). Lines
indicate h/hf = 1 + [h]f with
prediction [h] = 8 (dashed)
and fit [h] = 70.8 (solid).

Figure 7 (left below):

Relative thermal conductivi-
ty k/kf versus particle con-
centration f for nanofluids
with spherical particles:
filled circles for Al2O3 parti-
cles in oil (S1S3 and S1S4) and
open triangle for
Mn1/2Zn1/2Fe2O3 in water
(S4S1). Data taken from
INPBE study [16]. Lines indi-
[h] = 23.4 for the spherical particles and [h] = 70.8 is different for the nanofluids with rod-shaped cate k/kf = 1 + [k]f with pre-
for the rod-shaped particles. The shear thinning particles as shown in Figure 8. Here, the observed diction [k] = 3 (dashed) and
fit [k] = 4.0 (solid).
behavior observed in alumina nanorod suspen- dependence on particle concentration is weaker
sions can also be due to shear induced reorgani- than predicted: measured [k] = 5.6, and predicted Figure 8 (right below):
zation of nanorods under confined geometry [k] = 13.1. Apparently, particle agglomeration can Relative thermal conductivi-
where the packing can become more efficient. A significantly reduce the effective thermal conduc- ty k/kf versus particle con-
centration f for nanofluids
similar result has been reported in recently pub- tivity of nanofluids with rod-shaped particles. A with rod-shaped particles:
lished data for nanofluids with spherical particles second explanation for the reduction in effective filled circles for Al2O3 parti-
[813], although somewhat lower values of [h] = thermal conductivity in these nanofluids is inter- cles in oil (S1S5 and S1S6)
416 were found. A possible explanation for and open circle for Al2O3 in
facial thermal resistance [16]. water (S1S1). Data taken
these pronounced differences between theory from INPBE study [16]. Lines
and experiment is particle agglomeration, which 4 CONCLUSIONS indicate k/kf = 1 + [k]f with
would increase the effective volume fraction of prediction [k] = 13.1 (dashed)
the particles [1014]. Indeed, light scattering Viscosity data have been collected as part of an and fit [k] = 5.6 (solid).
results on these fluids reported elsewhere [16] International Nanofluid Property Benchmark
are consistent with the occurrence of particle Exercise (INPBE). These data are from approxi-
agglomeration. mately 10 different laboratories around the
For comparison, we show the relative ther- world on a series of 10 different nanofluids and
mal conductivity versus particle concentration their base fluids. In general, the agreement
data from the INPBE study [16] in Figures 7 and 8. between different laboratories was good with
As with the viscosity data, nanofluids with both variations of approximately 20%, which, in
spherical and rod-shaped particles show a linear part, could be explained by lab-to-lab tempera-
dependence of k/kf on f. The data for the fluids ture variations. Two of seven nanofluids showed
with spherical particles in Figure 7 show a slightly shear-thinning behavior; the remaining five
stronger dependence on particle concentration showed Newtonian behavior. For nanofluids
than predicted: measured [k] = 4.0, and predicted with both spherical and rod-shaped nanoparti-
[k] = 3. This observation is consistent with the pres- cles, the dependence of viscosity (relative to the
ence of particle agglomeration in these systems, base fluid viscosity) on particle concentration
and suggests that particle clustering has a larger (volume fraction) was significantly stronger than
effect on viscosity than thermal conductivity in predicted by dilute suspension theory. This dis-
nanofluids with spherical particles. The situation crepancy was attributed to particle agglomera-

Applied Rheology
Volume 20 Issue 4 44582 -6
Applied_Rheology_Vol_20_4:Applied_Rheology_ 08.10.10 11:44 Seite 242

