at SciVerse ScienceDirect

Applied Thermal Engineering 50 (2013) 1361e1368

Contents lists available

Applied Thermal Engineering

journal homepage: www.elsevier .com/locate/apthermeng

Experimental investigation on carbon nano tubes coated brassrectangular extended surfaces

Rajendran Senthilkumar a,b,*, Sethuramalingam Prabhu a, Marimuthu Cheralathan a

a School of Mechanical Engineering, SRM University, Chennai e 603203, Tamil Nadu, IndiabResearch Scholar, Indian Institute of Information Technology, Design & Manufacturing Kancheepuram, Chennai e 600127, Tamil Nadu, India

h i g h l i g h t s

* Corresponding author. Research Scholar, Mechanitute of Information Technology, Design & Manufacturi600127, Tamil Nadu, India. Tel.: þ91 (0)9952156288.

E-mail addresses: rajendiransenthilkumar@gmairediffmail.com (R. Senthilkumar).

1359-4311/$ e see front matter � 2012 Elsevier Ltd.http://dx.doi.org/10.1016/j.applthermaleng.2012.05.04

g r a p h i c a l a b s t r a c t

< Rectangular brass fins werepreferred for convective heat trans-fer process.

< The rectangular brass fins are coatedwith multi wall carbon nano tubesin EBPVD process with nanometerthickness.

< Temperature and heat transfer ratewere investigated for nanocoatedand non-coated fins by using Tagu-chi method.

< Multi wall carbon nanotubes act asa pin fin to enhance surface area foreffective convective heat transferrate.

a r t i c l e i n f o

Article history:Received 16 May 2011Accepted 30 May 2012Available online 15 June 2012

Keywords:Multi wall carbon nano tubeHeat transfer ratePhysical vapor deposition processNano coatingTaguchi methodRectangular brass fins

a b s t r a c t

Finned surface has been extensively used for free convection cooling of internal combustion engines andseveral electronic kits etc. Here rectangular brass fin was preferred for analysis. Thermocouples wereattached all over the surface of the fin in equal distances. The measurement of surface temperature andcalculated convective heat transfer rate were reported for several heat input values. The overall systemperformance can be improved by enhancing heat transfer rate of extended surfaces. Based on the aboverequirement, brass surface was coated by carbon nano tubes. The temperature and heat transfer char-acteristics were investigated using Taguchi method for experimental design. Finally the performances ofcoated and non-coated rectangular brass fins were compared. The average percentage of increase in heattransfer rate was proved around 12% for carbon nanocoated rectangular brass fins.

� 2012 Elsevier Ltd. All rights reserved.

1. Introduction

The operation of many engineering systems results in thegeneration of heat. This may cause serious overheating problems

cal Engineering, Indian Insti-ng Kancheepuram, Chennai e

l.com, Senthilkumar_mech@

All rights reserved.0

and sometimes leads to failure of the system. The heat generatedwithin a system must be dissipated to its surrounding in order tomaintain the system operating at its recommended workingtemperatures and functioning effectively and reliably. Extendedsurfaces are commonly used in many engineering applications toenhance heat transfer. A number of studies have been performed inorder to increase the heat transfer effectiveness and to reduce thedimensions and weight of heat exchangers. The necessity to reducethe volume and weight of heat exchanger has become more

R. Senthilkumar et al. / Applied Thermal Engineering 50 (2013) 1361e13681362

important in many engineering applications like electronicindustry, compact heat exchanger sector, power plants, etc. Effi-cient design of heat exchanger with fins can improve systemperformance considerably. Among several available techniques foraugmentation of heat transfer in heat exchanger tubes, the use ofinternal fin appears to be very promising method as evident fromthe results of the past investigations. This is especially important inmodern electronic systems, in which the packaging density ofcircuits is high. In order to overcome this problem, thermal systemswith effective emitters as fins are desirable [1]. In order to achievethe desired steady-state rate of heat dissipation, with the leastamount of material, the optimal combination of geometry andorientation of the finned surface is required [2].

