Experimental and Theoretical Study of Heat Transfer by Natural Convection of a Heat Sink Used for Cooling of Electronic Chip

International Journal of Engineering Inventions
e-ISSN: 2278-7461, p-ISSN: 2319-6491
Volume 2, Issue 2 (January 2013) PP: 01-09
www.ijeijournal.com P a g e | 1
Experimental and Theoretical Study of Heat Transfer by
Natural Convection of a Heat Sink Used for Cooling of
Electronic Chip
1
Sunil Hireholi, 2
K.S. Shashishekhar, 3
George. S. Milton
123
Dept. of Mechanical Engineering
12
S. I. T, 3
S. S. I. T
123
Tumkur, Karnataka, India
Abstract:- In the present work, heat transfer analysis of a commercially available heat sink is carried out. The
heat sink is used for cooling of electronic chip STK4141II and removes the heat generated by the chip through
natural convection. The work includes experimental investigations and theoretical modelling.
In experimental investigations electrical heating is used to supply heat to the heat sink and the temperature of
the heat sink is measured using RTD thermocouples attached to the heat sink. Theoretical temperature of heat
sink is predicted employing 2-D modelling of the heat transfer process based on fundamental heat transfer
principles. Theoretically predicted temperatures of the heat sink are compared with the measured temperatures.
Keywords:- heat sink, electronic cooling, RTD thermocouples.
I. INTRODUCTION
Advances in the field of electronics have resulted in significant increase in density integration, clock
rates and emerging trend of miniaturization of modern electronics. This resulted in dissipation of high heat flux
at chip level. In order to satisfy the junction temperature requirements in terms of performance and reliability,
improvements in cooling technologies are required. As a result thermal management is becoming important and
increasingly critical to the electronics industry. The task of maintaining acceptable junction temperature by
dissipating the heat from integrated circuit chips is significant challenge to thermal engineers. The electronic
cooling is viewed in three levels, which are non-separable. First, the maintenance of chip temperature at the
relatively low level despite the high local heat density. Second, the heat flux must be handled at system or
module level. Finally, the thermal management of the computer machine room, office space or
telecommunication enclosure [1].
In electronic systems, a heat sink is a passive component that cools a device by dissipating heat into the
surrounding . Heat sinks are used to cool electronic components such as high-power semiconductor devices, and
optoelectronic devices such as higher-power lasers and light emitting diodes (LEDs). In the present work we
carry out analysis of heat transfer by a commercially available heat sink used for chip level cooling.
II. LITERATURE SURVEY
The cooling aspects have been studied by many investigators. Among them Mobedi and Sunden [2]
investigated the steady state conjugate conduction-convection on vertically placed fin arrays with small heat
source inside. Thermal performance of a free standing fin Structure of copper heat sink has been reviewed by
Jan Bijanpourian [3]. In his work he had used CFD to evaluate and compare heat dissipation capabilities of
aluminium and copper heat sinks. Experimental Investigation of Pin Fin Heat Sink Effectiveness for forced
convection is reviewed by Massimiliano Rizzi et al [4]. One of the earliest studies about natural convection heat
transfer from fin arrays was conducted by Starner and McManus [5]. Four different fin array configurations with
three base types were investigated and heat transfer coefficients were calculated. Leung et al. [6] performed an
experimental study on heat transfer from vertically placed fin arrays produced from an aluminium alloy . It was
found that for different configurations the maximum heat transfer rate from the fin arrays was obtained at the fin
spacing value of 10 mm. Harahap et al. [7] performed experiments on miniaturized vertical rectangular fin
arrays in order to investigate the effect of miniaturizing on steady state rate of natural heat transfer. A numerical
analysis on natural convection heat transfer from horizontally placed rectangular shrouded fin arrays were
performed by Yalcin et al. [8]. Commercially available CFD package PHOENICS was used to solve the three
dimensional elliptic governing equations. Yüncü and Mobedi [9] investigated the three dimensional steady state
natural convection from horizontally placed longitudinally short rectangular fin arrays numerically. A finite
difference code in Cartesian coordinate system based on vorticity-vector potential approach was used to solve

