Proceedings of the spring 2014 Thermal Engineering Projects Presentation at UC Davis

Davis Thermal Solutions Inc.

May 12, 2014, Davis, USA

EME 165

HEAT DISSIPATION IN A 500W MICROCHIP

Kevin Omar Arias Garcia

Email: keoariasgarcia@ucdavis.edu

ABSTRACT
This report shows the optimization of a heat dissipation system for a microprocessor, and the effects of the changes in the
heat sink dimensions, fluid velocity and the heat conductor materials.

sink cross section, only one symmetric portion of the heat sink is
used in order to reduce significantly the amount of time needed
to do the simulation.
q00 p

Adiabatic
Isotherm

Microprocesor
Thermal Interfase
Heat Sink

Fluid
(25oC)

hf

Adiabatic
Isotherm

height

width
PROBLEM DESCRIPTION
Heat from a computer microprocessor is transfered trough a
thermal interface to a heat sink and then to a liquid as is shown in
Figure 1. The microprocessor generates 500[W ] as heat, and has
20[mm]x20[mm] in size. The thermal interface has a resistivity
of R00 = 2x10 5[oCm2 /W ]

Adiabatic
Surface

wf

(a) Thermal

(b) Dimensions

FIGURE 2: SECTION ANALYSIS.

The dashed section in Figure 1 shows the portion of the heat

sink to be simulated. The lateral sides, where the heat sink material continues, are modeled as adiabatic isotherm surfaces because of a negligible temperature gradient.At the bottom there is
also an adiabatic surface due to the insulation material in contact
with the fin tip, and finally heat from the chip is being applied at
the top surface and dissipated to the fluid channel, as shown on
Figure 2a.

Insulator
Liquid Channel

FIGURE 1: HEAT SYSTEM BASIC GEOMETRY

The parameters that will vary to get the optimum heat sink
dissipator are:
-Fin width w f and fin height h f (Figure 2b).
-Liquid channel velocity u.
-Material Thermal properties: Density (), Specific heat capacity (c p ), Thermal conductivity (k).

SOLUTION APPROACH
Because of the simulation time needed when a parameter
is changed, the system will be analyzed as an one dimensional
steady state heat transfer problem.
Another relevant point is the maximum temperature at the
surface of the chip must be 75[oC], with the thermal interface

In order to test the variations in the parameters of the heat

sink, the system is simulated in Matlab with a finite-difference
two dimensional model. Because of the geometry of the heat
1

resistance given (Rti00 = 2x10 5[oCm2 /W ]), we can get the temperature drop in this material:

T p Ts = q p

Rti00
= 25[oC]
A p

Simulation method
Once the optimal approximated dimensions (OAD) are
known, the same procedure as the approximation method will be
followed, but with the finite difference model, and only for points
near to the OAD, in order to validate the approximation method
and get the fin dimensions that give us the minimum temperature,
then materials and fluid speeds will be tested to get a better understand of their influence in the heat sink performance in order
to select the optimal configuration of these variables.

(1)

And with this value, the maximum permissible temperature

at the heat sink surface is Tsmax = 50[oC].

RESULTS
Approximation method

Simplified model
Assuming the steady state in the dotted symmetric section
from the heatsink in Figure 1, it can be modeled as a thermal
conductor with a uniform cross section fin , with boundary conditions of the case B (Insulated Tip) from the table 3.4 in the
textbook [1]. The simplified system is shown in Figure 3, that
is modeled as one dimensional thermal circuit, where heat is applied to the surface of the conductive material, modeled as Rcond ,
at the base surface heat is transfered by the fin as q f and by convection where the base surface is in contact with the fluid Rconv .

45
Temp [C]

qsection

50

qsection
Ts

Rcond

40

Rconv

qf

Tb

0
1
35
0

2
0.5

1
1.5
2

FIGURE 3: THERMAL RESISTIVE CIRCUIT

1
Rconv

Figure 4 shows that the principal dimension that affects the

heat transfer is the fin height, and the maximum temperature difference varying the fin width is less than 10[oC], so h f must be
the maximum possible and w f could be determined by other desired properties.

!
(2)

[C]

Ts = T + qsection Rcond +

hf [mm]

FIGURE 4: Ts FOR MULTIPLE DIMENSIONS.

The temperature at the surface of the heatsink is obtained

with detail in the Apendix A and is given by the equation (2).

