Experimental Thermal and Fluid Science 53 (2014) 111–118

Contents lists available at ScienceDirect

Experimental Thermal and Fluid Science

journal homepage: www.elsevier .com/locate /et fs

Experimental investigation on heat transfer enhancement of alumina/water and alumina/water–ethylene glycol nanofluids in thermallydeveloping laminar flow

0894-1777/$ - see front matter � 2013 Elsevier Inc. All rights reserved.http://dx.doi.org/10.1016/j.expthermflusci.2013.11.015

⇑ Corresponding author. Tel.: +98 2184063241.E-mail address: keshavarz@kntu.ac.ir (A. Keshavarz).

Mohammad Salemi Mojarrad, Ali Keshavarz⇑, Masoud Ziabasharhagh, Mohammad Mehdi RaznahanFaculty of Mechanical Engineering, K.N. Toosi University of Technology, Tehran, Iran

a r t i c l e i n f o

Article history:Received 5 July 2013Received in revised form 22 October 2013Accepted 16 November 2013Available online 26 November 2013

Keywords:Entrance regionAlumina nanofluidForced convective heat transferEthylene glycolLaminar flow

a b s t r a c t

In this paper, hydrodynamic and thermal behaviors of alumina/water and alumina/water–ethylene glycol50–50 by volume (WEG50) nanofluids in the thermal entrance region of a circular tube with constantwall temperature were studied experimentally. The flow regime was laminar and only hydrodynamicallyfully developed. The effects of base fluid, nanoparticles loading and Reynolds number on the convectiveheat transfer coefficient and pressure drop were studied. The experiments were conducted for 0%, 0.25%,0.5% and 0.7% nanoparticles volume fractions while Reynolds number varies between 650 and 2300. Thedynamic viscosity and the thermal conductivity were measured experimentally. Significant enhancementin nanofluids convective heat transfer coefficient was observed with respect to that of the base fluid. Theresults indicate that the average convective heat transfer and average Nusselt number increase withincreasing volume fraction as well as Reynolds number. But the average Nusselt number ratio was notimproves with increasing Reynolds number. Also it is found alumina/WEG50 nanofluids have more heattransfer increment compared to alumina/water nanofluids. The pressure drop behavior was the same asthe average convective heat transfer coefficient with the presence of nanoparticles in the base fluid.Finally the energy ratio was defined and showed adding nanoparticle to the base fluid caused incrementin energy ratio.

� 2013 Elsevier Inc. All rights reserved.

1. Introduction

A nanofluid is produced by dispersing some metallic or non-metallic nanometer sized solid particles in a base fluid like water,ethylene glycol or oil. One of the main goals of producing nanofluidis to improve heat transfer characteristics of the base fluid. Sus-pended nanoparticles have higher thermal conductivity than basefluid, thus nanofluids effective thermal conductivity and convec-tive heat transfer coefficient enhance. Recently, different experi-mental investigations on nanofluids convection have beenperformed, in both laminar [1–5] and turbulent regimes [6–9].

Choi and Eastman [10] showed that addition of a small amountof nanoparticles (less than 1% by volume) to conventional heattransfer liquids increase the thermal conductivity of the fluid upto approximately two times. Pak and cho [6] and Li and Xuan [7]provided the first empirical correlation for calculating the Nusseltnumbers in laminar and turbulent flows inside a tube using wateras a base fluid. Eastman et al. [11] found that with less than 1% vol-ume of CuO nanoparticles, the convective heat transfer coefficientof water increases more than 1%. Wen and Ding [1] showed that

the local heat transfer coefficient varies with u and Re at the tubeentrance region of laminar flow. Lai et al. [12] studied Al2O3

(20 nm)-deionized water nanofluids subjected to constant wallheat flux at low Reynolds number (Re < 270). They observed 8%enhancement in the Nusselt number for 1% volume concentrationof nanoparticles at Re = 270.