Experimental and Theoretical Study of Heat Transfer by Natural Convection…
the problem. Kundu and Das [10] performed an analytical study to investigate performance and optimum
design analysis of four fin array types viz, longitudinal rectangular fin array, annular rectangular fin array,
longitudinal trapezoidal fin array and annular trapezoidal fin array under convective cooling conditions are
investigated. Another experimental study regarding the natural convection heat transfer from both vertically and
horizontally oriented fin arrays was done by Leung and Probert [11]. The effects of fin spacing and fin height
were investigated for a limited number of fin array configurations. Vollaro et al. [12] analysed natural
convection from rectangular and vertical finned plates numerically in order to optimize the fin configuration.
The maximum heat transfer rate from fin array was investigated for the optimum fins spacing as a function of
parameters such as dimensions, thermal conductivity, fins absorption coefficient and fluid thermo-physical
properties.
III. EXPERIMENTAL STUDIES
The front, top and side views of the heat sink analysed are shown in Fig. 1. It consists of a vertical base
and seven horizontal fins as per the dimensions shown in the figure in millimetres. In the experimental studies
electrical heating is employed to supply heat to the heat sink. The electrical heater has same nominal dimensions
as the electronic chip and occupies the same physical location on the heat sink as the chip. The thermocouples
used for experiment were calibrated.
A. Experiment to determine heat transfer coefficient with heating coil
The method used for determining the heat transfer coefficient is to give a known heat input generated
by electrical heating and measure the temperature attained by the heat sink. The temperature is measured using
the RTD-thermocouples at three locations on the model and average of all the three values, T1, T2, T3, is taken
for the calculation of overall heat transfer coefficient. It is made sure that steady state conditions are reached
before the temperature are taken and then the corresponding heat transfer coefficient is computed. The
temperatures are taken at regular intervals till steady state is reached. The values of the temperature after it has
reached steady state are used in calculating the corresponding heat transfer co-efficient for the given heat input.
This is done for varying values of heat inputs
Fig.1 Front, top and side views of the heat sink
.
Fig. 2 shows a photograph of the Experimental set-up used for measurement of temperature. The heat
sink is kept at a height from the surface of the table to enable free flow of air from the bottom of the sink.
Alternating current from mains is supplied to the heating coil attached to the heat sink through a dimmerstat
(auto-transformer) as seen in the photograph given in Fig. 2. Voltage across the heating coil is measured using a
voltmeter and the current is measured using an ammeter. as described above. Heat dissipated and the associated
heat transfer co-efficient are calculated using the following basic equations [13], [14].
Qinput = V I (1)
𝑇𝑎𝑣𝑔 =
𝑇1+𝑇2+𝑇3
3
(2)
h =
𝑄input
𝐴(𝑇𝑎𝑣𝑔 −𝑇𝑎𝑚𝑏 )
(3)
Where,
Qinput = Heat input by the heating
coil (Watts)

V = Voltmeter reading (Volts),
I = Ammeter reading (Amperes),
T1,T2,T3=Temperature readings of
Thermocouples (o
C)
Tavg = Average temperature of heat sink (o
C),
h = Heat transfer coefficient (W/m2
K) ,
A = Effective area of the heat sink available for
heat transfer,
Tamb= Ambient temperature (o
C)
Results are given in Table 1.
Fig.2 Experimental setup
Table 1 shows time in seconds and the corresponding temperature readings of the three thermocouples
T1, T2, T3, for different values of the heat inputs along with their voltage and current values. The average
temperature, Tavg is also given in the table.
Table 1 Time history of temperature showing how steady state is reached for heat sink with heating coil kept in
open air for 26 o
C ambient temperature.
Volt-
age
(Volts)
Current
(Amps)
Power
Input
(W)
Time
(s)
T1
(0
C)
T2
(0
C)
T3
(0
C)
Tavg
(o
C)
10 0.23 2.3 0 26 26 26 26
65 28 28 28 28
289 30 30 31 30.33
608 32 33 34 33
1034 33 34 35 34
1584 34 34 36 34.66
2184 34 35 36 35
13 0.31 4.03 0 32 32 32 32
122 33 34 34 33.67
269 35 35 35 35
401 36 38 38 37.33
723 38 40 40 39.37
1055 41 41 41 41
1655 41 41 42 41.33
16 0.38 6.08 0 36 36 36 36
11’’ 38 39 39 38.67
188 40 40 41 40.33
367 42 43 44 43
812 44 46 46 45.33
1412 44 47 47 46