1
+ Mtanh(mL)

wf [mm]

T T

min

Aproximation method
For the approximation to the optimal dimensions, the heat
sink material will be aluminum (k = 237[W /mK]) and a fluid
velocity in the channel will be v = 4.65[m/s]. Ts will be calculated for different widths and heights of the fin and then it will be
plotted in a 3D surface .
Once having the best dimensions of the fin, the geometry
will remain constant and the material thermal conductivity will
vary and the temperature values for different k values will be
plotted. The same procedure will be done with the variation of
the fluid rate.

6
4
2
500

1000

1500
2000
k [W/(m K)]

2500

3000

FIGURE 5: Ts FOR DIFFERENT MATERIALS.

( w f = 1.9mm, h f = 3.9mm, v = 4.65m/s)

Figure 5 shows that the heat sink material isnt really important before k = 500[W /mK] the temperature improvement would
be less than 5[oC]. The only considerable change could be change
2

from a material with low conductivity to one like Diamond or

Graphite pyrolytic, but the change will be around 10[oC].

100

min

80
60

TsTs

100

40

Ts [C]

20

50
0

10

12

14

16

18

v [m/s]
0

FIGURE 9: Ts FOR DIFFERENT FLUID VELOCITIES.

v [mm/s]

(w f = 1.9mm, h f = 3.9mm, Aluminum)

x 10

FIGURE 6: Ts FOR DIFFERENT FLUID SPEEDS.

3

Figure 6 shows a very small range of velocities where there

is laminar flow and the temperature remains almost constant,
then, when the flow changes to turbulent (RE > 2300) the temperature tends to decrease exponentially, getting considerable
changes in the surface temperature with small variation in the
liquid velocity.

x 10

3.5

60

60

60

59

59

59

x 10
36

58

58

57

57

57

56

56

56

55

55

55

54

54

54

53

53

53

52

52

52

51

51

51

50

50

50

49

49

49

33

35

35
29
28
27

30

34

34

26

33

29

30

33

2.5 30

2.5

47

46

48

48

47

47

height (m)

27

28

31

28
30
1.5

1.5

26

48

height (m)

height (m)

1.5

32

29

29
2

440
442
43
41
39
38
36
35
34
33
32
31
30
29

36
36

31

31

31

3.5

28

42
41
40
3933
38
37
36
35
34
32
31
30
29

37

37

32

32

32

2.5

39
38

33

33

3.5

x 10

34
34

34

58

35
35

35

32
31
29 30

(w f = 1.9mm, h f = 3.9mm, Aluminum)

30

27

46

46

45
4443
42

45
4443
42

27

29

45

Simulation method

29

44

0.5

0.5

0.5

26

1
width (m)

1.5

2
3

x 10

min

10

TsTs

0.5

43

0.5

1
width (m)

1.5

0.5

x 10

1
width (m)

1.5

2
3

x 10

(a) Small channel

0

0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

(b) Thin channel

(c) Tall channel

FIGURE 10: TEMPERATURE GRADIENTS OF DIFFERENT

wf [mm]

CONFIGURATIONS

FIGURE 7: Ts FOR DIFFERENT WIDTHS.

( h f = 3.9mm, v = 4.65m/s,Aluminum)
convection will be predominant at the top and there will be a
small gradient of temperature added to a small amount of heat
transfer by convection at the bottom.

In the results of Table 2 in the Appendix B, the temperature variation for different heights and weights of the fin shows a
similar behavior as the one shown in Figure 4, with a difference
of temperatures of 7[oC] along w f and a temperature variation
greater than 50[oC] changing h f .

OPTIMAL SOLUTION SIMULATION

From the results obtained in the previous simulations, the
optimal solution proposed is :

TsTs

min

w f = 1.5[mm]
h f = 3.5[mm]
Material: Copper Pure
v=19.05 [m/s] w/ pump MCP655 (20 lt per min)

4
2
0

500

1000

1500

2000

2500

k [W/(m K)]

FIGURE 8: Ts FOR DIFFERENT MATERIALS .