Chandrasekar et al. [13] have found alumina/water nanofluidshave higher convective heat transfer coefficient with inserted wirecoil in tube. They did experiments with plain tube and two wirecoils with pitch ratio 2 and 3 inserted in a tube under fully devel-oped laminar flow and their results indicated the Nusselt numberincreased 12.24%, 15.91% and 21.53% respectively at Re = 2275.Heris et al. [3] investigated convective heat transfer of CuO andAl2O3–water nanofluids under laminar flow conditions through acircular copper tube. In their study, the heat transfer coefficientwas found to increase with increasing particle volume fraction aswell as Peclet number. Al2O3–water nanofluids showed higherenhancement of the heat transfer coefficient compared withCuO–water nanofluids. Arani and Amani [14] studied heat transfercoefficient and pressure drop of TiO2–water nanofluid in a horizon-tal double tube counter-flow heat exchanger experimentally. Theirexperiments were carried out in 0.01 and 0.02 volume concentra-tions with 10, 20, 30 and 50 nm particle diameters under turbulent

Nomenclature

A cross-sectional area of the tube (m2)CP specific heat capacity (J kg�1 K�1)D tube diameter (m)d diameter (m)f friction factorh(exp) experimental convective heat transfer coefficient of

nanofluid (W m�2 K�1)k thermal conductivity (W m�1 K�1)L tube length (m)Bc Boltzman constant (1.3807 � 10�23 J/K)l mean free path (m)Nu(exp) experimental Nusselt number of nanofluidPr Prandtl numberRe Reynolds numberT temperature (K)um average fluid velocity (m s�1)

Greek symbolsl dynamic viscosity (Pa s)u nanoparticle volume fractionq density (kg m�3)DP pressure drop (Pa)DTlm logarithmic mean temperature difference (K)

Subscriptsave averagebf base fluidnf nanofluidp particlein inletout outlet

Fig. 1. Nanofluids sample after 3 days.

112 M.S. Mojarrad et al. / Experimental Thermal and Fluid Science 53 (2014) 111–118

flow regime. They define thermal performance factor with bothNusselt number and pressure drop and showed that 20 nm particlediameter has the highest one. Heyhat et al. [15] performed exper-imental study on alumina/water nanofluids in a horizontal tubeunder laminar regime. They measured nanofluids properties exper-imentally and showed the convection heat transfer coefficient of2% nanofluids was 32% higher than the base fluid in fully developedregion.

In many industrial application the mixture of water and ethyleneglycol with different volume or weight percentages are used as cool-ant. Thus some researches were done for convective heat transferenhancement of nanofluids with mixture of water and ethylene gly-col as base fluid. Kulkarni et al. [16] investigated convective heattransfer enhancement of silicon dioxide–EG/water (60–40% byweight) experimentally. They studied the effect of Reynolds num-ber, volume fraction and three different particle sizes. Their resultsshowed heat transfer coefficient and pressure loss increased withincrease in Reynolds number, volume fraction and particle size.

Zamzamian et al. [17] carried out experiments on convectiveheat transfer of Al2O3 and CuO ethylene glycol nanofluids in doublepipe and plate heat exchangers under turbulent flow. They foundconvective heat transfer coefficient growth with increasing particleloading and nanofluid temperature. They reported augmentation ofconvective heat transfer coefficient from 2% to 50% more than basefluid. Yu et al. [18] investigated rheological and heat transfer prop-erties of alumina–water–ethylene glycol (55–45 by volume) exper-imentally. Their results indicated the thermal conductivity of 2%and 3% nanofluids are 7.7% and 11.6% much more than base fluid.Also the heat transfer coefficient of 1% and 2% nanofluids are 57%and 106% higher than base fluid when Re = 2000.

Raveshi et al. [19], the nucleate boiling heat transfer ofalumina–water–ethylene glycol (50–50 by volume) was studiedexperimentally. Six different volume concentrations of nanofluidswere used and results showed the heat transfer coefficient haveoptimum value in 0.75% volume concentration (64% higher thanbase fluid).

Main objective of the present study is to find out the effects ofbase fluid, nanoparticles volume concentrations and Reynoldsnumber on the average convective heat transfer coefficient andpressure drop along a circular tube under constant surface temper-ature boundary condition. The experiments were carried out forlaminar flow regime. It should be noted that the flow was onlyhydrodynamically fully developed. The obtained results for theNusselt number of nanofluids were compared with the availableequation for thermal length problem.

2. Nanofluid preparation

Nanoparticles used in this work are spherical shape Al2O3 pow-der with a size range of 20–30 nm and purity of +99%, supplied byDEGUSSA, Germany. A two-step method was used to prepare sta-ble nanofluids. Distilled water and the mixture of distilled waterand ethylene glycol by equal volume concentration (WEG50) wereused as the base fluid. For stabilizing the nanoparticles, sodiumdodecyl benzene sulfonate (SDBS) was used as the dispersant withthe amount of one tenth of the mass of nanoparticles. An electronicmass balance with accuracy 0.1 mg (model: Vibra LF 224R) wasused to weight suitable amounts of Al2O3 powder and SDBS disper-sant. The dispersant was first added to the base fluid and the mix-ture was sonicated in ultrasonic bath (model: Elmasonic S80 H) for2 h. Then, nanoparticles were dispersed into the mixture and son-icated for 4 h. Next, the mixture was agitated with magnetic stirrer(model: Stuart SB 162) for 5 h. Nanofluids were found very stableafter 3 days. The nanofluids were made with 0%, 0.25%, 0.5% and0.7% volume concentrations in each base fluid. Fig. 1 shows the alu-mina/WEG50 nanofluid samples for 0.25%, 0.5% and 0.7% volumefractions 3 days after preparation.

