Experimental Investigation of Heat Transfer by Natural

Available online at www.sciencedirect.com Available online at www.sciencedirect.com

ScienceDirect ScienceDirect

Available Availableonline onlineatatwww.sciencedirect.com www.sciencedirect.com Procedia Manufacturing 00 (2017) 000–000 Procedia Manufacturing 00 (2017) 000–000

ScienceDirect ScienceDirect

www.elsevier.com/locate/procedia www.elsevier.com/locate/procedia

Procedia Manufacturing 20 (2018) 311–317 Procedia Manufacturing 00 (2017) 000–000 www.elsevier.com/locate/procedia

2nd International Conference on Materials Manufacturing and Design Engineering 2nd International Conference on Materials Manufacturing and Design Engineering Experimental Investigation of Heat Transfer by Natural Convection with Perforated Pin Fin Array Experimental Investigation of Heat Transfer by Natural Convection with Perforated Pin Fin Array a*

b

c

d*

Shitole Pankaja*, Society BhosleInternational Santoshb, Kulkarni Kishor , Joshi Sarang c MESIC 2017,d*28-30 June Manufacturing Engineering Conference 2017, Shitole Pankaj , Bhosle Santosh , Kulkarni Kishor , Joshi Sarang a 2017, Vigo (Pontevedra), Spain JSPM’s BSIOTR, Pune, 412207, India b

G SaMandal’s MIT Aurangabad, 431001, JSPM’s BSIOTR, Pune, 412207, IndiaIndia G S Mandal’s MIT Aurangabad, 431001, India d c JSPM’s ICOER, Pune, 412207, IndiaIndia G S Mandal’s MIT Aurangabad, 431001, d JSPM’s ICOER, Pune, 412207, India

bc

Costing models for capacity optimization in Industry 4.0: Trade-off between used capacity and operational efficiency

Abstract Abstract a b heat transfer enhancement of The main objective of thisA. experimental to quantifya,*and compare theb,natural convection Santanastudy , P.isAfonso , A. Zanin R. Wernke The main of various this experimental study is to quantify comparediameter, the natural convection heat transfer enhancement of perforated finobjective array with fin spacing, perforation angle,and perforation pitch of perforation and heater inputs. The a perforated finthis array with various fin spacing, perforation angle, perforation diameter, pitch of perforation and heater University Minho, 4800-058 Guimarães, Portugal variables for natural convection cooling with ofthe help of finned surfaces are orientation and geometry. In thisinputs. study, The the b 89809-000 Chapecó, SC, Brazil variables forheat this transfer natural from convection cooling with the help of arrays finned surfaces are orientation geometry. In this study, the steady state the solid fin Unochapecó, and perforated fin are measured. The presentand study establishes optimized fin steady state heat transfer fromofthe fin and perforated are measured. study establishes optimized fin setup for various parameters finsolid geometry and its effectfin on arrays heat transfer results. The The present results obtained are matched well and setup forsimilar varioustrend parameters of fin geometry andfor itsheat effect on heat transfer results. The results areitmatched well that and showed and satisfactory agreement transfer under natural convection. Fromobtained all results is concluded o results it is concluded that showed satisfactory agreement for heat transferpitch under natural From45all the heat similar transfertrend rate and for the fins of perforation with constant and 4 mmconvection. diameter with geometry of perforation is Abstract the heat transfer ratearray for the finsfinofwith perforation with constant pitch and 4 mm rectangular diameter with 45o geometry of perforation is optimum fin and the of this 10 mm spacing is best suited horizontal fin array. optimumthe fin concept and the array of this fin with 10 production mm spacing isprocesses best suitedwill horizontal rectangular fin increasingly array. Under of "Industry 4.0", be pushed to be interconnected, © 2017 The Authors. Published bybasis Elsevier B.V. information based on a real time and, necessarily, much more efficient. In this context, capacity optimization © 2018 The Authors. Published by Elsevier B.V. © 2017 The Authors. Published byofcapacity Elsevier B.V. committee Peer-review under responsibility the scientific of the 2nd also International Conference on Materials goes beyond the traditional aim of maximization, contributing for organization’s profitability andand value. Peer-review under responsibility of the scientific committee of the 2nd International Conference on Materials Manufacturing Peer-review the scientific committee of the 2nd International Conference on Materials Manufacturing andresponsibility Design Engineering. Design Engineering. Indeed, lean under management and of continuous improvement approaches suggest capacity optimization instead of