Among the all geometrical variations, rectangular fins are themost commonly encountered because of their simple construction,cheaper cost and effective cooling capability. Two commonorientations of fin configurations, horizontally based vertical finsand vertically based vertical fins, have been widely used in theapplications. However, the horizontal orientation is not preferablebecause of its relatively poor ability to dissipate heat [3]. Compactheat exchangers have large surface-area-to-volume ratiosprimarily through the use of finned surfaces. An informativecollection of articles related to the development of compact heatexchanger has been presented by Shak et al. [4] A comprehensiveliterature on principles of enhanced heat transfer has been alsopresented by Webb [5]. Since the use of extended surfaces is oftenmore economical, convenient and trouble free, most proposedapplication of increasing surface area is adding fins to the surfacein order to achieve required rate of heat transfer. However, thedesigner should optimize the spacing or the number of fins onbase carefully; otherwise fin additions may cause the deteriorationof the rate of heat transfer. Although adding numerous finsincrease the surface area, they may resist the air flow and causeboundary layer interferences which affect the heat transferadversely [4]. The experimental investigations related to thethermal performances of rectangular fins were reported exten-sively in the literature [3e10]. However, except a few of them,studies were performed for limited ranges of fin configurations.Today’s designer has available a very wide range of materials fromwhich to choose. To determine the most cost-effective material forany application is no simple task when costs and performance areproperly assessed. Brass is ideally suited to the manufacture ofmany components because of the wide variety of forms and sizesavailable that minimize costs of machining to final dimensions. Ithas a unique combination of properties: strength, shock resis-tance, ductility and conductivity combined with good corrosionresistance and other attributes such as superb machinability.Other beneficial properties are good formability, good spark-resistance, low magnetic permeability and toughness retainedabove and below ambient temperatures. As for handling the heatsink problem, the size of its outward design, the amount of finflake, the gap of fin flake, the area of its outward surface all havean intimate relation on enhancing its convection effect andincreasing its heat sink ability [11,12]. For the operation of electricfan of heat sink, the experiment from Obinelo [13] and Wirtz et al.[14] show that changing the installation angle of fan’s blade andadjusting the distance between fan and heat sink can improve andenhance its heat sink capacity. From the above descriptive anal-ysis, the optimal design and selection of effective heat sink moduleis becoming one of the primary challenges of the computer scienceand technology industry. A numerical simulation of Parallel-PlainFin (PPF) heat sink module to understand the affecting situationof its related modeling parameters. Taguchi method for design ofexperiment (DOE) and the analysis of variance (ANOVA), which iscurrently most widely used on construction design and the

optimized analysis of production process [15], are employed toefficiently seek the combination of optimized design parameters.Using fins, is the most frequent application in heat transferenhancement for increasing the heat transfer area. Extendedsurface heat transfer is the study of high performance heat-transfer components with respect to the certain parameters suchas smaller weights, volumes, cost etc. In addition electric andelectronic circuitry, fin arrays are also found in airelandespacevehicles and their power sources, refrigeration, conventionalfurnaces and gas turbines, waste-heat boilers, nuclear-fuel models,and many more [16].

The only controllable variable to enhance the convection heattransfer rate is the geometry of the fins. The designer must opti-mize the size and the spacing of the fin arrays otherwise; usingfins can bring more disadvantages than its advantages to thedesign. Several studies of natural convection from rectangular finswere conducted previously [3,4,6e9,17e23]. Sadik Kacak et al. [24]proposed the nanofluids are considered to offer importantadvantages over conventional heat transfer fluids. During the2000s, researchers focused on measuring and modeling theeffective thermal conductivity and viscosity of nanofluids. Bilenet al. [25] proposed the effect of geometric position of wallmounted rectangular block on the heat transfer from the surface,taking into account the angular displacement of the block inaddition to its spanwise and streamwise disposition. The experi-ments were conducted with air in rectangular channel. Prabhuet al. [26] proposed the nano surface finish has become animportant parameter in the semiconductor, optical, electrical andmechanical industries. The materials used in these industries areclassified as difficult to machine materials such as ceramics,glasses and silicon wafers. Machining of these materials up tonano scale accuracy is a great challenge in the manufacturingindustry. The same group explored the machining characteristicsof AISI D2 tool steel with graphite as a tool electrode duringelectrical discharge machining process [27].

As a result of these studies, there are many data available in theliterature. However, rather limited data are available for an array ofcarbon nanocoated rectangular brass fins. CNTs are also attributedwith having excellent thermal and electrical conductivity in theaxial direction. Due to the unique structure of CNTs with the‘armchair’ morphology, superb electrical properties are believed toexist. Once heat has crossed the thermal interface, the heat sinkmaterial should be able to transport heat away from the source asmuch as possible. In this experimental study single rectangularbrass fin is mounted on a horizontal brass plate. The purpose of thisexperimental study is mainly dealing with heat transfer throughrectangular fin composed of brass base with carbon nano tubescoated fin. The rate of heat transfer from the fin effects of geometricparameters and base-to-ambient temperature difference on theheat transfer performance of the fin. In this study, the naturalconvection heat transfer protruding from brass base is investigatedexperimentally. This research finds temperature distributions andactual heat transfer rates for different heat inputs and corre-sponding fin efficiencies of the coated and non-coated rectangularbrass fins and also further investigation and optimization werecarried out by using Taguchi method and ANOVA analysis. The erroranalysis carried out by using regression model. The detailed anal-ysis is discussed further in the report.