3. Nanofluid properties

Determination of nanofluids thermophysical properties is veryimportant as they are so impressive on the final results. Many

M.S. Mojarrad et al. / Experimental Thermal and Fluid Science 53 (2014) 111–118 113

previous works use different suggested equations to calculate ther-mophysical properties of nanofluids [2,3,20,21]. Heyhat et al. [15]showed that the density and the specific heat capacity of nanofl-uids are much closed to suggested equations but the viscosityand the thermal conductivity of nanofluids that have important ef-fect on hydrodynamic and thermal behavior, should be measuredexperimentally. In this paper, the viscosity and the thermal con-ductivity are measured experimentally and the density and thespecific heat capacity calculated from previous equations. Theequations used in this work are as follows. The subscripts bf, nfand p refer to the base fluid, the nanofluids and the particles,respectively.

3.1. Density

The density of nanofluids was calculated by Pak and Cho corre-lation [22]:

qnf ¼ ð1�uÞqbf þuqp ð1Þ

3.2. Specific heat

Prior researchers have used the following equation [23]:

Cpnf¼ð1�uÞqbf Cpbf

þuqpCpp

qnfð2Þ

3.3. Thermal conductivity

The thermal conductivity of nanofluids is measured using a KDPro 2 thermal properties analyzer (model: Decagon devices Inc.,USA). Fig. 2 shows that the thermal conductivity enhancement ofnanofluids with respect to its base fluid is between 2.1% and7.8%. The experimental results for 0.25% alumina/water nanofluidscompared with Heyhat et al. [15] and Chandrasekar et al. [24] cor-relations. It is clear that these correlations predict underestimatevalue for the thermal conductivity. Also the thermal conductivityaugments with increasing volume fraction and temperature ofnanofluids. The following correlation has been derived by curve fit-ting the experimental thermal conductivity data:

knf

kbf¼ 9:9� 10�3 þ expð3:8� 10�4T þ 4:56� 10�2uÞ ð3Þ

where T(�C) is the nanofluids temperature (20 �C 6 T 6 50 �C) and uis the nanofluids volume fraction (u 6 0.7%).

Fig. 2. Nanofluids thermal conductivity at different volume fraction andtemperature.

This equation predicts the experimental thermal conductivitywithin 5% and �4% errors.

3.4. Dynamic viscosity

The dynamic viscosity of nanofluids is measured as a function oftemperature and volume concentration using a viscometer (model:Brookfield DV-II + Pro). All the measurements were done in steadystate condition and the accuracy was about ±2%.

Fig. 3 shows that the measured data in various nanofluids vol-ume concentrations and temperature:

Comparison the obtained results for 0.25% alumina/water nano-fluids with Heyhat et al. [15] and Chandrasekar et al. [24] correla-tions shows that the calculated values by these correlations aresmaller than the experimental results. It can be seen that viscosityof nanofluids increases with increasing volume concentrations.Also temperature has little effect on viscosity ratio of nanofluids.Similar to the thermal conductivity, with respect to the experimen-tal data, the following correlation is proposed to calculate the vis-cosity of the nanofluids:

lnf

lbf¼ �9:59� 10�2 þ expð2:84� 10�3T þ 4:58� 10�1uÞ ð4Þ

where T(�C) is the nanofluids temperature (20 �C 6 T 6 50 �C) and uis the nanofluids volume fraction (u 6 0.7%). The maximum errorsof this equation are ±5%.

Thermophysical properties of WEG50 were taken from ASHRAEhandbook (fundamental SI) to evaluate properties of nanofluidswith WEG50 base fluid [25]. The thermophysical properties ofAl2O3 nanoparticles are taken as:

qp ¼ 3600 kg=m3; Cpp¼ 733 J=kgK; kp ¼ 36 W=mK

4. Experimental setup

Schematic scheme of the experimental setup that was used inthis paper is shown in Fig. 4.

