Effectof Heat Spreadingonthe Performanceof Heat Sinkvia Vapor Chamber

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 11

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/269146978

EFFECT OF HEAT SPREADING ON THE PERFORMANCE OF HEAT SINK VIA


VAPOR CHAMBER

Conference Paper · January 2006


DOI: 10.1615/IHTC13.p22.220

CITATION READS

1 3,388

5 authors, including:

Yen-Shu Chen Kuo-Hsiang Chien


Institute of Nuclear erergy research, Taoyuan City, Taiwan Industrial Technology Research Institute
23 PUBLICATIONS   294 CITATIONS    35 PUBLICATIONS   1,249 CITATIONS   

SEE PROFILE SEE PROFILE

Chi-Chuan Wang
National Chiao Tung University
467 PUBLICATIONS   13,227 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Heat Spreading and Enhaced Heat Conduction @ substrate and PCB Level View project

Fin-and-tube heat exchangers View project

All content following this page was uploaded by Yen-Shu Chen on 01 June 2016.

The user has requested enhancement of the downloaded file.


EFFECT OF HEAT SPREADING ON THE PERFORMANCE OF HEAT
SINK VIA VAPOR CHAMBER

Y.S. Chen1, K.H. Chien2, T.C. Hung3, B.S. Pei1 and C.C. Wang2
1
Department of Engineering and System Science, National Tsing Hua University,
Hsinchu, Taiwan, Republic of China
2
Energy & Environment Research Laboratories, Industrial Technology Research Institute,
Hsinchu, Taiwan, Republic of China
3
Department of Mechanical and Automation Engineering, I-Shou University,
Kaohsiung County, Taiwan, Republic of China

Abstract
This study investigates the thermal spreading capability of the vapor chamber subject to the change
of heater size. The vapor chamber was assembled with a plate-fin heat sink and mounted in a wind
tunnel. Additionally, a copper plate and another aluminum plate having the same size were also
adapted for comparison. Experimental results indicate that the effective overall thermal resistance
across the heat spreaders increases with respect to smaller heater size. For vapor chamber, the
corresponding temperature difference between evaporator and condenser increases with the input
power and it is considerably augmented when the heater size is reduced. For the same heater size,
the thermal resistance of vapor chamber is about 15% less than the copper plate. In addition, the
condenser of vapor chamber shows an excellent uniform distribution in temperature and this
phenomenon is especially pronounced when its size is further reduced. Comparing with the metal
plates, the vapor chamber provides a lower temperature rise and better spreading performance,
which is helpful for the thermal management of the small chips.

Nomenclature
Af frontal area of the heat sink (m2) T average temperature (°C)
Ar area ratio between the heat source and U frontal air velocity (m/s)
the base plate ρ density
Cp specific heat (J/kg⋅K)
h convective heat transfer coefficient
(W/m2⋅K) Subscripts
k thermal conductivity (W/m⋅K) air air
Pin input power (W) amb ambient
q heat flux (W/m2) bot bottom surface
Qloss heat loss (W) c condenser
Qout energy absorbed by the air flow (W) e evaporator
RMS root mean square of the measured in inlet air flow
condenser temperature (°C) md maximum difference between the
Rth thermal resistance (°C /W) measured temperatures on the surface
T temperature (°C) out outlet air flow
∆T temperature rise from the ambient top top surface
temperature (°C)

1. Introduction
For electronic products, such as PCs and the consumer electronics, demands for more complex
functions along with requirements being integrated into a small volume, are inevitably resulting in a
higher power density. To effectively remove the generated heat, a heat sink is usually adhered to the
chip as the heat dissipation device. While a heat sink is attached on a heat source smaller than its
base area, the temperature distribution of the base surface is generally not uniform due to the
thermal spreading effect caused by the concentrated heat source. Thus, the thermal resistance of the
base plate is larger than that calculated by simply using the one-dimensional conduction method.
This discrepancy becomes significant provided that the heat source has a smaller contact area than
that of the base plate. Notice that smoothing the temperature distribution subject to the influence of
concentrated heat source is very crucial in thermal management of the electronic products. This is
because local overheating caused by hotspot may fail the electronic system entirely.