Manufacturing and Design Engineering. maximization. The study of capacity optimization and costing models is an important research topic that deserves Keywords: Experiment; Convection; Array; Temperaure contributions from the practical and theoretical perspectives. This paper presents and discusses a mathematical Keywords: Experiment;both Convection; Array; Temperaure model for capacity management based on different costing models (ABC and TDABC). A generic model has been developed and it was used to analyze idle capacity and to design strategies towards the maximization of organization’s 1. Introduction value. The trade-off capacity maximization vs operational efficiency is highlighted and it is shown that capacity 1. Introduction optimization might hide operational Convection is the mode of heat inefficiency. transfer between a surface and a fluid moving over it. The energy transfer in © 2017 The Authors. Published by Elsevier Convection is the mode of heat a surface a fluid though movingthe over it. The conduction energy transfer in convection is predominantly due to transfer theB.V. bulkbetween motion of the fluidand particles, molecular within Peer-review under responsibility of the scientific of the Engineering Society International Conference convection is predominantly the extent. bulkcommittee motion theManufacturing fluid particles, though the molecular conduction within the fluid itself also contributesdue to to some If this of motion is mainly due to the density variations associated with 2017. the fluid itself also contributes some to the variations associated temperature gradient within theto fluid, theextent. mode If ofthis heatmotion transferisismainly said todue be due to density free or natural convection. Onwith the temperature gradient within the fluid, the mode of heat transfer is said tovelocity be due to free(like or natural convection. On the other hand fluid motion is principally produced by some superimposed field a fan, blower or a pump), Keywords: Cost Models; ABC; TDABC; Capacity Management; Idle Capacity; Operational Efficiency other hand fluid motion is principally produced by some superimposed velocity field (like a fan, blower or a pump),

1. Introduction

* Corresponding author. Tel.: +91 9011066232 E-mail address: [email protected] *The Corresponding author. Tel.: +91 cost of idle capacity is 9011066232 a fundamental information for companies and their management of extreme importance E-mail address: [email protected] in modern production systems. In general, it is defined as unused capacity or production potential and can be measured 2351-9789 © 2017 The Authors. Published by Elsevier B.V. in several ways: tons of production, available hours of manufacturing, etc. The management of the idle capacity Peer-review underThe responsibility of theby scientific of the 2nd International Conference on Materials Manufacturing and 2351-9789 © 2017 Authors. Published Elsevier committee B.V. * Paulo Afonso. Tel.: +351 253 510 761; fax: +351 253 604 741 Design Engineering. Peer-review under responsibility of the scientific committee of the 2nd International Conference on Materials Manufacturing and E-mail address: [email protected] Design Engineering. 2351-9789 © 2017 The Authors. Published by Elsevier B.V. Peer-review under of the scientificbycommittee the Manufacturing Engineering Society International Conference 2017. 2351-9789 © 2018responsibility The Authors. Published Elsevier of B.V. Peer-review under responsibility of the scientific committee of the 2nd International Conference on Materials Manufacturing and Design Engineering. 10.1016/j.promfg.2018.02.046

312 2

Shitole Pankaj et al. / Procedia Manufacturing 20 (2018) 311–317 Shitole Pankaj/ Procedia Manufacturing 00 (2017) 000–000