(w f = 1.9mm, h f = 3.9mm, v = 4.65m/s)

34
33
32
Temperature (C)

In the Table 1 (Appendix B) is shown that with different materials the temperature improvement is small as the approximated
model suggested it, with a maximum temperature difference of
about 6[oC], from Aluminium to Diamond.
Figure 10 shows in a small chanel there will be a predominant conduction phenomena rather than convection, in a thin
channel there will be a conductive phenomena but, with a smaller
gradient in the vertical gradient, finally, with a tall channel the

31
30
29
28
27
26
25

500

1000
1500
2000
Time steps ( 5e07 seconds)

2500

3000

FIGURE 11: Ts for different fluid velocities.

combination of heat transfer by conduction and convection. The

fluid velocity with a constant amount of flow rate will vary with
the channel geometry, but will improve with thinner channels due
to the increment in the liquid velocity (u = V /A) which also increase the Reynolds number creating a turbulent flow and a bigger convection heat transfer coefficient. The choice of material of
the heat sink must be one as good as the aluminum, but selecting
one better than aluminum doesnt alter the results significantly,
leading the decision to other factors as weight, price, the manufacturing process or other physical properties of interest to the
application. The width of the fin, has a low impact in the heat
transfer performance, so it can be modified in order to make the
heat sink lighter, get a different liquid flow properties, or ease the
manufacturing processes of the heat sink. The one dimensional
resistive model proved to be inaccurate (10[oC]), but it shows
the same heat sink behavior than the simulation when parameters
changes, easing the understand of the relative importance of each
parameter and reducing the amount of time needed to reach the
optimal solution.

x 10

33

33

32

32
32
3.5

31
30
29
28

31
31

30

27

3
30

29

28

30

2.5

29

height (m)

29
2

28

28
27

1.5

REFERENCES
[1] Fundamentals of Heat & Mass Transfer, TL Bergman, AS
Lavine, FP Incropera & DP Dewitt, 7th Ed., John Wiley &
Sons, Inc. (2011)

27

0.5

0.5

1
width (m)

1.5

2
3

x 10

Appendix A: DERIVATION OF Ts FOR RESISTIVE THERMAL ANALYSIS

The heat flux per unit of area can be defined as:

FIGURE 12: Ts for different fluid velocities.

The maximum temperature at the surface of the chip is:

Ts p = 25[oC] + 33.1345[oC] = 58.1345[oC].

q00 p =

(5)

From the resistive thermal circuit

(3)

Radiation at the top of the surface:

(Assuming = 1 and Tsur = 25[oC])
4
Qrad = A p (Ts4 p Tsur
) = 0.0938[W ]

q p
qSection
=
A p
ASection

qsection = q f +
(4)

Tb T
Rconv

(6)

and

Radiation can be ignored because is less than 0.02% of the

total power generated by the microprocessor.

Ts Tb
Rcond

(7)

1
h(Asection A f in )

(8)

qsection =
where

DISCUSSION
As is shown in the results, the predominant variable to enhance the heat transfer in the system is the fluid velocity and the
fin height. The fin height enhances the heat transfer increasing
the temperature gradient in the horizontal direction, resulting in a

Rconv =
4

Rcond =

L
kAsection

(9)

TABLE 2: MAX AND MIN SURFACE TEMPERATURES FOR

ALUMINIUM WITH A VELOCITY OF 4.65m/s.

Temperature Range
q f = Mtanh(mL)(Tb T )

(10)

h f [mm]

w f [mm]

Tsmax [o ]

Tsmin [o ]

3.9

0.1

48.670

39.485

3.9

0.5

47.04

37.46

3.9

44.66

39.043

3.9

1.5

41.73

40.242

3.9

1.9

41.05

38.5212

3.5

1.9

43.716

42.29

3.0

1.9

46.9

46.22

2.0

1.9

55.33

55.14

0.1

3.9

>90.79

>90.78

with

M=

r
m=

(11)

hPkAc

hP
kAc

(12)

Solving for Ts

Ts = T + qsection Rcond +

1
Rconv

1
+ Mtanh(mL)

!
(13)

Ac and P are the cross sectional area and perimeter of the fin
respectively.
TABLE 3: MAX AND MIN SURFACE TEMPERATURES FOR

ALUMINUM WITH A VELOCITY VARIATION

Appendix B: SIMULATION VALUES PLOTTED IN RESULTS SECTION

Temperature Range

TABLE 1: SURFACE TEMPERATURE WITH DIFFERENT

Flow rate [m/s]

h[W/m.K]

Tsmax [o ]

Tsmin [o ]

0.1

1455 (L)

>124.737

>115.58

17853

>65.97

>62.50

MATERIALS
Temperature Range
Material

k[W /m.K]

Tsmax [o ]

Tsmin [o ]