The experimental apparatus consists of a pump, test section,cooling section, a reservoir, a flow meter, a pressure drop unit,pre heater, controller, two heaters, mixer and some thermocouples.A 5 L capacity tank was used as reservoir. To circulate the nanofl-uids through the system, a magnetic driven centrifugal type pump(model: CSE MP-010A/B) was used. The test section is a 1.46 mlong annular tube. The inner tube was made from copper with12.4 mm inner diameter and 0.5 mm thickness. To obtain a

Fig. 3. Nanofluids viscosity at different volume fraction and temperature.

Fig. 4. Schematic of experimental setup.

114 M.S. Mojarrad et al. / Experimental Thermal and Fluid Science 53 (2014) 111–118

hydrodynamic fully developed laminar flow, a 2 m tube was usedas a calming section before the test section to eliminate the en-trance effect. A stainless steel tube with 16 cm inner diameterwas used as the outer tube which is filled with constant tempera-ture water. To keep the water temperature constant, two heatersand 4 K-type thermocouples were located inside the outer tube.These thermocouples were in different location with differentdepth in the water bath. The heaters were turned on and off byan on–off controller with respect to the mean temperature of thefour thermocouples. Fig. 5 shows the locations of thermocouples(T11, T12, T13 and T14) and heaters in outer tube.

It should be noted that water inside the outer tube has no flowand only it is used to provide a uniform temperature conditionaround the copper tube. The range of wall temperature in all exper-iments is between 73 �C and 75 �C.

Nanofluids flowed inside the copper tube. The test section wasinsulated by fiber glass. Ten K-type thermocouples were mountedon the copper tube at different positions to measure the wall tem-perature as shown in Fig. 4. The inlet and outlet bulk temperatureof nanofluids were measured by 2 same type thermocouples. Thepressure drop along the test section was measured by an electronicdifferential pressure transmitter (model: ABB Kent-Taylor 505T)with an accuracy of ±3 Pa. The cooling section is a heat exchangerthat reduced the nanofluids outlet temperature significantly. Then,a pre heater with a controller was used to control the inlet

Fig. 5. location of thermocouple

temperature of nanofluids. The range of nanofluids inlet tempera-ture in all the conducted experiments is between 17 �C and 19 �C.All thermocouples, heaters and electronic differential pressuretransmitter were connected to a data acquisition system (model:ADAM 5000/TCP) to record experimental data. The flow rate wascontrolled by a three way valve and measured by a rotameter(model: AZM-LZB-35S) with an accuracy of 0.1 L per minute. Thisrotameter is calibrated for pure water and it must be recalibratedfor different fluids with various densities. The following equationis used to calculate the flow rate for different fluids:

Q2 ¼ Q 1

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiq1ðqf � q2Þq2ðqf � q1Þ

sð5Þ

where Q2 and Q1 are the volumetric flow rate of new fluid andwater, q2 and q1 are the density of new fluid and water and qf isthe float density.

5. Data analysis

5.1. Data reduction

By applying energy balance for a constant surface temperaturetube, the nanofluids average convective heat transfer coefficientand Nusselt number are calculated as follow:

s and heaters in outer tube.

M.S. Mojarrad et al. / Experimental Thermal and Fluid Science 53 (2014) 111–118 115

havðexpÞ ¼Cpnf� qnf � um � A � ðTout � TinÞ

p � D � L � DTlmð6Þ

DTlm ¼ðTw � TinÞ � ðTw � ToutÞ

ln½ðTw � TinÞ=ðTw � ToutÞ�ð7Þ

NuavðexpÞ ¼ havðexpÞ � Dknf

ð8Þ

where Tin and Tout are the measured bulk temperature at inlet andoutlet, um is the nanofluids average velocity, D is the tube diameterand DTlm is the logarithmic mean temperature difference. It shouldbe noted that the wall temperature is the average of data recordedfrom ten mounted thermocouple on the tube wall.

In order to determine theoretical pressure drop, Darcy–Weisbachequation was utilized:

DPnf ¼ fnf �LD�qnf u2

m

2ð9Þ

where L is the tube length, qnf is the nanofluids density and fnf isthe nanofluids friction factor which for laminar regime(Renf < 2300) calculated from:

fnf ¼64

Renfð10Þ

where nanofluids Reynolds number is:

Renf ¼qnf umD

lnfð11Þ

It should be noted that the thermophysical properties in theabove equations to calculate Nuav and DPnf must be evaluated atthe averaged values of the inlet and outlet bulk temperature. Allthe experiments were conducted at six different Reynolds numbers(650, 870, 1200, 1700, 2000 and 2300). For different nanofluidsconcentrations, the velocity of the flow was adjusted with respectto the nanofluids thermophysical properties changes in such a wayto keep the Reynolds number constant. All comparisons of the ob-tained results were done at constant Reynolds number to have abetter understanding for nanofluids thermo-fluids behavior.