the energy transport is said to be due to forced convection [1-3]. Mainly there are two ways to increase the rate of heat transfer either by increasing convection heat transfer coefficient or by increasing the surface area. Increasing heat transfer coefficient may require the installation of a pump or fan, or replacing the existing one with a larger one, but this approach may or may not be practical. Besides, it may not be adequate [3]. The alternative is to increase the surface area by attaching to the surface extended surfaces called fins made of highly conductive materials such as aluminium. Such finned surfaces are commonly used in practice to enhance heat transfer, and they often increase the rate of heat transfer from a surface several fold [4.5]. 2. Literature Review Heat conduction is the mode of heat transfer accomplished via two mechanisms, by molecular interaction where energy exchange takes place by kinetic motion or by direct impact of molecule, or by the drift of free electrons as in case of metallic solids. Convection is the mode of heat transfer between a surface and a fluid moving over it [3, 6]. A large number of studies have been conducted on shape modifications by cutting some material from fins to make holes, cavities, slots, grooves or channels through the fin body to increase the heat transfer area and/or the heat transfer coefficient. Bassam and Abu [1, 2] conducted the numerical analysis and found that the heat transfer through permeable fins resulted in significant enhancement over solid fins. They stated that increase of number of permeable fins always resulted in increase in Nusselt number unlike in solid fins. They used certain assumptions to make the analysis simple that the fins are made up of highly conducting material. They did not validate their results with experimental work. Ridouane and Campo [3] in their study showed the enhancement in heat transfer using grooved channels. They found that the grooves enhance the local heat transfer relative to flat passage. Jamin and Mohamad [4] quantified and compared the steady state heat transfer from a heated vertical pipe with and without porous medium. They found that the largest increase in Nusselt number was achieved by high thermal conductivity solid carbon foam sleeve, which was about 2.5 times greater than a bare copper pipe. Ahn et al. [5] in their experiment compared the heat transfer rates with rounded and elongated holes in rectangular plate. They showed that elongated holes enhance heat transfer rate more than rounded holes but at the cost of pressure drop. Layeghi [6] in his numerical analysis also showed that heat transfer can be enhanced using porous media, but there is an increasing in pressure drop. Abdullatif Ben-Nakhi et al [7] studied the natural convection in open cavity. They found that the heat transfer rate increases with the thin fins attached to the hot surface. Zhnegguo et al., [8] used threedimensional petal shaped finned tubes to enhance the heat transfer. Povel and Mohamad [9] in their experimental and numerical study, investigated the effect of metallic porous material, inserted in a pipe, on rate of heat transfer. Effects of porosity, porous material diameter, thermal conductivity as well as Reynolds number on heat transfer rate and pressure drop were investigated. Awasarmol et al. [11, 12, 13, 14] studied the effect of permeability of fins on natural and forced convection heat transfer. On the basis of temperature profile they experimentally and numerically found out that the permeable fins perform better than the solid fins. The objective of present study is to experimentally quantify and compare natural convection heat transfer enhancement in perforated aluminum fin array with various perforation configurations, different perforation diameters, different heat inputs and at different angles of inclination, consequently to check the suitability of perforated fins (as opposed to solid fins) for industrial applications as far as the heat transfer enhancement is concerned. The study investigates the steady state heat transfer from the vertical plate with solid and perforated fins in natural convection environment and compares heat transfer performance results of perforated fin array to solid fin array. 3. Experimental Setup An experimental setup is designed and developed to carry out the investigation on Horizontal Rectangular Fin Array under mixed convection. The length and height of fin flats is kept same as in case of natural convection. Insulating siporex block was also used to reduce the leakage of heat from bottom and sides of fin array. Thus side and bottom heat losses from the siporex block were measured an accounted for. Experiments are performed with natural and mixed convection to get better idea about the relative performance of same fin array. The main objective


313 3

of the experiments is to find out optimum spacing zone and to make use of more area effect for lower spacing zone under assisting mode of natural convection. 3.1. Requirement of Experimental Setup For the effective experiment main requirements are siporex concrete block, fin array, cartridge heater, temperature sensors, control panel and data logger. 3.2. Siporex concrete block Siporex concrete block to account for the conduction and radiation losses so as to calculate convective heat transfer accurately. Concrete block is used to minimize heat loss from bottom side of fin array assembly.

Fig.1. a) Siporex Box, b) Siporex Block with Perforated fin array 3.3. Fin array Fin array consist of Fin plates, Spacers and Bakelite plates. The basic dimensions of the fin array used for the experimentation are L=200mm, W=100mm,H=40mm.Fin array are formed by assembling fins, spacers and joint together by tie bolts. Fin plates are separated by spacers. Fin array assembly is mounted inside the rectangular cavity of concrete block. Fin array assembly can change by changing spacers. Spacers are used of thickness 2mm, 3mm, 4mm and 5mm.Three holes are drilled for inserting 20mm diameter cartridge heaters and two holes for inserting tie bolts. Bakelite plates are used to avoid heat loss from end fins .Bakelite plates having dimensions of L=200mm, W=75mm and 8mm thickness.

Fig. 2. a) Plane Fin, b) Perforated Fin As per the problem definition, we have seen that the central portion of the fin remains ineffective; hence we remove the material from central portion through perforations. Holes of different diameters like 4mm, 8mm are made with drilling operation.