2. Materials and methods

2.1. General materials

Natural convective heat transfer experimental setup was stan-dardized and calibrated. Over the hot surface the test specimenwas

Fig. 1. TEM image of MWNT’s 95 wt % <8 nm OD.

R. Senthilkumar et al. / Applied Thermal Engineering 50 (2013) 1361e1368 1363

located for heat transfer analysis. The test specimen was made ofbrass and rectangular in geometry (100 mm height, 63 mm widthand10mmthickness). The four thermocoupleswerefixed at 25mm,47 mm, 68 mm and 94 mm from base. The preferred heat inputvalues are 16.2 W, 49.5 W and 94.5 W. The corresponding temper-atures for different heat input values are measured at specifiedlocations.Deionizedwater (liquidmedium)andMWCNTswereusedin this work. Themulti wall carbon nano tubes spreads and suspendwith deionized water. Multi-walled nanotubes (MWNT) consist ofmultiple rolled layers (concentric tubes) of graphite. Carbon nano-tubes are a new form of carbon with unique electrical andmechanical properties. They can be considered as the result offolding graphite layers into carbon cylinders and may be composedof a several shells. The unique properties ofmultiwall nanotubes areproving to be a rich source of newphysics and could also lead to newapplications in materials and devices. The sources of carbon nanotubes are received fromCheap tubes Inc., USA [28]. The specificationof multi wall carbon nano tubes are given in Table 1.The nanotubeshadanaveragediameterof 10e20nmanda lengthof 10e30mm,andwere produced catalytically from hydrocarbon materials onnanocatalysts under high pressure. Multiwall carbon nano tubeswere coated on the test specimen using physical vapor deposition(PVD) processer [29]. Finally based on the temperature distribu-tion, the convective heat transfer rate and fin efficiency werecalculated and compared for coated and non-coated surfaces. Theexperimental work was carried out by using L9 orthogonal array ofTaguchi method for design of experiments. Fig. 1 shows a Trans-mission Electron Microscopy (TEM) image of the multiple wallcarbon nano tubes. It can be seen that the nanotubes are entangled,which makes it difficult to disperse and to stabilize them in water.

2.2. Experimental setup

The flat heater is connected to the Dimmer stat (voltage regu-lator) through the voltmeter and ammeter andpower is switched on(Fig. 2). The heaterwas used for converting the electrical energy intoheat energy. The amount of heat energy supplied by the heater wascalculated using voltmeter and ammeter readings. Initially thetemperatureof thefinwill increase, after some time the temperatureof the fin will not change with respect to time. This condition indi-cates that thefinhas attained the steady state. Steady state conditioncan be ensured by measuring the temperature sensed by the ther-mocouples that are attached to the surface of the fin at variouslocations. The rectangular brass fins were coated by multi wallcarbon nano tubes using physical vapor deposition (PVD) processwith nanometer thickness. The temperature readings and thedistance fromthebase atwhich the thermocoupleswere attached tothe surface of the nano coated and non-coated fins are observed.

2.3. Physical vapor deposition (PVD) of carbon nanotubes

The fin is cleaned thoroughly before coating and placed inside thecoating chamber. The pumping mechanism was started to createrequired vacuum levels inside the coating chamber (Fig. 3). Thechamberwill accommodate up to three ingots ranging in size from49

Table 1Specification of MWCNTs.

OD 10 nme20 nm

Length 10 mme30 mmPurity >95 wt%Ash <1.5 wt%Specific surface area 233 m2/gElectrical conductivity 10�2 S/cm

to 68 mm in diameter and 450mm long. The electron beam gunwasswitched on and the nitrogen gas supply was given to the chamber.The electron beamwas focused on the fin surface and the multi wallcarbon nanotube coating has been performed to the required thick-ness. After coating, the fins are going to be heated in a tray ata temperature of about 150 �C. Fins were heated continuously fora spanof 30min at the above set temperature. PVD is primarily a line-of-sight process, therefore uniformcoatings of complexparts (suchasturbineblades) canbeaccomplishedbycontinuously rotating thepartduring the coating process. The specimen were allowed to cool andtaken out from the coating chamber. The PVD process offers manydesirable characteristics suchas relativelyhighdeposition rates (up to150 urn/minute with an evaporation rate of app. 10e15 kg/h), densecoatings, controlled composition and microstructure, lowcontamination and high thermal efficiency. Coatings produced bythe PVD process usually have a good surface finish and a uniformmicrostructure. The microstructure and composition of the coatingcan be easily altered by manipulating the process parameters andingot compositions.