The theoretical analysis of the thermal spreading resistance has been studied by many researchers.
Kennedy (1960) developed the analytical solutions of the temperature distribution within a solid
cylinder where the bottom is assumed isothermal. However, the assumption of isothermal boundary
results in an optimistic solution (Song et al. 1994). Song et al. (1994) and Lee et al. (1995)
developed a closed-form solution for the solid cylinders with the convective boundary. Their
solutions can also be applied to the rectangular plates after geometrical transformation. For multi-
layer structures, Nguyen (1996) theoretically investigated the spreading resistance by using the idea
of spreading angle. Yovanovich et al. (1999) solve the three-dimensional conduction equation to
investigate the temperature rise of the rectangular plates being adhered to small heat sources.
Muzychka et al. (2003) extended the analytical model to the multi-source cases.

One of the solutions to spread heat more uniformly is to introduce high thermal conductivity
material at the heat sink base. Ma and Peterson (2002) investigated the influence of the thermal
conductivity on the spreading effect of an aluminum plate-fin heat sink having a base area of
6.75×6.25 cm2. The simulated concentrated heat source is a dummy heater with a size of 1.5×1.5
cm2, which was smaller than the base area. For comparison purpose, they conducted an experiment
with the heater size being the same as the base area, simulating the infinite thermal conductivity.
From their experimental and numerical results, they found that the thermal performance would be
close to the case obtained with infinite thermal conductivity when the conductivity reached 2370
W/m⋅K. Instead of analyzing the whole heat sink, Borgmeyer and Ma (2005) numerically solved the
heat sink base by using heat conduction equation. They concluded that the heat sink embedded with
a flat heat pipe can have an effective thermal conductivity exceeding 6000 W/m⋅K.

As is well known that heat pipe is an effective heat transfer device and has been applied on many
industrial applications. Because the heat pipe transports energy via the liquid-vapor phase change of
the inside working fluid, it has a better heat spreading ability comparing with the metal rod of the
same size. The fundamentals and theories of the heat pipes by numerous researchers and
professionals, such as Chi (1976), Dunn and Reay (1994), and Faghri, (1995) are available in the
literatures. The basic analysis of the traditional round heat pipes are described in their publications.
Due to its excellent heat-transferring ability, the heat pipe has been applied on electronic cooling. In
addition to the traditional round configuration, the vapor chamber (flat heat pipe) is receiving lots of
attention recently. The vapor chamber allows direct attachment with the chip and can provide a
uniform temperature distribution across the heat dissipation surface. Thus, it can reduce the
spreading resistance quite efficiently from a concentrated heat source. This feature is very
imperative for the thermal management of a high flux chip. While being heated, the inside liquid
fluid evaporates and the vapor flow carries the absorbed energy through the central chamber.
Accordingly the vapor flow condenses by releasing its energy and the condensed liquid returns to
the evaporator to complete a cycle. In that regard, the vapor chamber can provide a more uniform
temperature distribution across the heat dissipation surface.

In this study, quantitative information about the spreading ability of the vapor chamber is
investigated experimentally. Influence of the size of concentrated heat source on the performance of
various kinds of heat spreader is reported, and the temperature distributions across the vapor
chamber are also observed. A plate-fin heat sink was put on the vapor chamber for the heat
dissipation. The assembled heat sink was tested in a wind tunnel to examine its thermal
performance. Additionally, two conduction plates are adapted for comparison. The experimental
results of the assembled heat sink under different heater size and frontal air velocity are presented.

2. Experimental apparatus
The plate-fin heat sinks made of aluminum used in this study are shown in Figure 1. The fin
thickness and fin height are 1 mm and 20 mm; the thickness, width, and length of the base plate are
2 mm, 63 mm and 78 mm, respectively. To measure the temperature distribution across the vapor
chamber, five T-type thermocouples were attached on the base plate (Figure 1(b)). Additionally,
one thermocouple was installed on the center of the surface between the vapor chamber and the
heater to measure the hotspot temperature.

Figure 1. The plate-fin heat sink attached on the vapor chamber (a) and its base (b).

The heat sink is assembled with the vapor chamber as shown in Figure 2. The plate-fin heat sink is
attaching on the top surface of the vapor chamber, and the heater with specific area is attached on
the bottom surface of the vapor chamber. Thermal grease with k = 4.5 W/m⋅K (Shin-Etsu G-751) is
adhered to the surfaces to reduce the contact thermal resistance. The vapor chamber used in this
study is made of copper with water as the working fluid. The total length, width and thickness of the
vapor chamber are 86 mm, 71 mm and 5 mm, respectively. The wall thickness of the vapor chamber
is 1 mm, and the height of the central core is 1.4 mm. The wick is made from sintered copper
powder with a thickness of 0.8 mm and the corresponding porosity is 0.4. For further comparison of
the spreading ability of the vapor chamber, the copper and aluminum plates having the same size as
that of vapor chamber are also adopted as the base block under the same heat sink.