Fig.3. Fin Array 3.4. Cartridge heater Cartridge heaters are used to achieve certain temperature with specification of 100 watt. 3.5. Temperature sensors Thermocouples are used for temperature sensors and all thermocouple are made up of same spool of wire. Sixteen copper thermocouple are used in total. All thermocouples are separately calibrated. Calibrated thermocouples with temperature indicator are used to measure temperatures at various locations of fin array. Thermocouples are located on fin plates, channel base, Bakelite plate and concrete block.

314 4


3.6. Control Panel Calibrated digital voltmeter and ammeter are used to measure energy input of DC fan. A calibrated wattmeter is connected to measure heater input. 3.7. Data logger or temperature scanner

Fig. 4 Experimental Setup A 16 channel Data logger is used record the instantaneous pictures of flow pattern by temperature mapping for various fin spacing, flow velocities and heater input. The software and hardware of the instrument is so designed that it accepts nearly all types of instrumentation standard electrical signals, thermocouples and resistance. As per required input, one has to calibrate the analogy multiplexing card. In analog multiplexing card analog switches are used to select channel. 3.8. Experimental procedure Procedure of experimentation for convection: 1. The temperature on steady state readings at any point does not vary with time 2. The fin material is homogeneous with constant thermal conductivity. 3. The fin array is assumed as isothermal. 4. The coefficient of heat transfer is constant and uniform over the entire surface of the heat sink. 5. The temperature of surrounding fluid is uniform. 6. The average temperature of array after achieving steady state is used for calculating experimental results. 7. The average temperature of array after achieving steady state is used for calculating experimental results. 8. There are no heat sources within the fin itself. 9. The heat flow from the fin is proportional to the temperature difference i.e. excess temperature. 10. The fin array is assembled with spacers and fin flats by using tie bolts and nuts and placing the thermocouple at the appropriate locations. 11. Three cartridge heaters are inserted in the three drilled holes. These heaters are connected in parallel in the electrical circuit. 12. Assembled array is blackened with camphor and placed in siporex insulating block. 13. All the temperature of thermocouples T₁ to T7 are measured after steady state condition. 14. Predetermined heater input is given and kept constant by adjusting the dimmer stat, which is provided with stabilized voltage input. 15. The temperatures of assembled fin array at different locations and ambient temperature are recorded at the time intervals of one minute up to steady state condition using data logger. 16. Generally it took 4 hours to attain steady state condition. Steady state is attained to be reached when the difference in temperature recorded by thermocouple is negligible, less than 1°C in half an hour. 17. The temperatures above the array at different locations and ambient temperature are recorded at the time intervals of 3 seconds up to steady state condition using data logger. 18. The required heater input is adjusted with the help of dimmer stat. Wattmeter readings are recorded. Heater input is maintained at a constant value during the test run. 19. Thermocouple outputs are checked and recorded at all the measuring points.


315 5

20. The slowly increase in temperature values at every 15 minutes are recorded. When the thermocouples read more or less the same value for two or three successive observations, it is an indication of the steady state reading. A few sets are repeated to ensure the repeatability. 21. The temperatures of fin array assembly, Bakelite, and siporex block at different locations and ambient temperature are recorded at the time intervals of 3 seconds up to steady state condition using data logger. Test fin array blocks of solid and perforated fins were placed in enclosures. Tests were conducted over a range of conditions (Heater input: 15-35 W). For each case investigated, all the configurations of the perforated fin arrays and solid fin arrays were tested. The power supply to the heater was switched on, delivering approximately 15W to the heater. The apparatus was left running for approximately 3 hours and allowed to reach thermal steady state. Under steady state conditions, the temperatures at base and tip of each solid and perforated fin were recorded. The power supply to the heater was increased by 5W and the apparatus was allowed to reach the steady state. This procedure was repeated until a heater input of 35W was reached. The readings of temperatures were also taken by changing the angle of orientation of fins i.e. (00, 300, 450, 600 and 900). 4. Result and Conclusion Fig 9 shows temperature distribution along the fin array. It reveals that heating of fin array is uniform throughout the section which leads to confirmation of experiments at mentioned surface points of fin array.