5.2. Uncertainty

The uncertainty of each measured physical parameters is de-scribed in Table 1:

The experimental uncertainty can be determined using TSMuncertainty propagation [26] which are intended to yield a 95%level of confidence. The total uncertainty in the heat transfer coef-ficient and Reynolds number were mainly attributed to uncertain-ties in flow rate, wall and fluid temperature and correspondingoverall maximum were 3.2% and 2.2%, respectively. For each vol-ume concentration and each Reynolds number experiments wereconducted five times and the average of recorded temperatureswas used in the data analysis.

Table 1Typical uncertainty for measured variable.

Variable Uncertainty

Pipe diameter ±0.05 mmPipe length ±1 mmWall and bulk temperatures ±0.1 �CVolumetric flow ±0.1 l/sViscosity ±1.5% scalePressure drop ±0.2% scale

6. Result and discussion

6.1. Validation of experimental setup

In order to evaluate the reliability of the experimental setup,first a set of experiments were conducted with distilled water.The experimental Nusselt numbers were then compared to a corre-lation presented by Hausen [27] for the constant surface tempera-ture condition and thermal entry length problem where thevelocity profile is already fully developed. The correlation is asfollow:

Nuav ¼ 3:66þ 0:0668 � ðD=LÞ � Re � Pr

1þ 0:04½ðD=LÞ � Re � Pr�2=3 ð12Þ

Fig. 6 displays the comparison of the experimental data andNusselt number of distilled water as a function of Reynolds num-ber. As it is clear, there is a good accordance between experimentaland theoretical values and maximum deviation was 5%.

Also to evaluate the accuracy of pressure drop results, theoret-ical values from Eq. (9) were compared with experimental data fordistilled water. Fig. 7 indicates the comparison between experi-mental results and theoretical values of Darcy–Weisbach equation.It can be seen that experimental results have a good agreementwith theoretical values and the maximum deviation is 4% so theexperimental setup is successfully validated.

6.2. Nanofluids convective heat transfer

After validation, experiments were performed with four alu-mina/water volume concentrations (0%, 0.25%, 0.5% and 0.7%)and four alumina/WEG50 volume concentrations (0%, 0.25%, 0.5%and 0.7%) under laminar flow (650 < Re < 2300).

Fig. 8 shows the variation of average convective heat transfercoefficient for the base fluids and nanofluids in terms of Reynoldsnumber and volume fraction. Due to higher viscosity of WEG50base nanofluids, the velocity of it with respect to the water basenanofluids must be adjusted to keep the same Reynolds numbers.The major reason for having higher average convective heat trans-fer coefficient in WEG50 is higher Prandtl number in comparisonwith the water. The average convective heat transfer coefficientincreases with increasing Reynolds number as well as nanoparti-cles volume concentrations. As it is clear from Fig. 8, the averageconvective heat transfer coefficient of nanofluids are higher thanits own base fluid and increases with increasing Reynolds number.The maximum enhancement of 8%, 16% and 19% were observed for0.25%, 0.5% and 0.7% volume concentrations of alumina/water at

Fig. 6. Comparison of experimental and theoretical Nusselt number of distilledwater.

Fig. 7. Comparison of experimental and theoretical pressure drop of distilled water.Fig. 8. Average convective heat transfer coefficient of nanofluids with differentvolume concentrations versus Reynolds number.

Fig. 9. Average Nusselt number of nanofluids with different volume concentrationsversus Reynolds number.

Fig. 10. Average Nusselt number ratio of nanofluids with different volumeconcentrations versus Reynolds number.

116 M.S. Mojarrad et al. / Experimental Thermal and Fluid Science 53 (2014) 111–118

Re = 2300, respectively. Where these were 13%, 19% and 24% for0.25%, 0.5% and 0.7% alumina/WEG50 volume concentrations atsame Reynolds number. It is found that by adding nanoparticlesto the WEG50 base fluid causes more increment in the averageconvective heat transfer coefficient compared with water base alu-mina nanofluids at equal Reynolds number. It seems this highervelocity of WEG50 nanofluids increases chaotic movements ofnanoparticles in the base fluid and improves the energy exchangerate which caused the WEG50 base nanofluids to have higher con-vective heat transfer coefficient.

With respect to Eq. (6) by using nanofluids the density and tem-perature difference increase and the logarithmic mean tempera-ture difference and the specific heat capacity decrease whichcaused augmentation in convective heat transfer coefficient.

