1

Simulation of Flow Aerodynamics and Heat Transfer in a Plate-Fin Radiator

Winit jaiboon1 Niwat Phoocharoen3 Seiji Okawa4 Jarruwat Chareonsuk2 InternationalCollege, King Mongkuts Institute of Technology Ladkrabang, Ladkrabang, Bangkok, Thailand 10520
2 1

Department of Mechanical Engineering, Faculty of Engineering, King Mongkuts Institute of Technology Ladkrabang, Ladkrabang, Bangkok, Thailand 10520 Computer Service Center, King Mongkuts Institute of Technology Ladkrabang, Ladkrabang, Bangkok, Thailand 10520

Department of Mechanical Engineering, Tokyo Institute of Technology, Japan E-mail: sss_me05@yahoo.com, kpniwat@kmitl.ac.th, kcjarruw@kmitl.ac.th, sokawa@mech.titech.ac.jp Tel. +66-890-128-470

ABSTRACT
The objective of this study is to validate the mathematical model and to explore the effects of geometric parameters on heat transfer and friction loss of plate-fin radiator in terms of Stanton Number and friction factor. The numerical technique known as finite volume method is used. A three-dimensional model is constructed with the smallest fluid cell size next to the wall being smaller than or equal to the fins thickness as suggested by the previous worker. Conjugate heat transfer by Patankar, 1960, RNG kepsilon, with enhanced wall treatment are used for the modeling of heat transfer between solid and fluid and the fluid domain, respectively. The model is used to perform three cases of simulations with different values of fin pitch for validation purpose. Investigation on interactions between geometrical shape and flow aerodynamics is carried out, whereby the characteristics of heat transfer and friction loss is explored. Keywords: Heat exchanger, Louver fin, Stanton number, Friction factor

INTRODUCTION
Louver fin geometry (Figure 1) is used widely for industrial heat exchangers, air conditioning as well as automotive radiator, heater, oil coolers, and condenser. The outstanding characteristic of complex flow associated with this configuration has a high heat transfer effect thus inevitably creates high pressure drop. Both performance aspects are higher than the conventional louver plate fin. It is essential to find the reliable method in identifying the heat transfer and pressure performance of louver fin. Computational Fluid Dynamic has been widely used recently and has been a successful tool for understanding the flow and heat phenomena in many heat transfer devices. The underlying behavior of air flow through louvered fin array as well as its associated pressure distribution reflects the amount of heat transfer and the overall pressure drop. These parameters can be obtained from the experiment in relation to the geometric parameters of louver and fin. A study on air flow through louver fin is essential for understanding heat transfer mechanism. In the past, emphasize was made on experimental method with wind tunnel involving expensive instrument and tools with great difficulties on accuracy, because of the complexity in shape of the flow. During the past two decades, a lot of researchers had taken extensive studies on this problem. R.L. Webb [2] had conducted experimental results in order to correlate the geometry of louvered fin to the heat transfer and friction coefficients, j and f
13th Annual Symposium on Computational Science and Engineering (ANSCSE 13)

2
factors. Chang and Wang et al. [3] generalized heat transfer correlation for louver fin geometry at 91 types of radiator configurations. With their correlation factors, 90% of samples can be explained. T.A. Cowell et al. [4] obtained heat transfer and pressure drop characteristics of flat tube and louvered plate fin surfaces. From his work, friction factor and Stanton number were evaluated at certain range of Reynolds number. Later T.A.Cowell et al. [5] proposed an empirical formula for heat transfer on various kinds of louver fins. Tafti and Zhang [6] studied geometry effects on flow transition in multileveled fin, its propagation, and characteristic of vortex shedding using numerical technique called finite volume method and compared with experimental results. A. Bhattacharya et al. [7] presented a numerical investigation of fluid flow and heat transfer over louver fins in compact heat exchanger. The converted results were satisfactorily compared with Stanton number and Friction factor given as a function of Reynolds number of T.A. Cowell et al. [4].