19 0.44 8.36 0 32 32 32 32
84 36 37 37 36.67
350 39 40 40 39.67
619 44 46 46 45.33
1138 49 51 52 50.67
1738 49 53 53 51.67
Fig. 3 shows the plots of the time versus the average temperature, Tavg, , for different values of the heat
input. From this figure one can observe the temperature reaching steady state for each case of the heat input. The
final value of the average temperature Tavg , is plotted against the heat input in Fig. 4.
Fig. 3 Time history of the average temperature for different values of heat input
Fig. 4 Steady state values of average temperature of the heat sink plotted against the corresponding heat input
values.
The overall heat transfer co-efficient is computed for the above heat input values from equation (3).
The values of the heat transfer co-efficient are presented in Table 2 and are plotted against (Tavg – Tamb) in Fig.
5.
Table 2 Heat transfer coefficient values computed from steady state temperatures obtained from experiments for
heat sink with heating coil kept in open air (ambient temperature, Ta = 26 o
C).
Voltage
(Volts)
Current
(Amperes)
Heat
input
(Watts)
Temperature,
Tavg (o
C)
Heat
transfer
coefficient,
h
(W/m2o
C)
10 0.23 2.3 35 11.77
13 0.31 4.03 41.33 12.11
16 0.38 6.08 46 14.00
19 044 8.36 51.67 15.00
20
25
30
35
40
45
50
55
60
0 2000 4000
Temperature,oC
Time, s
2.3 W
4.03 W
6.08 W
8.36 W
0
10
20
30
40
50
60
0 5 10
Temperature,oC
Heat input, W

Fig. 5 Heat transfer coefficient values as computed from experimentally obtained steady state temperatures of
the heat sink.
IV. THEORETICAL MODELLING
2-D Analysis
In 2-D analysis, the 2-D conduction equation with heat generation is solved. The equation is as given
below [14]
(
𝜕2 𝑇
𝜕𝑥 2 +
𝜕2 𝑇
𝜕𝑦 2) +
𝑄
𝑘
=0 (4)
Where 𝑄 the source term and k is is the thermal conductivity of the material.
Eq. (4) is used to model the heat transfer in the vertical base plate. The finite volume discretization is
used to approximate the partial differential equation and the resulting algebraic equations are solved using
Gauss-Seidal approach. The vertical base plate is divided into a number of finite volume cells. A typical finite
control volume around a point P of the plate is shown in the Fig. 6. TP represents the temperature at P while TW,
TE, TS and TN represent the temperatures at the West, East, South and North neighbours of the point P. Δx and
Δy are the dimensions of the control volume while (δx)w, (δx)e, (δy)s, and (δy)n are the distances of the West,
East, South and North neighbours from the point P respectively. The finite volume discretization equation can
be written, following the notation in Patankar [15], as
aPTP=aETE+ aWTW+ aNTN+ aSTS+ b (5)
Where
aE=
ke tΔ𝑦
(𝛿𝑥 ) 𝑒
, (5a)
aW=
kw tΔ𝑦
(𝛿𝑥 ) 𝑤
, (5b)
aN=
kn tΔX
(𝛿𝑦 ) 𝑛
, (5c)
aS=
kstΔ𝑦
(𝛿𝑦 ) 𝑠
(5d)
t = thickness of the plate
The source term Q can have the following components
Q=Qgen+Qflux+Qconv+Qfin (6)
where
Qgen =heat generated within the control volume
= qgen tΔx Δy, (6a)
Qflux=heat flux through the surface
= qflux Δx Δy, (6b)
10
12
14
16
0 5 10
Heattransfer
coefficient,W/m2oC
Heat input, W
∆𝑦