2.4. Fin analysis

2.4.1. Fin efficiencyA dimensionless parameter that describes howwell the short fin

end insulated functions as an extension of the base surface. The finefficiency (hf) can be calculated from the following equation.

hf ¼ Qfin=Qmax ¼ Qfin=hAðTb � TNÞ (1)

mLc ¼ffiffiffiffiffiffihPkA

rLc ¼

ffiffiffiffiffiffi2hkt

rLc (2)

Fig. 2. Schematic diagram of experimental setup.

Table 3L9 orthogonal array.

Exp. no Heat input (W) Fin distance (mm)

1 1 12 1 23 1 34 2 15 2 26 2 37 3 18 3 29 3 3

Fig. 3. Physical vapor deposition (PVD).

R. Senthilkumar et al. / Applied Thermal Engineering 50 (2013) 1361e13681364

3. Results and discussions

3.1. Taguchi method

The studies of process parameters which influence the objectivefunction and which parameters will impact largely to this convec-tive heat transfer is analyzed. The parameters control factors andtheir levels were identified in Table 2. Thus the design of experi-ments with optimization of control parameters to obtain bestresults is achieved in the Taguchi Method. “Orthogonal Arrays” (OA)provide a set of well balanced (minimum) experiments and Signal-to-Noise ratios (S/N), which are log functions of desired output,serve as objective functions for optimization, help in data analysisand prediction of optimum results. Three levels and two parame-ters were taken based on Taguchi design of experiments L9orthogonal array was taken in Table 3. Totally nine experimentswere conducted. The temperature with and without nanocoatedbrass rectangular fin and S/N ratio values are tabulated in Table 4.The temperature distribution over rectangular fin surface wasfound out by thermocouples attached at different locations on thesurface of fin for different power inputs. These values were used tofind other parameter such as convective heat transfer coefficient(h). The temperature distribution over the surface of fin was foundout by conducting the experiments and values were tabulated.

h ¼ Q=AðTs � TNÞ (3)

3.2. Signals to noise ratio

S/N ratio is calculated based on quality of the characteristics. Theobjective function of this method is to improve the heat dissipatingrate through heat transfer apparatus. If heat is transferred effec-tively, the corresponding surface temperature will be lower. If the

Table 2Control factors and levels.

Item Control factor Units Level 1 Level 2 Level 3

A Heat input (I) Watts 16.2 49.5 94.5B Fin distance (X) mm 25 47 68

calculated S/N ratio (smaller the best) is smaller that will give bestresults. Based on this concept, the following expression for S/N ratiowas selected.

S=N ratioðhÞ ¼ �10 log101= ni¼1

Xy2 (4)

The predicted S/N ratio (h�) using the optimal levels of the param-

eters can be calculated as:

h� ¼ hm þXpi¼1

hi � hm (5)

where hm ¼ total mean of S/N ratio, hi ¼ mean of S/N Ratio at theoptimum level and p ¼ the number of main parameters thatsignificantly affect the performance.

The predicted temperature for without nanocoating(T) ¼ 68.94 �C.

Actual temperature through experiment for without nano-coating (T) ¼ 68.80 �C.

Level 3 of A and Level 1 of B which gives the maximum effect ofimproving temperature distribution.

The predicted temperature for nanocoated fin (T) ¼ 58.05 �C.Actual temperature through experiment for nanocoated fin

(T) ¼ 57.90 �C.Interestingly, A3 and B1 is the best combination for heat input

94.5 W and fin distance of 25 mm will give the minimum surfacetemperature for coated and non-coated surfaces. From the factoreffect graph shows that heat input giving more impact to improvethe temperature distribution of fin (Figs. 4 and 5). The based onexperiments, the optimum level setting of parameters is A3B1.

In fin analysis, themain variables are heat input and fin distance,among these two parameters in optimization technique heat inputis ranked one for both without and with nano coating whichinfluences the heat transfer characteristics (Tables 5 and 6).