31085

52.48

49.9

Aluminium

273

41.212

39.49

42995.6

46.699

44.658

Copper

401

39.46

38.05

74859.7

39.53

38.4294

Silver

429

39.29

37.95

12

130338

36.02

34.27

Diamond

2300

35.38

34.95

17.1

172982

34.7839

32.6642

Appendix C: SIMULATION CODE

.1 EME165Project1Microchip.m code
1
2

%% Clear Workspace/Command/Figures
clear all; clc; close all;

3
4
5
6
7
8
9
10
11
12

%% Define Geometry
width=2e-3;
%width of symmetry element (m)
height=4e-3;
%height of symmetry element (m)
wf=1.5e-3;
%width of fin in symmetry element (m)
hf=3.5e-3;
%height of fin in symmetry element=channel height (m)
wc=width-wf;
%channel width (m)
hs=height-hf;
%thickness of base (where there is no fin) (m)
dx=0.1e-3;
%node size delta x (m)
dy=0.1e-3;
%node size delta y (m)

13
14
15
16
17

Ncolumn=int16(width/dx)+1;
%number of rows;
Nrow=int16(height/dy)+1;
%number of columns;
fluid_yloc=int16(hs/dy)+1; %row number location of fluid y contact
fluid_xloc=int16(wf/dx)+1; %column number location of fluid x contact

18
19
20
21
22
23
24
25
26
27
28
29
30

%% Define temperatures, properties, constants

km=[237,401,429,2300];
%material conductive heat transfer coeff.(W/m.K)
rhom=[2702,8933,10500,3500];%heat sink material density (kg/m3)
com=[903,385,235,509];%heat sink material specific heat (J/kg.K)
Tf=25;
%fluid temperature (degC)
material=2; %Material: 1=Aluminum 2=Copper 3=Silver 4=Diamond
k=km(material);
%heat sink material conductive heat transfer coeff.
rho=rhom(material);
%heat sink material density (kg/m3)
c=com(material);
%heat sink material specific heat (J/kg.K)
CPUWatts=500;
%CPU heat dissipation (W)
CPUSize=0.02;
%dimension of one side of CPU (m)
heatflux=CPUWatts/(CPUSize)2;
%heatflux from top surface (W/m2)

31
32

%% Liquid properties

33
34
35
36
37
38
39

rhoh=1000; %Liquid density (kg/m3)

muh=118e-6; %Liquid viscosity (N s /m2)
kfh=0.642; %Thermal conductivity of the liquid (W/mK)
Prh=0.86; %Prandtl number
h_f=hf;
w_f=wf;

40
41
42
43
44
45
46

Ach=2*h_f*(width-w_f); %Cross-section area of channel

vp=0.02/(5*Ach*60); %Liquid velocity from flow rate of the pump
v=vp; %Liquid velocity
Ph=2*h_f+4*(width-w_f); %Perimeter of the channel
Dh=4*Ach/Ph; %Hydraulic diameter
Reh=rhoh*v*Dh/muh; %Reynolds number calculation

47
48
49

if Reh>2300 %Turbulent flow

h=0.023*Reh0.8*Prh0.3*kfh/Dh;

%convective heat transfer coeff.

6

50
51
52

else
h=3.61*kfh/Dh; %convective heat transfer coeff. to the fluid (W/m2.K)
end

53
54
55
56
57

%% Initial temperature of sink

Tnew=zeros(Nrow,Ncolumn)+Tf; % initially heat sink is at fluid temperature
incr=zeros(Nrow,Ncolumn);
% "incr" matrix stores change in temperature at each timestep

58
59
60
61

%% Initialize indexing and other variables for Euler forward time stepping method
ti=0;
% initialize time index to use for time stepping
err=1;
% initial error start the while loop

62
63
64
65
66
67
68
69
70

%% Solve ODE using Euler forward time stepping method

% Initialize Euler forward time stepping method
% You can continue solving from the last solution by
% doing "Run Section" here, after changing tolerance and step limit.
dt=0.8e-6;
% time step (s)
tol=0.3e-5;
% Temperature tolerance to determine steady state
SetStepLimit=1100000; % iteration limit
tic

71
72
73
74
75
76
77

while err>tol
% loop while error is larger than preset tolerance
ti=ti+1;
% move time step by one
Temp(:,:,ti)=Tnew;
T=Tnew;
% use new temperature array from the previous step to determine change in tem
incr=zeros(Nrow,Ncolumn);
% reset "incr" array, which tracks temperature change at this tim

78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98

%% Left Boundary (n=1, where m is row number and n is column number)