Qconv=convective heat transfer with the surroundings
= hface Δx Δy (TP-Tamb), (6c)
Qfin=heat transferred with the surroundings through fin
=𝜂hfin Afin(TP-Tamb) (6d)
The coefficient ap and the constant b in Eq. 5 can now be given as,
aP = aE+ aW+ aS+ aN+ hface Δx Δy+ 𝜂hfin Afin, (7)
b = qgen tΔx Δy+Qflux Δx Δy+ hface Δx Δy Tamb+ 𝜂hfin AfinTamb . (8)
In the equations above qgen is the heat generated per unit volume, qflux is the heat flux per unit area,
hface is the heat transfer coefficient for exchanging heat with surroundings, 𝜂 is fin efficiency, hfin is the
effective heat transfer coefficient of fin, Afin is the effective surface area of fin through which heat is exchanged
with surroundings and Tamb is the ambient temperature.
The heat transfer coefficient hface is modelled using the standard correlations [13], [14]:
Nu = 0.59 (Gr Pr)0.25
(9)
Where
Nusselt number Nu =
hface y
𝑘
(9a)
Grashof number Gr =
𝑔𝛽 ∆𝑇𝑦3
𝜗2 (9b)
Prandtl number Pr=
𝐶 𝑝 𝜇
𝑘
(9c)
Fin efficiency 𝜂 is given by
𝜂=
𝑡𝑎𝑛 𝑕 (𝑚𝐿)
𝑚𝐿
(10)
Where
mL= Lc
1.5(
(
𝑕
𝑘𝐴 𝑚
)0.5
(10a)
Lc= L+
𝑡
2
(10b)
Am= tfin Lc (10c)
tfin is thickness of fin.
The heat transfer coefficient hfin is modelled using the standard correlations [13][14]:
for upper surface of fin
Nu = 0.54 (Gr Pr)0.25
(11)
for lower surface of fin
Nu = 0.27 (Gr Pr)0.25
(12)
hfin is obtained from equations (11) and (12) by first obtaining the average Nusselt number as
Nu = 0.405 (Gr Pr)0.25
(13)
Nu =
hface Le
𝑘
(14)
The Grashof number in equations (11), (12) and (13) is given by
Gr =
𝑔𝛽 ∆𝑇𝐿𝑒3
𝜗2 (14a)
Pr=
𝐶 𝑝 𝜇
𝑘
(14b)
where Le is given by Le =
𝑙 𝑓𝑖𝑛 𝑏 𝑓𝑖𝑛
2(𝑙 𝑓𝑖𝑛 + 𝑏 𝑓𝑖𝑛 )
(14c)
Afin in equation (6d) is given as
Afin=2Δ𝑥𝑙𝑓𝑖𝑛 (15)
where
𝑙𝑓𝑖𝑛 is length of the fin,
𝑏𝑓𝑖𝑛 is breadth of the fin.
Boundary Conditions:
Convective heat transfer boundary condition has been applied on all the four edges of the plate viz.
At the boundary, in general,
k(
𝜕𝑇
𝜕𝑛
)B =h(TB-Tamb) (16)
Where TB is the temperature at the boundary,
(
𝜕𝑇
𝜕𝑛
)B is the temperature gradient at the boundary.
The resulting algebraic equations are solved using Gauss-Seidal approach as mentioned earlier. Iteration
method has been used for achieving the convergence of the solution of the equations. A computer programme
has been written in C language for computation of temperature field.

A mesh of size 60x32 has been used for the solution. It is uniform in x-direction and non-uniform in y-
direction. The details of mesh co-ordinates are given in the appendix.
V. RESULTS AND DISCUSSION
The value of heat input and the corresponding heat flux value is given as input to the C-program to get
the temperature field by using the approach as given in section IV above. Temperature values corresponding to
the three thermocouple locations are noted from the temperature field and average of these three temperatures,
Tavg is calculated. This is done for different values of the heat input. Heat input values, and the corresponding
heat flux values, the average temperature of the heat sink and heat transfer coefficient is shown in Table 7. Heat
transfer coefficient is computed from equation similar to Eq. (3). Fig. 7 shows the average temperature of the
heat sink plotted against the corresponding heat inputs. Fig. 8 shows the heat transfer coefficient values plotted
against the corresponding heat inputs. Temperature profile showing the variation of temperature along the mid-
line (x=60e-3) as function of the vertical distance y is shown in Fig. 9.
Table 7 Heat sink temperature as evaluated by 2D code for heat sink with heating coil kept in open air
Heat
input
(W)
Heat
flux
(W/m2
)
Temperature,
Tavg (o
C)
Heat
transfer
coefficient,
h (W/m2o
C)
2.3 1288.52 34.88 11.93
4.03 2257.70 39.72 13.54
6.08 3406.16 44.60 15.06
8.36 4683.47 49.99 16.06
Fig. 7 Temperature obtained by 2-D analysis
Fig. 8 Heat transfer coefficient obtained by 2-D analysis
Fig. 9 Temperature profile ,
, ,
0
20
40
60
0 5 10
Temperature,oC
Heat input ,W
0
10
20
0 5 10
Heattransfer
coefficient,W/m2o
C
Heat input ,W
0
0.02
0.04
0.06
300 310 320 330
Verticaldistance
frombaseofheat
sink,m
Temperature, K