[7] Russel WB, Saville DA, Schowalter WR: Colloidal

Dispersions, Cambridge University Press, Cam-
bridge (1989).
[8] Das SK, Putra N, Thiesen P, Roetzel W: Pool boil-
ing characteristics of nano-fluids, Int. J. Heat
Mass Transfer 46 (2003) 851.
[9] Prasher R, Song D, Wang J: Measurements of
Map: tion. In contrast, the observed enhancement in
Map showing locations of
nanofluid viscosity and its implications for ther-
thermal conductivity was slightly larger for the mal applications, Appl. Phys. Lett., 89 (2006)
laboratories participating in
the International Nanofluid spherical particle fluids, and significantly lower 133108.
Properties Benchmark Exer- for the rod-shaped particle fluids, than predicted [10] Chen H, Ding Y, He Y, Tan C: Rheological behav-
cise (INPBE). by effective medium theory. As noted in the iour of ethylene glycol based titania nanofluids,
introduction, criteria for the overall effectiveness Chem. Phys. Lett. 444 (2007) 333.
of nanofluids as heat transfer fluids have been [11] Murshed SMS, Leong KC, Yang C: Investigations
of thermal conductivity and viscosity of nanoflu-
proposed [9, 12], which suggest [h] should be 4 -
ids, Int. J. Therm. Sci. 47 (2008) 560.
5 times smaller than [k]. Clearly, the nanofluids [12] Garg J, Poudel B, Chiesa M, Gordon JB, Ma JJ,
considered in this study would fail; this suggests Wang JB, Ren ZF, Kang WT, Ohtani H, Nanda J,
that the overall effect of adding nanoparticles to McKinley GH, Chen G: Enhanced thermal con-
the base fluid is negative in terms of heat trans- ductivity and viscosity of copper nanoparticles in
fer performance. ethylene glycol nanofluid, J. Appl. Phys. 103
(2008) 074301.
ACKNOWLEDGEMENTS [13] Anoop KB, Kabelac S, Sundararajan T, Das SK: Rhe-
ological and flow characteristics of nanofluids:
This work was made possible by the support of Influence of electroviscous effects and particle
the National Science Foundation under grant agglomeration, J. Appl. Phys. 106 (2009) 034909.
CBET-0812804. The authors are also grateful to [14] Shima PD, Philip J, Raj B: Magnetically control-
Sasol and W. R. Grace & Co. for donating some of lable nanofluid with tunable thermal conductiv-
the samples used in INPBE. ity and viscosity, Appl. Phys. Lett. 95 (2009) 133112.
[15] Bergman TL: Effect of reduced specific heats of
nanofluids on single phase, laminar internal
REFERENCES forced convection, Int. J. Heat Mass Transfer. 52
[1] Eastman JA, Phillpot SR, Choi SUS: Keblinski P, (2009) 1240.
Nanofluids for thermal transport, Annu. Rev. [16] Buongiorno J, Venerus DC, Prabhat N, McKrell T,
Mat. Res. 34 (2004) 219. Townsend J, Christianson R, Tolmachev YV,
[2] Keblinski P, Eastman JA, Cahill DG: Nanofluids for Keblinski P, Hu L-W, Alvarado JL, Bang IC, Bishnoi
thermal transport, Materials Today, June (2005) SW, Bonetti M, Botz F, Cecere A, Chang Y, Chen G,
36. Chen H, Chung SJ, Chyu MK, Das SK, Di Paola R,
[3] Das SK, Choi SUS, Patel HE: Heat transfer in Ding Y, Dubois F, Dzido G, Eapen J, EscherW, Fun-
nanofluids a review, Heat Transfer Eng. 27 fschilling D, Galand Q, Gao J, Gharagozloo PE,
(2006) 3. Goodson KE, Gutierrez JG, Hong H, Horton M,
[4] Yu W, France DM, Routbort JL, Choi SUS: Review Iorio CS, Jarzbski AB, Jiang Y, Jin LW, Kabelac S,
and comparison of nanofluid thermal conductiv- Kamath A, Kedzierski MA, Kim C, Kim JH, Kim S,
ity and heat transfer enhancements, Heat Trans- Kieng LG, Leong KC, Manna I, Michel B, Ni R, Patel
fer Eng. 29 (2008) 432. HE, Philip J, Poulikakos D, Reynaud C, Savino R,
[5] Maxwell JC: A Treatise on Electricity and Mag- Singh PK, Song P, Sundararajan T, Timofeeva E,
netism, Clarendon, Oxford (1881). Tritcak T, Turanov AN, Van Vaerenbergh S, Wen
[6] Einstein A: Eine neue Bestimmung der Molekl - D, Witharana S, Yang CC, Yeh W.-H., Zhao X-Z,
dimensionen, Ann. der Physik 19 (1906) 289; Ein- Zhou S-Q: A Benchmark Study on the Thermal
stein A: Berichtigung zu meiner Arbeit: Eine Conductivity of Nanofluids, J. Appl. Phys. 106
neue Bestimmung der Molekldimensionen, (2009) 094312.
Ann. der Physik 34 (1911) 951.

Applied Rheology
44582 -7 Volume 20 Issue 4