Figure 2. The schematic of the heat sink assembled with the vapor chamber.
To examine the thermal performance of the assembled heat sink, it was tested in a wind tunnel
(Figure 3) which can provide a controllable air flow rate. This experimental system was comprised
of a wind tunnel, two differential pressure transducers, a power supply, a power meter and a data
acquisition unit. The heater was directly contacted with the bottom center of the vapor chamber. To
investigate the influence of the heater size on the heat spreading, three heaters with different sizes
were used. The areas of the heaters were 40×40 mm2, 20×20 mm2 and 10×10 mm2, respectively.
The area ratio Ar is defined as the ratio between the heater and the base plate, so that the according
area ratios of the heaters are 0.262, 0.065 and 0.016, respectively. The input power varied from 30
to 60 W, and the ambient temperature Tamb was kept around 27 °C. The frontal air velocity entering
the heat sink was fixed as 3, 5 or 7 m/s, and the airflow rate measurement was determined by
following the ANSI/AMCA standard 210-99 released by Air Movement and Control Association
International Incorporation (1999).

Figure 3. Schematic of the wind tunnel.

3. Data reduction
Under the steady state, the energy absorbed by the air flow is calculated by:

Qout = ( ρCp)airUAf (Tout − Tin ) (1)

The heat loss is defined as the energy not absorbed by the air flow:

Qloss ≡ Pin − Qout (2)

Hence the thermal resistance of the base block is defined as:

Te − Tc T −T
Rth = = e c for the vapor chamber; (3a)
Pin − Qloss Qout
Tbot −Ttop
Rth = for the conduction plates. (3b)
Qout

To present the temperature distribution of the condenser surface, the RMS (root mean square) of the
five measured temperatures is calculated by:

1 5
RMS= ∑
5 −1 i=1
(Ti −T )2 (4)
The uncertainty of the measurements is ±0.1 °C for the temperature measurement, ±0.9% for the
pressure difference, and ±0.3% for the input power. Based on the error analysis proposed by Kline
and McClintock (1953), the uncertainties are ±0.07 °C for calculating the root mean squares of the
condenser surface temperature, ±0.3-0.5% for the airflow rate, ±2.3 % for the energy absorbed by
the airflow, and ±2.4-8.6 % for the thermal resistance of the vapor chamber.

4. Results and discussions


With the input power of 50 W, and the heater size of 40×40 mm2, the evaporator temperature rise
( ∆Te = Te − Tamb ) and condenser temperature rise ( ∆Tc = Tc −Tamb ) are listed in Table 1. As stated
previously, five thermocouples are installed on the condenser surface to present the temperature
distribution. The average condenser surface temperature rise and its RMS calculated by using
equation (4) with the five measured temperatures are both listed in Table 1. The maximum
difference between the five measured temperatures on the condenser surface Tc,md is also list in the
table. The results with smaller heaters size of 20×20 mm2 and 10×10 mm2 are listed in Table 2 and
3, respectively.

As seen in Table 1-3, the effective overall thermal resistance across the heat spreaders increases
when the heater size is reduced. This is associated with the increase of spreading resistance. For the
same heater size, the vapor chamber shows lower thermal resistance than either copper or aluminum
plate. Since the thermal conductivity of copper (380 W/m⋅K) is higher than that of aluminum (171
W/m⋅K), the thermal resistance of the aluminum plate is the highest among the three blocks.
Additionally, the RMS values for vapor chamber are also the smallest; this is attributed to the much
faster heat transporting mechanism of the vapor chamber.

However, one should note that the ratio of thermal resistance between vapor chamber and
conduction plate (such as copper or aluminum) decreases when the heater size is reduced from
40×40 mm2 to 10×10 mm2. Despite heat in the evaporator (lower part) is effectively turned into
vapor and spreads toward the condenser (upper part), the reduced contact size between evaporator
and heater degrades the spreading effect of the evaporator wall and inevitably reduced heat flowing
along the lower part. As a consequence, one can see that the ratio of resistance between vapor
chamber and conduction plate is reduced.