% Top left corner (heat flux and adiabatic)
n=1;
%column number = 1
m=1;
%row number = 1
incr(m,n)=4*dt/(rho*c*dx*dy)*( ...
heatflux*dx/2 + ...
% heat flux from above
k*dx/2*(T(m+1,n)-T(m,n))/dy + ...
% conduction from below
k*dy/2*(T(m,n+1)-T(m,n))/dx );
% conduction from right
% Left wall (symmetry adiabatic)
for m=2:Nrow-1
incr(m,n)=2*dt/(rho*c*dx*dy)*( ...
k*dx/2*(T(m-1,n)-T(m,n))/dy + ...
% conduction from above
k*dx/2*(T(m+1,n)-T(m,n))/dy + ...
% conduction from below
k*dy*(T(m,n+1)-T(m,n))/dx );
% conduction from right
end
% Bottom left corner (adiabatic and adiabatic)
m=Nrow;
incr(m,n)=4*dt/(rho*c*dx*dy)*( ...
k*dx/2*(T(m-1,n)-T(m,n))/dy + ...
% conduction from above
k*dy/2*(T(m,n+1)-T(m,n))/dx );
% conduction from right

99
100
101

%% "fin" area, where the heat sink extends to the bottom insulation
for n=2:fluid_xloc-1
7

% top boundary condition (heat flux)

m=1;
incr(m,n)=2*dt/(rho*c*dx*dy)*( ...
heatflux*dx + ...
k*dy/2*(T(m,n-1)-T(m,n))/dx + ...
k*dx*(T(m+1,n)-T(m,n))/dy + ...
k*dy/2*(T(m,n+1)-T(m,n))/dx );

102
103
104
105
106
107
108

%
%
%
%

heat flux from above

conduction from left
conduction from below
conduction from right

109

% inside nodes (no boundary condition)

for m=2:Nrow-1
incr(m,n)=dt/(rho*c*dx*dy)*k*( ...
dx/dy*(T(m-1,n)-T(m,n)) + ...
dy/dx*(T(m,n-1)-T(m,n)) + ...
dx/dy*(T(m+1,n)-T(m,n)) + ...
dy/dx*(T(m,n+1)-T(m,n)));
end

110
111
112
113
114
115
116
117

% conductions from four sides

118

% bottom boundary condition (adiabatic)

m=Nrow;
incr(m,n)=2*dt/(rho*c*dx*dy)*( ...
k*dy/2*(T(m,n-1)-T(m,n))/dx + ...
k*dy/2*(T(m,n+1)-T(m,n))/dx + ...
k*dx*(T(m-1,n)-T(m,n))/dy );

119
120
121
122
123
124
125

% conductions from three sides

% as bottom is adiabatic

end

126
127
128
129
130
131
132
133
134
135

%% Right boundary of "fin" area where fluid touches the fin

n=fluid_xloc;
% top boundary condition (heat flux)
m=1;
incr(m,n)=2*dt/(rho*c*dx*dy)*( ...
heatflux*dx + ...
k*dy/2*(T(m,n-1)-T(m,n))/dx + ...
k*dx*(T(m+1,n)-T(m,n))/dy + ...
k*dy/2*(T(m,n+1)-T(m,n))/dx );

136
137
138
139
140
141
142
143
144

% inside nodes (no boundary condition)

for m=2:fluid_yloc-1
incr(m,n)=dt/(rho*c*dx*dy)*k*( ...
dx/dy*(T(m-1,n)-T(m,n)) + ...
dy/dx*(T(m,n-1)-T(m,n)) + ...
dx/dy*(T(m+1,n)-T(m,n)) + ...
dy/dx*(T(m,n+1)-T(m,n)));
end

145
146
147
148
149
150
151
152
153

% internal corner (with convection)

m=fluid_yloc;
incr(m,n)=4/3*dt/(rho*c*dx*dy)*( ...
k*dx*(T(m-1,n)-T(m,n))/dy + ...
k*dy*(T(m,n-1)-T(m,n))/dx + ...
k*dx/2*(T(m+1,n)-T(m,n))/dy + ...
k*dy/2*(T(m,n+1)-T(m,n))/dx + ...
h/2*(dx+dy)*(Tf-T(m,n)) );
8