Figure 1. Flat-side tube and louver fin heat exchanger [3]. The objective of this project is to introduce a method associated with simulation technique as an alternative for cost reduction of new product design and development cycle. The numerical technique called finite volume method is used in this study to investigate the effects of geometric parameters on heat transfer and fiction factor of louvered fin radiator. Table 1. Dimension details of louver plate-fin geometry. Fin Louver Louver Tube Number pitch pitch angle Pitch Fp (mm) Lp (mm) (deg) Tp(mm) 1 2.02 1.4 25.5 11 2 3.25 1.4 25.5 11 3 1.65 1.4 25.5 11

Number of tube n 2 2 2

Hydraulic diameter dh(mm) 3.35 5.00 2.79

GEOMETRICAL PARAMETER OF LOUVER PLATE-FINS

The geometrical parameter for a plate-fin in this study is shown in Figure 2. All samples in this analysis have the tube width 16 mm, depth 2 mm and the copper louver fin thickness is 0.05 mm. The length of the fins in the airflow direction is 41.6 mm associating with doublerow tubes. Simulations are performed for different geometries with various fin pitch (Fp), louver pitch (Lp), tube pitch (Tp) and louver angle (). The values of these parameters are listed in Table 1. The scope of this paper is limited to 3 from 15 configurations of T.A. Cowell and A. Achaichia [4].

BOUNDRY CONDITION AND COMPUTATIONAL DETAILS

The three-dimensional computational domain and boundary condition is shown in Figure 3. One-half of the tube pitch is considered with symmetry plane at one side and constant wall temperature at the other. The top and bottom surfaces of the domain are defined as periodic boundaries. Previous literature A. Bhattacharya et al. [7] suggests that the three-dimensional model is to be constructed with the smallest fluid cell size next to the wall being smaller than or equal to the fins thickness. For simplicity of the computational mesh and modeling reason of heat transfer between solid and fluid interface, however, its thickness next to fin is
13th Annual Symposium on Computational Science and Engineering (ANSCSE 13)

3
assigned to be equal to the fin heig with symm o ght metrical geo ometry at upp and lowe surface per er of the fin s surface. The numerical s simulations are performe for differe Reynolds number a ed ent s within the range of 100 to 4000. Fr 0 rom experim ment of T.A. Cowell and A. Achaichia [4] the temperature variation o the tube s of surface is sm mall, thus can be assume constant at 358 K. n ed a The inlet a temperatu of the lou air ure uver fin is se at 288 K. A thermo-p et All physical prop perties of the tube an fin surface are also as nd es ssumed const tant.

Figure 2 Geometric parameter of the 2. cal r louver plate- fin geometry (m red n mm).

nsional comp putational Figure 3. Three-dimen dom and bou main undary cond dition.

The SIMPL algorithm is used to c LE m couple fluid pressure and velocities. T discretiz p d The zations of the momen ntum, turbule ence kinetic energy, turb bulence dissip pation rate, a energy equations and e are set to second-orde upwind. T under-r er The relaxation fa actors for ea iteration are, for ach n pressure = 0.3, moment tum = 0.7, tu urbulence kin netic energy = 0.8, turbu ulence dissipa ation rate = 0.8 and e energy = 1. T residuals for converg solution of the conti The s ged inuity, compo onents of velocities, turbulent kin netic energy and turbule dissipatio rate are b y ent on below 105, while for w energy it is below 107 similarly to A Bhattacha s A. arya [7].

CALCU ULATION P POCEDUR OF PER RE RFORMAN NCE PARA AMETER

The efficie ency of louve ered plate-fin can be considered in terms of frictio factor and Stanton n on d number de epending on interaction between geometrical shape and flow aerod n ns dynamics. Stanton nu umber is based on the am mount of he transfer w eat which is rela ated to the total heat t transfer are and the v ea velocity thro ough the min nimum flow area. The log-mean-tem mperature difference can be calcu ulated by the coolant an area weig e nd ghted average air temper ratures of inlet and o outlet planes. The conven . ntion from Kays and London [8] is adopted. The friction K factor is de efined by an equivalent shear force in the flow d n i direction per unit of hea transfer r at area. Diffe erent terms o Reynolds numbers ar defined b of s re based on hydraulic diam meter and louver pitch h
Re d =

ULp 4A L Ud h d ; Re Lp = ; d h= c s A

(1)