Table 1. Surface temperature rises (∆T ≡ T-Tamb), heat loss and thermal resistance under different
frontal air velocity (heater: 40×40 mm2; Ar = 0.016).
Spreader U (m/s) ∆Te or ∆Tbot ∆Tc or ∆Ttop Tc,md or Qloss / Pin Rth
Block (°C) (°C) Ttop,md (°C) (%) (K/W)
Vapor 3 22.5 20.61 ± 0.38 0.9 9.0
chamber 5 17.8 15.85 ± 0.35 0.9 5.1 0.041
7 15.1 13.06 ± 0.36 0.9 3.5
Copper 3 23.2 20.64 ± 1.22 2.7 7.9
plate 5 18.7 16.07 ± 1.22 2.7 5.1 0.054
7 15.6 13.26 ± 1.20 2.3 3.4
Aluminum 3 24.8 20.50 ± 1.49 3.5 11.0
plate 5 20.3 15.97 ± 1.43 3.4 7.8 0.093
7 17.6 13.25 ± 1.41 3.4 4.3
Table 2. Surface temperature rises, heat loss and thermal resistance under different frontal air
velocity (heater: 20×20 mm2; Ar = 0.065).
Spreader U (m/s) ∆Te or ∆Tbot ∆Tc or ∆Ttop Tc,md or Qloss / Pin Rth
Block (°C) (°C) Ttop,md (°C) (%) (K/W)
Vapor 3 24.4 20.36 ± 0.41 1.1 10.0
chamber 5 19.5 15.34 ± 0.39 1.1 6.9 0.087
7 16.9 12.71 ± 0.37 1.1 1.8
Copper 3 25.1 20.48 ± 1.38 3.6 7.9
plate 5 20.6 15.89 ± 1.37 3.5 6.9 0.100
7 17.6 12.90 ± 1.32 3.4 5.9
Aluminum 3 30.0 20.48 ± 2.63 6.4 8.4
plate 5 25.5 15.94 ± 2.66 6.5 6.7 0.199
7 22.5 13.24 ± 2.62 6.4 2.3

Table 3. Surface temperature rises, heat loss and thermal resistance under different frontal air
velocity (heater: 10×10 mm2; Ar = 0.262).
Spreader U (m/s) ∆Te or ∆Tbot ∆Tc or ∆Ttop Tc,md or Qloss / Pin Rth
Block (°C) (°C) Ttop,md ( ° C) (%) (K/W)
Vapor 3 31.8 20.22 ± 0.44 1.2 9.7
chamber 5 26.5 15.24 ± 0.47 1.2 6.6 0.239
7 23.4 12.48 ± 0.49 1.2 4.2
Copper 3 31.5 19.55 ± 1.75 4.7 11.7
plate 5 26.8 15.10 ± 1.68 4.5 7.1 0.255
7 23.8 12.40 ± 1.62 4.4 6.5
Aluminum 3 35.7 20.17 ± 3.05 7.9 10.4
plate 5 30.8 15.38 ± 2.97 7.7 9.1 0.335
7 28.1 12.76 ± 2.93 7.6 6.0

Despite of this shortcoming, the temperature of the condenser (upper part) for vapor chamber shows
excellent uniform distribution when compared to the conduction plate. In addition, comparatively
better distribution along the condenser surface is seen when the size is further reduced. This is
because of the phase change phenomenon of the working fluid, by which energy transported from
the evaporator to the condenser is much faster than that of the conduction plates. Since the
temperature of the evaporated vapor maintains at the saturated temperature and the vapor flows to
the condenser, the energy absorbed from the concentrated heat source is uniformly released to the
condenser. Hence the internal flow helps to make the condenser temperature distribution more
uniform. Thus, it has a better thermal performance than the other metal plates, which transfer
energy via heat conduction only.