Fig. 5. Temperature distribution in fin set up

Fig. 6. Comparison between perforation geometry of 300 and 450 Fig 6 indicates that ha values are better in case of 45 0 geometry of perforation with 4mm diameter perforation with constant pitch for spacing of 10mm. Fig 7 deals with comparison between perforation geometry of 30 0 and 450, the constant pitch with 4mm and 8mm diameter of perforation. It indicates that ha values are better in case of 45 0 geometry of perforation with 4mm diameter perforation with constant pitch for spacing of 10mm. Fig 8 deals with comparison between perforation

316 6


geometry of 450, the constant pitch with 4mm and 8mm diameter of perforation. It indicates that ha values are better in case of 450 geometry of perforation with 4mm diameter perforation with constant pitch for spacing of 10mm

Fig. 7. ha vs spacing for 30º and 45º constant pitch dia 4mm and 8mm

Fig. 8. ha vs spacing for 45º constant pitch dia 4mm and 8mm The experimental study of perforated heated horizontal rectangular fin array has been carried out under natural convection for various fin spacing, perforation angle, perforation diameter, pitch of perforation and heater inputs. In summary, the present study establishes optimized fin setup for various parameters of fin geometry and its effect on heat transfer results. Based on experimental analysis under natural convection, following conclusions are drawn: As the fin spacing increases, the average heat transfer coefficient (ha) increases for the fin array, as expected. The smallest value of ha is 7.78 W/m2K at 6 mm spacing with constant pitch of 450 perforation and 4 mm diameter. The highest value of ha is 12.61 W/m2K at 10 mm spacing with constant pitch of 450 perforation and 4 mm diameter. It is observed that the average nusselt (Nua) increases gradually as the ratio of fin spacing to height (S/H) increases for fin array. The value of Nuais very small for S/H = 0.15 to 0.2, as compared to that of S/H = 0.25 to 0.3. The highest Nua value is obtained about 19.23 W/m2K for S/H = 0.3 at 73W heater input with constant pitch 450 angle of perforation, fin spacing 12 mm and diameter of perforation 4 mm. The lowest Nua value is obtained about 8.14 W/m2K for S/H = 0.15 at 52W heater input with constant pitch 450 angle of perforation, fin spacing 6 mm and diameter of perforation 4 mm. The highest Nub value is obtained about 147.81 W/m2K for S/H = 0.3 at 73W heater input with constant pitch 450 angle of perforation, fin spacing 12 mm and diameter of perforation 4 mm. The lowest Nub value is obtained about 88 W/m2K for S/H = 0.15 at 52W heater input with constant pitch 450 angle of perforation, fin spacing 10 mm and diameter of perforation 4 mm. It is observed that temperature difference is maximum for fin spacing 6 mm at all heater inputs, as there is choking of air flow due to smaller spacing. For constant pitch, 450 perforation angle with perforation diameter 4 mm maximum convection (Q convection = 64.31W) at heater input 73W is obtained. Minimum heat convection (Qconvection = 43.70W) is found at constant pitch, 450 perforation angle with perforation diameter 4 mm when heater input is 52W. Experimental study has been carried out and compared with ANSYS results. The results obtained are matched well and showed similar trend and satisfactory agreement for heat transfer under natural convection. From all results


317 7

it is concluded that the fins of constant pitch 4 mm perforation, 45o perforation angle is optimum fin. And the array of this fin with 10 mm spacing is best suited horizontal rectangular fin array. 5. Future Scope    

Using the calibrated optimized fin setup the experimental investigation could be expanded for forced and mixed convection with the appropriate arrangement. CFD analysis could be made on the optimized setup for studying the fluid flow pattern. This setup could be modified more till the imagination of an individual, like use of different type of fins and perforation patterns and many more. After getting all the results of the setup this could be used in electronic equipment’s for application purpose.