Where Ls is the length of the fins in the airflow direction, A is the total heat transfer area, Ac n w r is the mini imum flow a area, and U is the veloc through the minimum flow area Stanton city m a. number and Friction fac can be d d ctor defined by
St =
f =

hc Uc p
p U 2 A 2 Ac

(2) (3)

and hc is th heat trans coefficie he sfer ent, P this is the total pressure dr s l rop. The hea transfer at coefficient can be defin terms of the log-mean ned n-temperatur difference re e
th 13t Annual Sympo osium on Computational Science and Engineering (ANSCSE 13) g

4
hc = Q A LMTD

(4)

LMTD =

To Ti ln(To / Ti )

(5)

Ti is the temperature difference between the coolant and the air at the inlet plane, Ti = Tw Ta,i . To is the temperature difference between the coolant and the air at the outlet plane, To = Tw Ta,o . It is assumed that there is small difference in coolant and tube temperature thus Tw is equal to the temperature of tube wall. Q is the heat transfer rate. LMTD is the log mean temperature difference. The heat transfer rate is given by
Q = UAc (Ta , o Ta, i )

(6)

Therefore the heat transfer coefficient and the Stanton number can be re-written as
hc = Uc p Ac (Ta ,o Ta ,i ) A LMTD (Ta ,o Ta ,i ) A St = c A LMTD

(7) (8)

RESULTS AND DISCUSSIONS

Flow aerodynamic and temperature distribution Figure 4 shows the results of velocity and temperature contours for two different Reynolds numbers; (a) and (b) at low Reynolds number (Red = 100), and for (c) and (d) at high Reynolds number (Red = 1000). At low Reynolds number, the boundary layers thickness around the lover is very thick as the air flows through the gap between the fins rather than flowing into the gap between the louvers. Generally, the temperature difference between the fin surface and the air is not much. The result of heat transfer performance of the fin is poor. At high Reynolds numbers the air flow is directed by the louvers, the boundary layers thickness around the lover is very thin. In this case temperature of the air increases along the flow direction and a significant temperature difference is observed between air and fin. The result of heat transfer is thus higher at high Reynolds number. Heat transfer and pressure drop characteristics Heat transfer and friction loss of plate-fin radiator are expressed in terms of Stanton number (St) and friction factor (f ). Figures 8 and 9 show the result from this study compared with empirical results of Stanton number and friction factor as functions of Reynolds number for configuration number 1 and 3 of Table 1. The values from the experiment and from the formulation of A. Achaichia and T.A. Cowell [4] and those computed from Stanton number and friction factors from A. Bhattacharya et al. [7] are plotted on the same graph. The predictive quality of the selected model is good in general with the minimum and maximum deviation in heat transfer of 0.3% and 12% respectively. As far as friction factor is concerned, the minimum and maximum deviation is 0.2% and 35% respectively. The heat transfer characteristics of louvered plate fin are discussed in relation with flow aerodynamic and the distribution of pressure and temperature.

13th Annual Symposium on Computational Science and Engineering (ANSCSE 13)

5 Velocity Temperature

(a) Velocity contours (m/s) Red 100.

(b) Temperature contours (K) Red = 100.

Velocity Temperature

(c) Velocity contours (m/s) Red = 1000.

(d) Temperature contours (K) Red = 1000. Figure 4. Velocity contour and Temperature contour, (a) and (b) at low Reynolds number Red=100, and; b) and d) at high Reynolds number Red=1000.
1
St-expt St-comp St-std f-expt f-comp f-std

Stanton Number

Friction factor

0.1

0.1

0.01 100 1000 Reynolds Number

0.01 100 1000 Reynolds Number

Figure 8. Stanton number and Friction factor of configuration 1.

Figure 9. Stanton number and Friction factor of configuration 3.

13th Annual Symposium on Computational Science and Engineering (ANSCSE 13)

Friction factor

St-expt St-comp St-std f-expt f-comp f-std

6
500 450 400 350 300 250 200 150 100 50 0 0
Lp=1.4mm Fp=1.65mm Lp=1.4mm Fp=2.02mm Lp=1.4mm Fp=3.25mm

Heat tranfer coc. (W/m2K)

Pressure drop(Pa)

Lp=1.4mm Fp=1.65mm Lp=1.4mm Fp=2.02mm Lp=1.4mm Fp=3.25mm

1200 1000 800 600 400 200 0 0

1000 2000 3000 4000 Reynolds Number

1000 2000 3000 4000 Reynolds number

Figure 10. Effects of Fp on heat transfer coefficients.