Adding nanoparticle to the base fluid increases the thermal con-ductivity and viscosity. It seems higher nanofluids thermal conduc-tivity against base fluid has substantial effect on augmentation ofaverage convective heat transfer coefficient. Also due to chaoticmotion of nanoparticle, the energy exchange rate betweennanoparticle and base fluid has improved and the temperature ofthe fluid adjacent to tube surface increases. Therefore, the fluidtemperature gradient near the tube wall increases which resultsan enhancement of convective heat transfer of nanofluids withrespect to base fluids. Other probably mechanism which canincreases the nanofluid thermal conductivity is particle migrationto near the wall. In other hand augmentation in the viscositycaused an increase in the boundary layer thickness which reducesthe convective heat transfer coefficient. But Brownian motion ofnanoparticles reduces boundary layer thickness. The obtained re-sults from Fig. 8 shows that the convective heat transfer coefficientincreases with adding nanoparticle which seems that the thermalconductivity increase effect overcome the viscosity growth effect.

Fig. 9 displays the experimental average Nusselt number of basefluids and nanofluids with 0.25% 0.5% and 0.7% nanoparticles vol-ume concentrations. Nanoparticles volume concentration and Rey-nolds number have the same influence on Nusselt number asconvective heat transfer coefficient but the amount of augmenta-tion is less than enhancement of convective heat transfer coeffi-cient. With respect to Eq. (8), this is because of the fact that boththermal conductivity and convective heat transfer coefficients af-fect Nusselt number.

Average Nusselt number ratio as a function of Reynolds numberis shown in Fig. 10. The maximum augmentations were 11%, 16%and 20% for 0.25%, 0.5% and 0.7% alumina/WEG50 volume concen-trations. As it was indicated, this ratio was not growth withincreasing Reynolds number for all nanofluids concentrations.These results have an accord with Kayhani et al. [28] and Mojarrad

et al. [29] results where it is opposite to Duangthongsuk and Won-gwises results [30].

Generally the Nusselt number is a function of Reynolds number,Prandtl number and nanofluids volume concentrations. Based oncurve fitting the experimental Nusselt number data, a new correla-tion has been derived to predict the nanofluids Nusselt number:

Nuav ¼ 0:1899 Re0:37Pr0:496u0:67 ð13Þ

M.S. Mojarrad et al. / Experimental Thermal and Fluid Science 53 (2014) 111–118 117

It should be noted that in this equation the Reynolds number isbetween 650 to 2300, the volume concentration is less than 0.7%and the Prandtl number is greater than 4. There is a maximumdeviation of 7% between the experimental results and predictedvalues using suggested correlation.

Fig. 13. Comparison of energy ratio for different nanofluids versus Reynoldsnumber.

6.3. Nanofluids pressure drop

Figs. 11 and 12 show the comparison of the measured and pre-dicted pressure drop of alumina/water and alumina/WEG50 nano-fluids along the tube respectively. The predicted pressure drop wascalculated by employment Eq. (9) using nanofluids properties. As itis seen experimental results have good agreement with predictedvalues (11% maximum deviation). Also pressure loss increases withincreasing particle concentrations and Reynolds number and thishappens due to larger viscosity of nanofluids. In a same way withthe Nusselt number, a new friction factor correlation as a functionof Reynolds number and nanofluids volume concentration isproposed as follow:

f ¼ 473:66Re�1:344u0:0586 ð14Þ

In this equation the Reynolds number is between 650 and 2300and the volume concentration is less than 0.7%. This equationpredicts the nanofluids friction factor within +11% and �8%.

Fig. 11. Comparison of experimental and predicted pressure drop by Eq. (9) foralumina/water nanofluids.

Fig. 12. Comparison of experimental and predicted pressure drop by Eq. (9) foralumina/WEG50 nanofluids.

6.4. Comparison nanofluids thermal and hydraulic behavior

In order to compare thermal and hydraulic behavior of nanofl-uids, the ratio of transferred thermal energy to required mechani-cal energy should be examined. This ratio defined as energy ratio:

ER ¼havnf

=havbf

� �DPnf =DPbf

� � ð15Þ

where h and DP are convective heat transfer coefficient andpressure drop respectively. Fig. 13 shows the energy ratio ofalumina/water and alumina/WEG50 nanofluids versus Reynoldsnumber. As it is clear, the energy ratio increases with increasingnanoparticle loading as Reynolds number. Also at same Reynoldsnumber, it enhances with increasing nanofluid concentrations.