References [1] Bassam A/K Abu Hijleh, Natural convection heat transfer from a fin with high conductivity permeable fins, ASMEJ. Heat Transfer 125 (2003) 282-288. [2] Bassam A/K Abu Hijleh, Enhanced forced convection heat transfer from a fin with high conductivity permeable fins, ASME J. Heat Transfer 125 (2003) 804-811. [3] EI Hassan Ridouane, Antonio Campo, Heat Transfer Enhancement of air flowing Across Grooved Channels: Joint Effects of channel height and Grooved Depth, ASME J. Heat Transfer 130 (2008) 1-7. [4] Yorweart L. Jamin, A.A. Mohamad Natural Convection Heat Transfer Enhancement from Fin Using Porous Carbon Foam ASME J. Heat Transfer 130 (2008) 1-6. [5] Ahn H.S., Lee S.W., Lau S.C., Heat Transfer Enhancement For Turbulent Flow Through Blockages With Round and Elongated holes in a Rectangular Channel ASME J. Heat Transfer 120 (2007) 1611-1615. [6] Mohammad Layeghi, Numerical Analysis of Wooden Porous Media Effects on Heat Transfer From Staggered Tube Bundles ASME J. Heat Transfer 130 (2008) 1-6. [7] Abdullatif Ben-Nakhi, M.M. Eftekhari,D.I. Loveday, Natural Convection Heat Transfer in a Partially Open Square Cavity With Thin Fin Attached to the Hot Wall ASME J. Heat Transfer 130 (2008) 1-9. [8] Zhang Zhnegguo, Xu Tao, Fang Xiaoming, Experiental study on heat transfer enhancement of helically baffled heat exchanger combined with three dimensional finned tubes Applied Thermal Engineering 24 (2004) 2293-2300. [9] Bogdan I. Povel, Abdulmajeed A. Mohamad, An experimental and numerical study on heat transfer enhancement for gas heat exchangers fitted with porous media International J. Heat and Mass Transfer 47 (2004) 4939-4952. [10] Abdullah H. AlEssa1, Mohamad I. Al-Widyan, Enhancement of natural convection heat transfer from a fin by triangular perforation of bases parallel and toward its tip Appl. Math. Mech. -Engl. Ed., 29(8) (2008) 1033–1044. [11] Awasarmol U.V., Pise A.T., “Investigation of Enhancement of Natural Convection Heat Transfer from Engine Cylinder with Permeable Fins” International J. Mechanical Engineering and Technology, 1(1) 2010 238-247 [12] Awasarmol U.V., Pise A.T., , “Experimental Study of Effect of Angle of Inclination of Fins on Natural Convection Heat Transfer through permeable fins” International Conference on Thermal Energy and Environment (INCOTEE) 2011, Krishnankoil, India. [13] Awasarmol U.V., Pise A.T., Sandikar A.N., “Numerical Analysis Of Heat Transfer Enhancement Using Perforated Fins At Different Base Inclinations And Base Temperatures” 21st National & 10TH ISHMT-ASME Heat and Mass Transfer Conference December 27-30, 2011, IIT Madras, India. [14] Awasarmol U.V., Pise A.T., Bhosale S.Y., Desai A.D., “Experimental Investigation Of Forced Convection Heat Transfer Enhancement Using Permeable Fins” 21st National & 10TH ISHMT-ASME Heat and Mass Transfer Conference December 27-30, 2011, IIT Madras, India. [15] Abdullah H. AlEssa, Ayman M. Maqableh1 and Shatha Ammourah, Enhancement of natural convection heat transfer from a fin by rectangular perforations with aspect ratio of two, International J. Physical Sciences 4 (10) (2009) 540-547. [16] Shaeri M.R., Yaghoubi M., Jafarpur K., Heat transfer analysis of lateral perforated fin heat sinks, Applied Energy 86 (2009) 2019–2029. [17] Shaeri M.R., Yaghoubi M., Numerical analysis of turbulent convection heat transfer from an array of perforated fins, International J. Heat and Fluid Flow 30 (2009) 218–228. [18] Gulay Yakar and Rasim Karabacak, Effects of holes placed on perforated finned heat exchangers at different angles on the Nusselt and Reynolds numbers, Scientific Research and Essay 5 (2) (2010) 224-234. [19] S. Kiwan and O. Zeitoun, Natural convection in a horizontal cylindrical annulus using porous fins, International J. Numerical Methods for Heat & Fluid Flow, 18 (5) (2008) 618-634. [20] S. D. Suryawanshi and N K Sane, Natural Convection Heat Transfer from Horizontal Rectangular Inverted Notched Fin Arrays, ASME J. Heat Transfer 131 (2009) 082501-082506. [21] Rama Subba Reddy Gorla , A.Y. Bakier, Thermal analysis of natural convection and radiation in porous fins, International Communications in Heat and Mass Transfer 38 (2011) 638–645.