% conduction from four sides (with varying ar

% convection from lower right side

154

% fin right boundary (with convection)

for m=fluid_yloc+1:Nrow-1
incr(m,n)=2*dt/(rho*c*dx*dy)*( ...
k*dx/2*(T(m-1,n)-T(m,n))/dy + ...
k*dy*(T(m,n-1)-T(m,n))/dx + ...
k*dx/2*(T(m+1,n)-T(m,n))/dy + ...
h*dy*(Tf-T(m,n)));
end

155
156
157
158
159
160
161
162

% conduction from three sides

% convection from right side

163

% fin bottom right corner (convection and adiabatic)

m=Nrow;
incr(m,n)=4*dt/(rho*c*dx*dy)*( ...
k*dx/2*(T(m-1,n)-T(m,n))/dy + ...
k*(dy/2)*(T(m,n-1)-T(m,n))/dx ...
+ h*(dy/2)*(Tf-T(m,n)));

164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179

%% Base area where there is no fin (top "cover" of fluid channel)

for n=fluid_xloc+1:Ncolumn-1
% top boundary condition (heat flux)
m=1;
incr(m,n)=2*dt/(rho*c*dx*dy)*( ...
heatflux*dx + ...
k*dy/2*(T(m,n-1)-T(m,n))/dx + ...
k*dx*(T(m+1,n)-T(m,n))/dy + ...
k*dy/2*(T(m,n+1)-T(m,n))/dx );

180

% inside nodes (no boundary condition)

for m=2:fluid_yloc-1
incr(m,n)=dt/(rho*c*dx*dy)*k*( ...
dx/dy*(T(m-1,n)-T(m,n)) + ...
dy/dx*(T(m,n-1)-T(m,n)) + ...
dx/dy*(T(m+1,n)-T(m,n)) + ...
dy/dx*(T(m,n+1)-T(m,n)));
end

181
182
183
184
185
186
187
188
189

% bottom boundary condition (convection)

m=fluid_yloc;
incr(m,n)=2*dt/(rho*c*dx*dy)*( ...
k*dy/2*(T(m,n-1)-T(m,n))/dx + ...
k*dx*(T(m-1,n)-T(m,n))/dy + ...
k*dy/2*(T(m,n+1)-T(m,n))/dx + ...
h*dx*(Tf-T(m,n)));

190
191
192
193
194
195
196
197

end

198
199
200
201
202
203
204
205

%% Right Boundary (symmetry adiabatic)

%Top right corner (heat flux and adiabatic)
n=Ncolumn;
%rightmost column
m=1;
%row number = 1
incr(m,n)=4*dt/(rho*c*dx*dy)*( ...
heatflux*dx/2 + ...
k*dx/2*(T(m+1,n)-T(m,n))/dy + ...
9

k*dy/2*(T(m,n-1)-T(m,n))/dx );

206
207

% Right Boundary (symmetry adiabatic)

for m=2:fluid_yloc-1
incr(m,n)=2*dt/(rho*c*dx*dy)*( ...
k*dx/2*(T(m-1,n)-T(m,n))/dy + ...
k*dx/2*(T(m+1,n)-T(m,n))/dy + ...
k*dy*(T(m,n-1)-T(m,n))/dx );
end

208
209
210
211
212
213
214
215

% Bottom right corner of no fin area (adiatic and fluid)

m=fluid_yloc;
incr(m,n)=4*dt/(rho*c*dx*dy)*( ...
k*dy/2*(T(m,n-1)-T(m,n))/dx + ...
k*(dx/2)*(T(m-1,n)-T(m,n))/dy + ...
h*(dx/2)*(Tf-T(m,n)));

216
217
218
219
220
221
222

%% Update temperature array (at time step ti+1) with change in temperature calculated by temp
Tnew=T+incr;
err=abs(max(incr));
% Determine when to stop solution
if ti>SetStepLimit
disp(Tolerance not met within set iteration limit);
break;
end

223
224
225
226
227
228
229
230

end

231
232
233

% Plot and display the results

234
235
236
237
238

MaxSurfaceTemp=max(Tnew(1,:));
MinSurfaceTemp=min(Tnew(1,:));
TempDifference=MaxSurfaceTemp-MinSurfaceTemp;
TimePassed=ti*dt;

%
%
%
%

Maximum surface temperature

Minimum surface temperature
Temperature difference on the top surface
Time to satisfy tolerance

[X,Y]=meshgrid(0:dx:width, 0:dy:height);