3.3. ANOVA analysis

The purpose of analysis of variance is to find the significantfactors affecting the heat transfer process, to improve the temper-ature distribution of heat transfer apparatus. ANOVA analysis givesclearly, how the parameters affect the response and the level ofsignificance of the factors. In this study, the two main parametersconsidered are heat input and fin distance. The ANOVA analysisclearly indicates that the R2 value (99.91%) for coated surface ishigher than non-coated surface (99.32%). The high R2 value indi-cates the better fitness of theoretical model with experimentalobservation. Here 0.000 p value of heat input is significant. Themain output of ANOVA analysis on variance arranged in Tables 7and 8. Larger FAo values (289.66 and 2329.05) indicate that thevariation of heat input makes the significant changes on heattransfer rather than fin distance. Interestingly, the ANOVA analysisconfirms the Taguchi results obtained.

Table 4Temperature and S/N ratio value for rectangular brass fins.

Exp. No Coded values Temperature withoutnanocoating (�C)

Temperature withnano coating (�C)

S/N ratio withoutnano coatingtemperature (h)

S/N ratio withnano coatingtemperature (h)

Heat input (I) Fin distance (X)

1 1 1 37.7 34.2 �31.5268 �30.68052 1 2 37.1 34.1 �31.3875 �30.65513 1 3 36.8 34 �31.317 �30.62964 2 1 59.2 53.5 �35.4464 �34.56715 2 2 57.2 50.2 �35.1479 �34.01416 2 3 56.2 48.7 �34.9947 �33.75067 3 1 68.8 57.9 �36.7518 �35.25368 3 2 67.5 56.6 �36.5861 �35.05639 3 3 66.4 55.3 �36.4434 �34.8545Mean (m) �34.4002 �33.2735

1 2 3 1 2 3

-37

-3

6

-

35

-34

-

33

-32

-

31

Levels of input

Mea

n of

SN

rat

io

Heat Input ( )

Fig. 4. Factor effect diagram for S/N ratio without nano coating for rectangular brass fintemperature.

R. Senthilkumar et al. / Applied Thermal Engineering 50 (2013) 1361e1368 1365

3.4. Confirmation test and error analysis

The experimental confirmation is the final step of experimentalprocess. The purpose of the experimental confirmation is to validateconclusions drawn during the analysis phase and experimentswereconducted by setting the process parameters at optimum level.

Regression model is determining the relationship betweenindependent variable with dependent variables. Here fin distanceand heat input are independent variables and temperature isa dependent variable.

1 2 3 1 2 3

-35

-34

-33

-32

-31

-30

Mea

n of

SN

rat

io

Levels of input

Fig. 5. Factor effect diagram for S/N ratio with nano coating for rectangular brass fintemperature.

In this model, the relationship between temperature (T), heatinput (X1) and fin distance (X2) at different locations (I, d, T) forcoated and non-coated were analyzed (Fig. 6).

The empirical model was developed based on relationship existin between temperatures and heat input (T & I), temperatures andfin distance (T & D) in natural heat convection process.

The empirical model is

T ¼ AðX1ÞaðX2Þb (6)

The above non linear equation is converted to linear form by

log T ¼ log Aþ a logðX1Þ þ b logðX2Þ (7)

Now the above equation can be written as

y� ¼ b0 þ b1x1 þ b2x2 (8)

where y�true value of dependent temperature output on a loga-

rithmic scale, x1, x2 are the logarithmic transformation of thedifferent input parameters, b0, b1, b2 are corresponding parametersto be estimated. Minitab 15 software has been used to estimate theparameters of the above first order model and shown in Table 9.a and b are coefficients determined by regression analysis.

The regression analysis of the experimental data yields the semiempirical model

TemperaturewithoutnanocoatingðTÞ¼25:8ðIÞ15:2ðXÞ�1:05 (9)

Temperature with nanocoatingðTÞ ¼ 27:2ðIÞ11:2ðXÞ�1:27 (10)

The error between experimental values with regression models arecalculated by using the following equation

Table 5S/N ratio effect for without coated rectangular brass fins.

Factor Level 1 Level 2 Level 3 Delta Rank

A-Heat input �31.41 �35.20 �36.59 5.18 1B-Fin distance �34.58 �34.37 �34.25 0.32 2

Table 6S/N ratio effect for with nano coated brass rectangular fins.

Factor Level 1 Level 2 Level 3 Delta Rank

A-Heat input �30.66 �34.11 �35.05 4.40 1B-Fin distance �33.50 �33.24 �33.08 0.42 2

Table 7ANOVA for the temperature with nanocoating.