V. COMPARISION OF EXPERIMENTAL RESULTS WITH THEORETICAL RESULTS
Comparison between heat sink temperatures from experiment and theory is shown in Fig. 10.
Fig. 10 Comparison between heat sink temperatures from experiment and theory
Comparison between heat transfer coefficient from experiment and theory is shown in Fig. 11
Fig. 11 Comparison between heat transfer coefficient from experiment and theory
VI. SUMMARY AND CONCLUSION
Temperature of heat sink is measured experimentally. The experimentally measured temperatures have
been compared with those predicted by the theory and have been found to compare well with each other.
REFERENCES
1) Padmakar A. Deshmukh, Kamlesh A. Sorate and Ravi Warkhedkar, “Modeling and Analysis of
Rectangular Fin Heat Sinks under Natural Convection for Electronic Cooling”, International Journal of
Engineering Research and Technology, ISSN 0974-3154 Volume 4, Number 1, pp 67-74, (2011).
2) Mobedi M., Sunden B., “Natural Convection Heat Transfer from a Thermal Heat Source Located in a
Vertical Plate Fin”, International Communications in Heat and Mass Transfer 33, 943-950, (2006).
3) Jan Bijanpourian,” Experimental determination of the thermal performance of a free standing fin
Structure copper heatsink”, Dept. of Public Technology, Mälardalen University, MdH
4) MassimilianoRizzi, Marco Canino, Kunzhong Hu, Stanley Jones, Vladimir Travkin, Ivan Catton,
“Experimental Investigation of Pin Fin Heat Sink Effectiveness”, MAE Department, 48-121
Engineering IV, UCLA, Los Angeles, CA 90024-1597
5) Starner K.E. and McManus H.N., “An Experimental Investigation of Free Convection Heat Trasnfer
from Rectangular Fin Arrays”, Journal of Heat Transfer, 273-278, (1963).
6) Leung C.W., Prober S.D. and Shilston M.J. “Heat Exchanger: Optimal Separation for Vertical
Rectangular Fins Protruding from a Vertical Rectangular Base”, Applied Energy, 77-85, (1985).
7) Harahap F., Lesmana H., Dirgayasa A.S., “Measurements of Heat Dissipation from Miniaturized
Vertical Rectangular Fin Arrays under Dominant Natural Convection Conditions”, Heat and Mass
Transfer 42, 1025-1036, (2006).
8) Yalcin H.G., Baskaya S., Sivrioglu M., “Numerical Analysis of Natural Convection Heat Transfer from
Rectangular Shrouded Fin Arrays on a Horizontal Surface”, International Communications in Heat and
Mass Transfer 35, 299-311, (2008).
9) Yüncü H. and Mobedi M., “A Three Dimensional Numerical Study on Natural Convection Heat
Transfer from Short Horizontal Rectangular Fin Array”, Heat and Mass Transfer 39, 267-275, (2003).
10) Kundu B., Das P.K., “Performance and Optimum Design Analysis of Convective Fin Arrays Attached
to Flat and Curved Primary Surfaces”, International Journal of Refrigeration, 1-14, (2008).
0
20
40
60
0 5 10
Temperature,oC Heat input ,W
0
5
10
15
20
0 5 10
Heattransfer
coefficient,W/m2K
Qinput, W

11) Leung C.W. and Probert S.D., “Thermal Effectiveness of Short Protrusion Rectangular, Heat
Exchanger Fins”, Applied Energy, 1-8, (1989).
12) Vollaro A.D.L., Grignaffini S., Gugliermetti F., “Optimum Design of Vertical Rectangular Fin
Arrays”, International Journal of Thermal Sciences 38, 525-259, (1999).
13) C P Kothandaraman, “Heat and mass transfer data book”, New age international publishers, fifth
edition, 2005.
14) J P Holman, “Heat transfer”, McGraw- Hill, 1989
15) Patankar,”Numeral heat transfer and fluid flow”, Hemisphere publishing corporation, 2005
APPENDIX
The mesh used in this work (section IV) of 60x32 i.e. it has 60 divisions in x-direction and 32 divisions in y-
direction.
The mesh is uniform in x-direction with ∆𝑥 = 0.002 𝑚. The y co-ordinates are as shown below:
y (in m) = 0.000000, 0.001563, 0.003125, 0.004688, 0.006250, 0.007813, 0.009375, 0.010938, 0.012500,
0.014063, 0.015625, 0.017188, 0.018750, 0.020313, 0.021875, 0.023438, 0.025000, 0.026563, 0.028125,
0.029688, 0.031250, 0.032813, 0.034375, 0.035937, 0.037500, 0.039062, 0.040625, 0.042187, 0.043750,
0.045312, 0.046875, 0.048437, 0.050000.

Experimental and Theoretical Study of Heat Transfer by Natural Convection of a Heat Sink Used for Cooling of Electronic Chip

Related slideshows

More Related Content

Experimental and Theoretical Study of Heat Transfer by Natural Convection of a Heat Sink Used for Cooling of Electronic Chip