7. Conclusions

An experimental study on the heat transfer performance andpressure drop of alumina/water and alumina/WEG50 nanofluidsflowing in the thermal entrance region of a circular tube was per-formed. Experiments were carried out for the laminar flow regimeunder constant surface temperature boundary condition. The con-clusions from the present study are as follows:

� The thermal conductivity and the dynamic viscosity ofnanofluids were measured experimentally in different tem-peratures. The results show the nanofluids have higherdynamic viscosity and thermal conductivity with respectto the base fluids. This increment augments with addingnanoparticle to the base fluid.

� The presence of alumina nanoparticles in both base fluidsincreases the convective heat transfer coefficient, Nusseltnumber and pressure drop of the nanofluids and this aug-mentation increases with Reynolds number as well as theparticle concentration.

� The convective heat transfer coefficient enhancementrange was between 5% and 24% for 0.25%, 0.5% and 0.7%nanofluids volume fractions (650 6 Re 6 2300). The maxi-mum enhancements were 13%, 19% and 24% for 0.25%,0.5% and 1% alumina/WEG50 volume concentrations atRe = 2300, respectively.

� Adding nanoparticles to WEG50 base fluid have more influ-ence on the heat transfer coefficient and Nusselt number.

� The measured pressure drop of nanofluids compared withpredicted values using Darcy–Weisbach equation withnanofluids properties. A maximum deviation between

118 M.S. Mojarrad et al. / Experimental Thermal and Fluid Science 53 (2014) 111–118

experimental and theoretical results was 11% and itshowed that this equation can be used for estimate nanofl-uids pressure drop.

� Adding nanoparticles at the base fluid improves energyratio at equal Reynolds number. So, nanofluids can be usedas working fluid at mechanical engineering application andit helps engineers to design more efficient heat exchanger.

References

[1] D. Wen, Y. Ding, Experimental investigation into convective heat transfer ofnanofluids at the entrance region under laminar flow conditions, Int. J. HeatMass Transfer 47 (2004) 5181–5188.

[2] S. Zeinali Heris, M. Nasr Esfahany, S.G. Etemad, Experimental investigation ofconvective heat transfer of Al2O3/water nanofluid in circular tube, Int. J. HeatFluid Flow 28 (2007) 203–210.

[3] S. Zeinali Heris, S.G. Etemad, M. Nasr Esfahany, Experimental investigation ofoxide nanofluids laminar flow convective heat transfer, Int. Commun. HeatMass Transfer 33 (2006) 529–535.

[4] U. Rea, T. McKrell, L. Hu, J. Buongiorno, Laminar convective heat transfer andviscous pressure loss of alumina–water and zirconia–water nanofluids, Int. J.Heat Mass Transfer 52 (2009) 2042–2048.

[5] K.S. Hwang, S.P. Jang, S.U.S. Choi, Flow and convective heat transfercharacteristics of water-based Al2O3 nanofluids in fully developed laminarflow regime, Int. J. Heat Mass Transfer 52 (2009) 193–199.

[6] B.C. Pak, Y.I. Cho, Hydrodynamic and heat transfer study of dispersed fluidswith submicron metallic oxide particles, Exp. Heat Transfer 11 (1998) 151–170.

[7] Y. Xuan, Q. Li, Investigation on convective heat transfer and flow features ofnanofluids, J. Heat Transfer 125 (2003) 151–155.

[8] W. Williams, J. Buongiorno, L.W. Hu, Experimental investigation of turbulentconvective heat transfer and pressure loss of alumina/water and zirconia/water nanoparticle colloids (nanofluids) in horizontal tubes, J. Heat Transfer130 (2008) 042412.

[9] W. Duangthongsuk, S. Wongwises, Heat transfer enhancement and pressuredrop characteristics of TiO2–water nanofluid in a double-tube counter flowheat exchanger, Int. J. Heat Mass Transfer 52 (2009) 2059–2067.

[10] S.U.S. Choi, J. Eastman, Enhancing thermal conductivity of fluids withnanoparticles, in: Developments and Applications of Non-Newtonian Flows,Argonne National Lab, IL (United States), New York, 1995, pp. 99–105.

[11] J. Eastman, U. Choi, S. Li, G. Soyez, L. Thompson, R. DiMelfi, Novel thermalproperties of nanostructured materials, Mater. Sci. Forum 312–314 (1999)629–634.

[12] W. Lai, B. Duculescu, P. Phelan, R. Prasher, Convective heat transfer withnanofluids in a single 1.02-mm tube, in: Proceedings of ASME InternationalMechanical Engineering Congress and Exposition, ASME, Chicago, USA, 2006.