Factors Degree offreedom (f)

Sum ofsquares (SSA)

Variance(VA)

FAo P Contribution(%)

A 2 1436.25 718.12 2329.05 0.000a 99.45B 2 6.74 3.37 10.93 0.024 0.47Error 4 1.23 0.31 0.08Total 8 1444.22 100

a Significant, S ¼ 0.555278; R-Sq ¼ 99.91%; R-Sq(adj) ¼ 99.83%.

Table 8ANOVA for the temperature without nanocoating.

Factors Degree offreedom (f)

Sum ofsquares (SSa)

Variance(Va)

FAo P Contribution(%)

A 2 818.78 409.39 289.66 0.000* 98.14B 2 9.81 4.90 3.47 0.134 1.17Error 4 5.65 1.41 0.69Total 8 834.24 100

* S ¼ 1.18884; R-Sq ¼ 99.32%; R-Sq(adj) ¼ 98.64%.

Fig. 6. Response of temperature in rectangular brass fin surface (T) with and without nanocoating to heat input (I) and fin distance (X). A. Temperature without nanocoating vs HeatInput (I), Fin Distance (X). B. Temperature with nanocoating vs Heat Input (I), Fin Distance (X).

R. Senthilkumar et al. / Applied Thermal Engineering 50 (2013) 1361e13681366

Errorð%Þ¼ððExperimentalvalue�predictedvalueÞ=ExperimentalvalueÞ*100 (11)

Results of regression analysis were compared with experiments inTable 9 for 9 check sets. The maximum test errors for regressionmodel of surface temperature are 4.05% for without nanocoatingand 3.14% for with nanocoating.

This method is suitable for estimating temperature in an accept-able error ranges. From the results, errors ofmeasurements occurs intemperaturewithnanocoating is less thanthewithoutnanocoatingofbrass rectangular fin. The theoretical temperatures obtained throughregression model were compared with experimentally observedtemperature for coated and non-coated brass fin (Fig. 7A and B). TheR2 value of regressionmodel for non-coated surface is 99.32% and forcoated surface is 99.91%. The high R2 value for both coated and non-coated brass fin indicate, the better fitness exist among regressionmodel with experimental observation.

Fig. 6A and B is compared, the slope of the generated surface ishigh for non-coated fin that indicates the level of heat intensity(temperature) on the surface is more. Similarly, for nano coated

Table 9Comparison of regression model with experimental measurements for heatconvection process.

Experiment no. Without nanocoatingtemperature (�C)

With nanocoatingtemperature (�C)

Experimentalmeasurements

Regressionmodel

Experimentalmeasurements

Regressionmodel

1 37.7 39.95 34.2 37.132 37.1 38.9 34.1 35.863 36.8 37.85 34 34.594 59.2 55.15 53.5 48.335 57.2 54.1 50.2 47.066 56.2 53.05 48.7 45.797 68.8 70.35 57.9 59.538 67.5 69.3 56.6 58.269 66.4 68.25 55.3 56.99

surface the slope of the surface is less that indicates the drop insurface temperature due to coated material.

3.5. Heat dissipation and temperature distribution

The comparison of experimental temperature distribution forcarbon nano coated and non-coated fins, the following things wereobserved (Fig. 8). The first one is surface temperature was reducedfor coated fin. The second one is temperature difference for coated

Fig. 7. A. Error showing actual vs predicted temperature without nanocoating. B. Errorshowing actual vs predicted temperature with nanocoating.

30

40

50

60

70

2.5 4.7 6.8 9.4

Tem

pera

ture

in °

C

Thermocouple Location in cm

Nano coated

Noncoated

Fig. 8. Temperature distributions of coated and non coated fin for different headinputs.

Fig. 9. Temperature distribution in different location for various heat inputs.

0

40

80

120

160

25 46 68 94

Non coated

Nano coated

Thermocouple location in cmCon

vect

ive

Hea

t T

rans

fer

Co

-Eff

icie

nt i

n W

/m2 K

Fig. 10. Comparison of convective heat transfer coefficient of nano coated and non-coated rectangular brass fins.

40

50

60

70

80

94.5 94.5

Fin

eff

icie

ncy

in p

erce

ntag

e

Uniform heat input in Watts

Non coated

Nano Coated

Fig. 11. Rectangular fin efficiency.

Table 10Experimental uncertainties.