Figure 11. Effects of Fp on pressure drop.

The heat transfer coefficients and the values of pressure drop are given for different values of Reynolds numbers in Figure 10 and 11. The effects of fin pitch (Fp = 1.65mm, 2.02mm, 3.25mm) can be observed. Heat transfer coefficients proportionally relate with Reynolds number. Moreover, the heat transfer coefficients and pressure drop increase with decreasing in fin pitch. This is congruent with the previous result carried out by Junnqi Dong et al. [11] in Heat transfer and pressure drop correlation for the multi-louvered fin compact heat exchangers.

CONCLUSION
The effects of geometric parameters have been investigated by considering heat transfer and friction loss of louver plate fins in terms of Stanton number and friction factor, similar to A. Bhattacharya et al. [7]. The study is performed within the range of Reynolds number from 100 to 4000. Comparison study with empirical equation of A. Achaichia and T.A. Cowell [4] is made to evaluate the predictive quality of the technique within the specified range of operating conditions and geometries. Good agreement is observed as far as Stanton number is concerned but large deviation is found friction factors. The predictive quality of the selected model is good in general with the deviation in heat transfer of 0.3% up to 12%. As far as friction factor is concerned, the deviation is 0.2% up to 35%. The heat transfer characteristic of louvered plate fin is discussed in relation with flow aerodynamic and the distribution of pressure and temperature. The heat transfer coefficient and pressure drop increase with decreasing in fin pitch. Work is on progress to investigate the underlying mechanism of flow aerodynamic on heat transfer and pressure drop as it is very important for design of radiators and heat exchangers.

13th Annual Symposium on Computational Science and Engineering (ANSCSE 13)

7
LITERATURE CITED
[1] S.V. Patankar, Numerical Heat Transfer and Fluid Flow, Hemisphere/McGraw-Hill, Washington, DC.1998, p41-77. [2] R. L. Webb, Principle of Enhanced Heat Transfer, John Wiley & Son, 1993. [3] Yu-Juei Chang and Chi-Chuan Wang, A Generalized Heat Transfer Correlation for Louver Fin Geometry, Int. J. Heat Mass Transfer, Vol.40, No.3, 1997, p533- 544. [4] T.A. Cowell and A. Achaichia, Heat Transfer and Pressure Drop Characteristics of Flat Tube and Louver Plate FinSurfaces, Experimental Thermal and Fluid Science, No.1, 1988, p147-157. [5] T.A. Cowell, M.R.Heikal and A.A chaichia, Flow And Heat transfer in Compact Louvered Fin Surfaces, Experimental Thermal and Fluid Science, No.10, 1995, p192199. [6] D.K.Tafti and X.Zhang, Geometry Effects on Flow Transition in Multilouvered Fins Onset, propagation, and Characteristic Frequencies, International Journal of Heat Mass Transfer, Vol. 44, No.22, 2001, p4195-4210. [7] V.P. Malapure, Sushanta K. Mitra , A.Bhattacharya, Numerical Investigation of Fluid Flow and Heat Transfer Over Louvered Fins in Compact Heat Exchanger, International Journal of Thermal Sciences, No. 46, 2007, p 199211. [8] W.M. Kays, A.L. London, Compact Heat Exchangers, third ed., McGraw-Hill, New York, 1984. [9] A.A. Antoniou, M.R. Heikal, T.A. Cowell, Measurements of local velocity and turbulence levels in arrays of louvered plate fins, Proceedings of the Ninth International Heat Transfer Conference, Jerusalem, 1990, pp.105-110. [10] H.K. Versteeg, W. Malalsekara, An Introduction To Computational Fluid Dynamics The Finite Volume Method, Prentice-Hall, New York, 1995. [11] j. Dong, J. Chen, Z. Chen,W. Zhang, Y. Zhou, Heat transfer and pressure drop correlateion for the mulit-louvered fin compact heat exchangers, Energy Conversion and Management, No.48, 2007, p 1506-1515. [12] FLUENT, Users Guide, Release 6.3, Fluent Inc., Lebanon, New Hampshire.

ACKNOWLEDGMENTS
The authors acknowledge the financial support of CH. Watanayont Co., Ltd

13th Annual Symposium on Computational Science and Engineering (ANSCSE 13)