[13] M. Chandrasekar, S. Suresh, A. Chandra Bose, Experimental studies on heattransfer and friction factor characteristics of Al2O3/water nanofluid in a

circular pipe under laminar flow with wire coil inserts, Exp. Therm Fluid Sci.34 (2010) 122–130.

[14] A. Abbasian Arani, J. Amani, Experimental investigation of diameter effect onheat transfer performance and pressure drop of TiO2–water nanofluid, Exp.Therm Fluid Sci. 44 (2013) 520–533.

[15] M. Heyhat, F. Kowsary, A. Rashidi, M. Momenpour, A. Amrollahi, Experimentalinvestigation of laminar convective heat transfer and pressure drop of water-based Al2O3 nanofluids in fully developed flow regime, Exp. Therm Fluid Sci.44 (2013) 483–489.

[16] D.P. Kulkarni, P.K. Namburu, H.E. Bargar, D.K. Das, Convective heat transfer andfluid dynamic characteristics of SiO2 ethylene glycol/water nanofluid, HeatTransfer Eng. 29 (2008) 1027–1035.

[17] A. Zamzamian, S.N. Oskouie, A. Doosthoseini, A. Joneidi, M. Pazouki,Experimental investigation of forced convective heat transfer coefficient innanofluids of Al2O3/EG and CuO/EG in a double pipe and plate heat exchangersunder turbulent flow, Exp. Thermal Fluid Sci. 35 (2011) 495–502.

[18] W. Yu, H. Xie, Y. Li, L. Chen, Q. Wang, Experimental investigation on the heattransfer properties of Al2O3 nanofluids using the mixture of ethylene glycoland water as base fluid, Powder Technol. 230 (2012) 14–19.

[19] M.R. Raveshi, A. Keshavarz, M.S. Mojarrad, S. Amiri, Experimental investigationof pool boiling heat transfer enhancement of alumina–water–ethylene glycolnanofluids, Exp. Therm Fluid Sci. 44 (2013) 805–814.

[20] S. Fotukian, M. Nasr Esfahany, Experimental study of turbulent convective heattransfer and pressure drop of dilute CuO/water nanofluid inside a circulartube, Int. Commun. Heat Mass Transfer 37 (2010) 214–219.

[21] S. Fotukian, M. Nasr Esfahany, Experimental investigation of turbulentconvective heat transfer of dilute c-Al2O3/water nanofluid inside a circulartube, Int. J. Heat Fluid Flow 31 (2010) 606–612.

[22] B.C. Pak, Y.I. Cho, Hydrodynamic and heat transfer study of dispersed fluidswith submicron metallic oxide particles, Exp. Heat Transfer An Int. J. 11 (1998)151–170.

[23] K. Khanafer, K. Vafai, M. Lightstone, Buoyancy-driven heat transferenhancement in a two-dimensional enclosure utilizing nanofluids, Int. J.Heat Mass Transfer 46 (2003) 3639–3653.

[24] M. Chandrasekar, S. Suresh, A. Chandra Bose, Experimental investigations andtheoretical determination of thermal conductivity and viscosity of Al2O3/waternanofluid, Exp. Thermal Fluid Sci. 34 (2010) 210–216.

[25] F. Handbook, ASHRAE, SI Edition, American Society of Heating, Refrigeratingand Air-Conditioning Engineers, Inc, Atlanta, GA, 2005.

[26] H.W. Coleman, W.G. Steele, Experimentation, validation, and uncertaintyanalysis for engineers, Wiley, 2009.

[27] H. Hausen, Darstellung des Warmeuberganges in Rohren durchverallgemeinerte Potenzbeziehungen, Z. VDI Beih. Verfahrenstech 4 (1943) 91.

[28] M. Kayhani, H. Soltanzadeh, M. Hayhat, M. Nazari, F. Kowsari, Experimentalstudy of convective heat transfer and pressure drop of TiO2/water nanofluid,Int. Commun. Heat Mass Transfer 39 (2012) 456–462.

[29] M.S. Mojarrad, A. Keshavarz, A. Shokouhi, Nanofluids thermal behavioranalysis using a new dispersion model along with single-phase, Heat MassTransfer 49 (2013) 1333–1433.

[30] W. Duangthongsuk, S. Wongwises, An experimental study on the heat transferperformance and pressure drop of TiO2–water nanofluids flowing under aturbulent flow regime, Int. J. Heat Mass Transfer 53 (2010) 334–344.