Variables Uncertainty (%)

Non coated surface Coated surface

Voltage 0.100 0.100Current 0.720 0.720Mean temperature 0.200 0.250Density of the air 0.008 0.008Thermal conductivity of air 0.340 0.340

R. Senthilkumar et al. / Applied Thermal Engineering 50 (2013) 1361e1368 1367

and non-coated surface was very high at higher heat input value.Finally the carbon nanocoated rectangular brass fin was moreefficient for transferring the heat with higher heat transfer coeffi-cient (Fig. 9). The plot of heat transfer coefficient and thermocouplelocation at different heat inputs for coated and non-coated finsvisibly demonstrated, the level of heat intensity maintained onnon-coated surface was high (Fig. 10). Fig. 10 clearly shows the dropin temperature for different heat input values at different locationsfor nano coated rectangular brass fin.

Exactly over the coated surface the temperature was compa-rably too low because, the carbon nano tube acts as a pin fin. It wasclearly noted that there is a significance changes in convective heattransfer coefficient for coated and non-coated fins (Fig. 11). Theefficiency of rectangular brass coated and non-coated fins werecompared for same heat input value (94.5 W). The increase in finefficiency for coated fin was 14%. In literature, similar type of workcarried out by Pathaka et al. [30] for SMA wire cooling and theyachieved 24% increase in natural convection.

The individual contributions of measured physical propertieswith uncertainties are tabulated (Table 10).

4. Conclusion

The present research brings about the convective heat transfercharacteristics of carbon nano tube coated brass surface undernatural convection. Experimentally the temperature distributionfor coated and non-coated fins were observed and optimized usingTaguchi method and ANOVA analysis.

There is a significant drop in surface temperature for nanocoated brass surface for three different heat inputs. Simultaneously,the convective heat transfer increases for coated brass surface dueto considerable increase in surface area of carbon nano tube andhuge temperature difference between coated surface and ambienttemperature. Interestingly, the individual carbon nano tubes coatedon the brass surface almost acts as a pin fin. The averageimprovement in fin efficiency is 12% for coated brass surface. Theerror analysis was carried out and the percentage of errors istabulated.

Hereby it is confirmed that the coating is efficient in enhancingnatural convection while considering the thermal properties ofcarbon nano tubes. This research work focused into naturalconvection on nano coated surfaces and it indicates the furtherinvestigations to clearly understand the physical phenomenon andpercentage of influence of surface forces using suitable dimen-sionless numbers.

Acknowledgements

All experimental facilities in the department were created byfunding respectively from School of Mechanical Engineering, SRMUniversity, Chennai, Tamil Nadu, India. Prof. D. Kingsly Jeba Singh,Professor and Head, School of Mechanical Engineering, SRMUniversity, is gratefully acknowledged for his precious support. Theauthors acknowledged for constant support provided by SRM Nanotechnology centre for nano coating process.

R. Senthilkumar et al. / Applied Thermal Engineering 50 (2013) 1361e13681368

References

[1] Chiang Ko-Ta, Optimization of the design parameters of parallel-plain fin heatsink module cooling phenomenon based on the Taguchi method, Int. Com-mun. Heat Mass Transfer 32 (2005) 1193e2120.

[2] B. Moshfegh, R. Nyiredy, Comparing RANS Models for Flow and ThermalAnalysis of Pin Fin Heat Sinks, 15th Australasian Fluid Mechanics Conference,The University of Sydney, Sydney, Australia 13e17 December (2004).

[3] J.R. Welling, C.N. Wooldridge, Free convection heat transfer coefficients fromvertical fins, J. Heat Transfer 87 (1965) 439e444.

[4] F. Harahap, H.N. McManus, Natural convection heat transfer from horizontalrectangular fin arrays, J. Heat Transfer 89 (1967) 32e38.

[5] C.W. Leung, S.D. Probert, M.J. Shilston, Heat exchanger: optimal separation forvertical rectangular fins protruding from a vertical rectangular base, Appl.Energy 19 (1985) 77e85.

[6] H. Yü ncu€, A. Gü venc, An experimental investigation on performance ofrectangular fins on a vertical base in free convection heat transfer, Heat MassTransfer 37 (2001) 409e416.

[7] K.E. Starner, H.N. McManus, An experimental investigation of free convectionheat transfer from rectangular fin arrays, J. Heat Transfer 85 (1963) 273e278.

[8] C.D. Jones, L.F. Sparrow Smith, Optimum arrangement of rectangular fins onhorizontal surfaces for free convection heat transfer, J. Heat Transfer 92 (1970)6e10.

[9] H. Yu€ncu€, G. Anbar, An experimental investigation on performance of rect-angular fins on a horizontal base in free convection heat transfer, Heat MassTransfer 33 (1998) 507e514.

[10] H. Yü ncü, S. Yildiz, An experimental investigation on performance of annularfins on a horizontal cylinder in free convection heat transfer, Heat MassTransfer 40 (2004) 239e251.