% meshgrid to display temperatures with correct g

239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254

F1=figure(1);
set(F1,units,inches,OuterPosition,[0 0 10*width/height 10]); %set plot size
MinChanTemp=min(min(Tnew(fluid_yloc,:)),min(Tnew(:,fluid_xloc))); %minimum surface temperature o
ContourSpacing=1;
%contour line spacing (degC)
ContourRange=ceil(MinChanTemp:ContourSpacing:MaxSurfaceTemp);
contour(X,height-Y,Tnew,ShowText,on,LevelList,ContourRange);
boundary1=line([wf wf],[0 hf],linewidth,2);
% draw boundary lines where fluid and hea
boundary2=line([wf width],[hf hf],linewidth,2);
set(boundary1,Color,k);
set(boundary2,Color,k);
rectangle(Position,[wf,0,wc,hf],FaceColor,b);
% "white out" channel area from plot
xlabel(width (m));
%label axis
ylabel(height (m));

255
256
257

disp([Fluid velocity: ,num2str(v), m/s]);

disp([Reynolds number:,num2str(Reh)]);
10

258
259
260
261
262
263
264
265
266
267
268
269

disp([Fin width: ,num2str(wf), meters]);

disp([Fin height: ,num2str(hf), meters]);
disp([Number of nodes: ,num2str(Nrow),x,num2str(Ncolumn)]);
disp([Time step: ,num2str(dt), seconds]);
disp([Tolerance (per time step): ,num2str(tol)]);
disp([Convective heat transfer coeff: ,num2str(h), W/m2 K]);
disp([Maximum surface temperature is ,num2str(MaxSurfaceTemp), degC.]);
disp([Minimum surface temperature is ,num2str(MinSurfaceTemp), degC.]);
disp([Temperature difference is ,num2str(TempDifference), degC.]);
if ti<SetStepLimit
disp([Tolerance satisfied in ,num2str(TimePassed), seconds,, after ,num2str(ti), itera
end

270
271
272
273
274
275
276
277
278
279

F2=figure(2);
Teo=Temp(1,:,:); %Surface temperature
Teo=squeeze(Teo); %
Teo=Teo(:,1:100:length(squeeze(Teo(1,:))));
surf(Teo,EdgeColor,none,LineStyle,none,FaceLighting,phong);
xlabel(Time x 2 (\mu s));
%label axis
ylabel(width (m));
zlabel(Temperature ( C ))
system=toc

.2
1
2
3
4
5
6
7
8
9
10
11
12

ApproxMicrochip.m code

%% Clear Workspace/Command/Figures
clear; close all; clc;
Tf=25;
%Temperature of the fluid (degC)
CPUWatts=500;
%CPU heat dissipation (W)
width=2e-3;
%width of symmetry element (m)
height=4e-3;
%height of symmetry element (m)
Z=0.02;
%dimension of one side of CPU (m)
k=237;
%material conductive heat transfer coeff. (W/m.K)
rhoh=1000
%Fluid density (kg/m3)
muh=118e-6
%Fluid viscosity (N s/ m2)
kfh=0.642
% Fluid conductive heat transfer coeff. (W/m.K)
Prh=0.86
%Prandlt Number

13
14
15

w_f=1e-5:5e-5:width-.5e-4; %Fin weight array

h_f=1e-5:5e-5:height-.5e-4; %Fin height array

16
17
18
19

%% Loop width-height of the fin variable

for i=1:length(w_f)
for j=1:length(h_f)

20
21
22
23

Q=CPUWatts/Z2*(Z*2*width); %Heat into symetric portion

P=2*Z+2*(2*w_f(i)); %symetric portion fin perimeter
A_c=(2*w_f(i))*(Z); %symetric portion fin CS area
11

L=h_f(j);

24

%Lenght of the fin

25

Ach=2*h_f(i)*(width-w_f(i));
v=0.02/(5*Ach*60)
Ph=2*h_f(i)+4*(width-w_f(i));
Dh=4*Ach/Ph;
Reh=rhoh*v*Dh/muh;

26
27
28
29
30

%Channel area
%Fluid velocity
%Channel perimeter
%Hydraulic diameter
%Reynolds number

31

if Reh>2300
h=0.023*Reh0.8*Prh0.3*kfh/Dh; %Laminar convection coeff.
else
h=3.61*kfh/Dh;%Turbulent convection coeff.
end
m=sqrt((h*P)/(k*A_c));
M=sqrt(h*P*k*A_c);
R_cond=(height-h_f(j))/(k*Z*2*width);
R_conv=1/(h*Z*2*(width-w_f(i)));
q_f(i,j)=M*tanh(m*L);
T_ch(i,j)=Q*R_cond+Q/(1/R_conv+q_f(i,j))+Tf; %Temperature at base
if T_ch(i,j)>50
%discard values greater than max temperature
T_ch(i,j)=50;
end
end