Since the convective effect outside the plate-fin heat sink is greatly augmented by increasing the
frontal air velocity, hence the temperature rises are reduced for higher airflow velocity. Also shown
in Table 1-3, the mean temperature of the condenser surface is relative insensitive to change of the
heater size. The RMSs of the condenser surface temperature are also increased for the cases with
smaller heater. However, the RMSs of the vapor chamber increase rather slightly when the heat size
is reduced (from 0.35 to 0.49 °C). Even for the smallest heater (10×10 mm2), the measured
maximum difference of the condenser surface is only 1.2 °C, which is much smaller than that of the
copper plate (4.7 °C) or aluminum plate (7.9 °C). In this respect, the results in Table 1-3 indicate
that the vapor chamber has the better thermal performance than the other two metal plates.
The measured thermal resistances of the aluminum plate are also compared with the available
models as shown in Figure 4. The approximate solutions of the spreading resistance were developed
by Song et al. (1994) and Lee et al. (1995). They defined the spreading resistance based on the
average or the maximum temperature of the heated surface, respectively. The three-dimensional
analytical solution for rectangular plates was developed by Yovanovich et al. (1999), and the
spreading resistance in Yovanovich’s model is based on the average temperature of the heat source.
Additionally, result by using the correlation proposed by Negus and Yovanovich (1984) for the
isoflux boundary is also presented. As seen in Figure 4, the experimental results show a good
agreement with the model prediction. It is also found that the result of the approximate solution
based on the average temperature is close to that calculated by the analytical solution, which is also
based on the temperature of the heat source. However, the prediction of Negus and Yovanovich’s
correlation shows a significant departure to the present data. This is attributed to the plate thickness
is not considered in their correlation.

Figure 4. Thermal resistance of the aluminum plate: experimental results and model predictions

Despite of the thermal resistance of the plate, the temperature distribution across the plate is also
important for thermal management of concentrated heat sources. For a fixed frontal velocity of 3
m/s, the RMSs and the maximum differences between the measured temperatures on the top surface
of the three base blocks are shown in Figure 5. As seen, the RMSs and the maximum temperature
differences of the metal plates increase almost linearly with the rise of the input power. However,
the results of the vapor chamber vary slightly with the input power. Again, this is because the inside
two-phase transport makes the condenser temperature distribution more uniform and less sensitive
to the input power.

Figure 5. The RMSs (a) and the maximum difference (b) of the condenser surface temperature.
The temperature difference between the evaporator and condenser surfaces also varies with the
input power. For the copper and aluminum plates, the temperature differences across the plate
increase linearly with the input power. Thus, the thermal resistance of the base plate defined by
equation (3b) is almost independent of the input power. As shown in Figure 6, the average thermal
resistances of the copper/aluminum plates are 0.108 and 0.169 K/W, respectively. Note that it is
well understood that the thermal resistance of copper and aluminum is fixed when the power varies.
However, the thermal resistance of the vapor chamber is reduced from 0.092 to 0.074 K/W when
the input power increases from 30 W to 60 W. The slight drop of thermal resistance at higher power
input is related to the nature of two-phase heat transport within vapor chamber. One possible
explanation is that the heat transfer coefficient is strongly related to the heat flux for the nucleate
boiling heat transfer. For example, the famous Cooper correlation (1984) indicates h ~ q0.67 for
nucleate boiling of the plain surface of pure refrigerant. Another explanation is that the high input
power results in higher evaporating flow rate of the working fluid. Therefore, the liquid film
thickness within the evaporator wick would be thinner and the thermal resistance of the evaporator
might be reduced. As a consequence, one can see a slight drop of thermal resistance of vapor
chamber when the input power is increased.

Figure 6. The thermal resistance of the three base block.

The temperature difference between condenser and ambient of the vapor chamber subject to the
influence of heater size is shown in Figure 7. With the same frontal air velocity (3 m/s) and the
same heat sink (21 plate-fins) attached, the effect of heater size on the condenser temperature is
negligible. This is because the condenser surface temperature is mainly controlled by the airflow
and the attached heat sink instead of the heater size. Additionally, the RMSs are also shown in
Figure 7. However, the RMSs (less than 0.5 °C) are comparably small than the mean temperature
rises due to the excellent spreading ability of the vapor chamber. Even for the smallest heater
(10×10 mm2, Ar = 0.016), the condenser temperature is still quite uniform.

Although the condenser temperature is not influenced by the heater size, the evaporator temperature
is strongly related to the heater size. In Figure 8, the temperature differences between the evaporator
and condenser surfaces are shown. Under the same input power, the surface temperature difference
of the smaller heater is larger than that of the larger heater. One of the reasons is because the
reduced heater area would increase the conduction thermal resistance, which is proportional to the
reciprocal of the heat transfer area. The temperature difference of the smallest heater (10×10 mm2)
are about 2.5 times than that of the middle-size heater (20×20 mm2), and are about five times than
that of the largest heater (40×40 mm2). The other explanation is due to the increase of spreading
resistance along the evaporator when the heater size is reduced. The thermal resistance of the
evaporator wall and wick will be increased for a smaller heater. However, the thermal resistance of
the condenser part would not be affected so much since the heat is released to the attached heat sink.