[11] A.D. Kraus, A. Bar-Cohen, Design and Analysis of Heat Sinks, John Wiley andSons, Inc., New York, 1995.

[12] S.H. Kim, N.K. Ananda, Laminar developing flow and heat transfer betweena series of parallel plates with surface mounted discrete heat sources, Int. J.Heat Mass Transfer 37 (15) (1994) 2231.

[13] I.F. Obinelo, ASME Interpak’97-Advances in Electronic Packaging. Hawaii, June(1997).

[14] R.A. Wirtz, R. Sohal, H. Wang, Thermal performance of pin-fin fan-sinkassemblies, ASME J. Electr. Pack 119 (1997) 26.

[15] C. Taguchi, Introduction to Quality Engineering, Asian Productivity Organi-zation, Tokyo, 1990.

[16] D.Q. Kern, A.D. Kraus, Ext. Surface Heat Trans, McGraw-Hill, New York, 1972.[17] N.D. Fitzroy, Optimum spacing of fins cooled by free convection, J. Heat

Transfer (1971) 462e463.[18] E.M. Sparrow, P.A. Bahrami, Experiments on natural convection heat transfer on

thefinsof afinnedhorizontal tube, Int. J.HeatMassTransfer23 (1980)1555e1560.[19] E.M. Sparrow, I.E. Niethammer, Effect of vertical separation distance and

cylinder-to-cylinder temperature imbalance on natural convection for a pairof horizontal cylinders, J. Heat Transfer e Trans. ASME 103 (1981) 638e644.

[20] C.W. Leung, S.D. Probert, M.J. Shilston, Heat exchanger design: thermalperformances of rectangular fins protruding from vertical or horizontal rect-angular bases, Appl. Energy 20 (1985) 123e140.

[21] D.S. Kadle, E.M. Sparrow, Numerical and experimental study of turbulent heattransfer and fluid flow in longitudinal fin arrays, J. Heat Transfer 108 (1986)16e23.

[22] C.W. Leung, S.D. Probert, Thermal effectiveness of short protrusion rectan-gular, heat-exchanger fins, Appl. Energy 34 (1989) 1e8.

[23] C.B. Sobhan, S.P. Venkateshan, K.N. Seetharamu, Experimental studies onsteady free convection heat transfer from horizontal rectangular finarrays, Waerme und Stoffuebertragung 25 (1990) 345e352. SpringerVerlag.

[24] S. Kakac, A. Pramuanjaroenkij, Review of convective heat transfer enhance-ment with nanofluids, Int. J. Heat Mass Transfer 52 (2009) 3187e3196.

[25] K. Bilen, S. Yapici, C. Celik, A Taguchi approach for investigation of heattransfer from a surface equipped with rectangular blocks, Energy Convers.Manage. 42 (2001) 951e961.

[26] S. Prabhu, B.K. Vinayagam, Nano surface generation of grinding process usingcarbon nano tubes, Sadhana 35 (2010) 747e760.

[27] S. Prabhu, B.K. Vinayagam, AFM surface investigation of Inconel 825 withmulti wall carbon nano tube in electrical discharge machining process usingTaguchi analysis, Arch. Civil Mech. Eng. 9 (2011) 149e170.

[28] www.cheaptubes.com.[29] Jogender Singh, Farhat Quli, Douglas E. Wolfe, J.T. Schriempf, Jason Singh, An

overview: electron beam-physical vapor deposition technology-present andfuture.

[30] A. Pathaka, J.A. Buchonb, D. Breia, J. Shawc, J. Luntza, S. Jinb, Carbon Nanotube(CNT) fins for the enhanced cooling of shape memory alloy wire, in: Proc. ofSPIE 6926:6969292K91K-2 (2009).

Nomenclature

As: surface area, m2

h: convective heat transfer coefficient, W/m2KLc: characteristic length, mP: perimeter, mQ: heat transfer rate, Whfin: fin efficiencyn: kinematic viscosity, m2/sTs: surface temperature, �CTN: ambient temperature, �CQmax: maximum heat transfer rate, WT: A (X1)a (X2)b

T: temperature (�C)A: coefficientX1: heat input (W)X2: fin distance (Cm)n: no of experimentsy: no of response valueh�: predicted S/N ratio

h: S/N ratiohfin: fin efficiencyn: kinematic viscosity, m2/sTs/Tw: surface temperature, �CTb: base temperature, �CTN: ambient temperature, �Ck: thermal conductivity of brass, W/mKurn: uniform resource number