32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47

end

48
49
50
51
52
53
54

figure
%Surface of Temperature(width,height)
surf(T_ch,EdgeColor,none,LineStyle,none,FaceLighting,phong);
xlabel(h_f [mm]);
ylabel(w_f [mm]);
zlabel(Temp [ C ]);
colormap(gray)

55
56

%% Loop material heat transfer coef. variable

57
58
59
60
61
62
63

clear; clc; Tf=25;

CPUWatts=500;
%CPU heat dissipation (W)
width=2e-3;
%width of symmetry element (m)
height=4e-3;
%height of symmetry element (m)
Z=0.02;
%dimension of one side of CPU (m)
k=237;
%heat sink material conductive heat transfer coeff. (W/m.K)

64
65
66
67
68
69

v=4.64511
rhoh=1000
muh=118e-6
kfh=0.642
Prh=0.86

%Fluid velocity
%Fluid density (kg/m3)
%Fluid viscosity (N s/ m2)
% Fluid conductive heat transfer coeff. (W/m.K)
%Prandlt Number

70
71
72

w_f=1.9e-3; %Constant fin width

h_f=3.9e-3; %Constant fin height

73
74
75

Ach=2*h_f*(width-w_f);
Ph=2*h_f+4*(width-w_f);
12

76
77
78
79
80
81
82

Dh=4*Ach/Ph;
Reh=rhoh*v*Dh/muh
if Reh>2300
h=0.023*Reh0.8*Prh0.3*kfh/Dh;
else
h=3.61*kfh/Dh;
end

83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100

for i=1:3000
Q=CPUWatts/Z2*(Z*2*width);
P=2*Z+2*(2*w_f);
A_c=(2*w_f)*(Z);
L=h_f;
m=sqrt((h*P)/(i*A_c));
M=sqrt(h*P*i*A_c);
R_cond=(height-h_f)/(i*Z*2*width);
R_conv=1/(h*Z*2*(width-w_f));
q_f(i)=M*tanh(m*L);
T_ch(i)=Q*R_cond+Q/(1/R_conv+q_f(i))+Tf;
end
figure
plot(T_ch-min(T_ch),k,linewidth,1.5)
ylabel(T_s-T_{min} [ C ])
xlabel(k [W/(m K)])
grid on

101
102

%% Loop fluid velocity variable

103
104
105
106
107
108
109

clear; clc; Tf=25;

CPUWatts=500;
%CPU heat dissipation (W)
width=2e-3;
%width of symmetry element (m)
height=4e-3;
%height of symmetry element (m)
Z=0.02;
%dimension of one side of CPU (m)
k=237;
%heat sink material conductive heat transfer coeff. (W/m.K)

110
111
112
113
114
115

v=4.64511
rhoh=1000
muh=118e-6
kfh=0.642
Prh=0.86

%Fluid density (kg/m3)

%Fluid viscosity (N s/ m2)
% Fluid conductive heat transfer coeff. (W/m.K)
%Prandlt Number

116
117
118

w_f=1.9e-3;
h_f=3.9e-3;

119
120
121
122
123
124
125
126

clear q_f T_ch

for i=1:50000
v=i/1000;
Ach=2*h_f*(width-w_f);
Ph=2*h_f+4*(width-w_f);
Dh=4*Ach/Ph;
Reh=rhoh*v*Dh/muh;

127

13

128
129
130
131
132

if Reh>2300
h=0.023*Reh0.8*Prh0.3*kfh/Dh;
else
h=3.61*kfh/Dh;
end

133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150

Q=CPUWatts/Z2*(Z*2*width);
P=2*Z+2*(2*w_f);
A_c=(2*w_f)*(Z);
L=h_f;
m=sqrt((h*P)/(k*A_c));
M=sqrt(h*P*k*A_c);
R_cond=(height-h_f)/(k*Z*2*width);
R_conv=1/(h*Z*2*(width-w_f));
q_f(i)=M*tanh(m*L);
hh(i)=h;
T_ch(i)=Q*R_cond+Q/(1/R_conv+q_f(i))+Tf;
end
figure
plot(T_ch-min(T_ch),k,linewidth,1.5)
ylabel(T_s [ C ])
xlabel(v [mm/s])
grid on

14