Figure 7. The condenser surface temperature rises of the vapor chamber while attached on the
heaters of different sizes.

Figure 8. Temperature differences between the evaporator and condenser surfaces under different
heater sizes.

5. Conclusion
The study for the heat spreading capability of the vapor chamber subject to the change of heater size
has been conducted. For comparison purpose, a copper and an aluminum plate with the same
thickness are tested as the reference heat spreading plate. The spreading plates were assembled with
a plate-fin heat sink and mounted in the wind tunnel to examine the thermal performance.

For the copper/aluminum plates, the temperature distribution across the plates becomes more
concentrated when attached on the smaller heat source. The experimental results of the conduction
plate show good agreement with the model prediction. Compared the metal plates, the condenser
temperature is much uniform for the vapor chamber even adhered to the smallest heater.
Additionally, the thermal resistance of the vapor chamber is about 15% less than the copper plate
and about 50% less than the aluminum plate. It has been also found that the thermal resistance of
the vapor chamber is slightly influenced by the heating power due to the inner complicated two-
phase phenomena. Comparing with the conduction plates, the vapor chamber can offer a lower
temperature rise and obtain a more uniform temperature distribution. Thus, the vapor chamber
provides a better choice as a heat spreader for the concentrated heat sources.

6. Acknowledgments
The authors are indebted to the financial support of the Department of Industrial Technology from
the Ministry of Economic Affairs, Taiwan.

7. References
Air Movement and Control Association International Inc. and American Society of Heating,
Refrigerating and Air Conditioning Engineers, Inc., 1999, Laboratory Method of Testing Fans for
Aerodynamic Performance Rating.
Borgmeyer, B.V. and Ma, H.B., 2005, Heat Spreading Analysis of a heat Sink Base Embedded with
a Heat Pipe, Proceedings of IPACK2005, ASME InterPACK ’05, 1-4.
Chi, S.W., 1976, Heat Pipe Theory and Practice, Hemisphere Publishing Corporation.
Cooper, M.G., 1984. Saturation nucleate pool boiling - a simple correlation, Int. Chem. Engng.
Symp. Ser. 86, 785-792.
Dunn, P.D. and Reay, D.A., 1994, Heat Pipes, 4th edition, Pergamon.
Faghri, A., 1995, Heat Pipe Science and Technology, Taylor & Francis.
Kennedy, D.P., 1960, Spreading Resistance in Cylindrical Semiconductor Devices, J. Applied
Physics, 31, 1490-1497.
Kline, S.J. and McClintock, F.A., 1953, Describing uncertainties in single sample experiments,
Mechanical Engineering, 75, 3-8.
Krane, Matthew John M., 1991, Constriction Resistance in Rectangular Bodies, ASME, J.
Electronic Packing, 113, 392-396.
Lee, S., Song, S. and Au, V., 1995, Constriction/Spreading Resistance Model for Electronic
Packaging, Proceedings of the 4th ASME/JSME thermal Engineering Joint Conference, 4, 199-
206.
Ma, H.B. and Peterson, G.P., 2002, The Influence of the Thermal Conductivity on the Heat Transfer
Performance in a Heat Sink, ASME, J. Electronic Packing, 124, 1-6.
Muzychka, Y.S., Culham, J.R. and Yovanovich, M.M., 2003, Thermal Spreading Resistance of
Eccentric Heat Sources on Rectangular Flux Channels, ASME, J. Electronic Packing, 125, 178-
185.
Negus, K.J., and Yovanovich, M.M., 1984, Constriction Resistance of Circular Flux Tubes with
Mixed Boundary Conditions by Linear Superposition of Neumann Solutions, ASME, 84-HT-84.
Nguyen, N.B., 1996, Properly Implement Thermal Spreading Will Cut Cost while Improving
Device Reliability, ISHM ’96 Proceedings, 383-388.
Song, S., Lee, S. and Au, V., 1994, Closed-Form Equation for Thermal Constriction/Spreading
Resistances with Variable Resistance Boundary Condition, Proceedings of the 1994 IEPS
Conference, 111-121.
Yovanovich, M.M., Muzychka, Y.S. and Culham, J.R., 1999, Spreading Resistance of Isoflux
Rectangles and Strips on Compound Flux Channel, J. Thermophysics and Heat Transfer, 13, 495-
500.

View publication stats

You might also like