This article was downloaded by: [Ingenta Content Distribution TandF titles]
On: 10 August 2009
Access details: Access Details: [subscription number 791939330]
Publisher Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK
Heat Transfer Engineering
Publication details, including instructions for authors and subscription information:
http://www.informaworld.com/smpp/title~content=t713723051
Evaluation of Jet Impingement, Spray and Microchannel Chip Cooling Options
for High Heat Flux Removal
Satish G. Kandlikar a; Akhilesh V. Bapat a
a
Mechanical Engineering Department, Rochester Institute of Technology, Rochester, New York, USA
Online Publication Date: 01 November 2007
To cite this Article Kandlikar, Satish G. and Bapat, Akhilesh V.(2007)'Evaluation of Jet Impingement, Spray and Microchannel Chip
Cooling Options for High Heat Flux Removal',Heat Transfer Engineering,28:11,911 — 923
To link to this Article: DOI: 10.1080/01457630701421703
URL: http://dx.doi.org/10.1080/01457630701421703
PLEASE SCROLL DOWN FOR ARTICLE
Full terms and conditions of use: http://www.informaworld.com/terms-and-conditions-of-access.pdf
This article may be used for research, teaching and private study purposes. Any substantial or
systematic reproduction, re-distribution, re-selling, loan or sub-licensing, systematic supply or
distribution in any form to anyone is expressly forbidden.
The publisher does not give any warranty express or implied or make any representation that the contents
will be complete or accurate or up to date. The accuracy of any instructions, formulae and drug doses
should be independently verified with primary sources. The publisher shall not be liable for any loss,
actions, claims, proceedings, demand or costs or damages whatsoever or howsoever caused arising directly
or indirectly in connection with or arising out of the use of this material.
Heat Transfer Engineering, 28(11):911–923, 2007
C Taylor and Francis Group, LLC
Copyright ISSN: 0145-7632 print / 1521-0537 online
DOI: 10.1080/01457630701421703
Evaluation of Jet Impingement,
Spray and Microchannel Chip Cooling
Options for High Heat Flux Removal
Downloaded By: [Ingenta Content Distribution TandF titles] At: 19:10 10 August 2009
SATISH G. KANDLIKAR and AKHILESH V. BAPAT
Mechanical Engineering Department, Rochester Institute of Technology, Rochester, New York, USA
Thermal management for high heat flux removal from microelectronic chips is gaining critical importance in many earthbased and space-based systems. Heat fluxes greater than 1 MW/m2 (100 W/cm2 ) have already been realized in high-end server
applications, while cooling needs in next generation chips and advanced systems such as high-power electronics and electrical
systems, pulsed power weapons systems, solid-state sensors, and phased-array radars are expected to reach 5–10 MW/m2
(500–1000 W/cm2 ). After evaluating the contributions from different thermal resistances in the chip-to-ambient thermal
path, this paper presents a critical review and research recommendations for three prominent contending technologies: jet
impingement, spray cooling, and microchannel heat sinks.
INTRODUCTION
Oktay et al. [1] presented an overview of various man-made
thermal systems identifying the heat fluxes and associated surface temperatures. The ballistic re-entry encountered one of
the highest heat fluxes of 10 MW/m2 (1 kW/cm2 ), but the
surface temperature was also high, about 3500 K. Their diagram is redrawn in Figure 1 to include the current and projected needs in chip-cooling application. It is seen that heat
fluxes of the same magnitude as the ballistic re-entry are projected by 2011, but the surface temperature is limited to only
350 K.
A schematic of the thermal resistance elements in a chip
thermal path is shown in Figure 2. For the present discussion,
heat generation is assumed to be uniform over the surface of
the silicon chip (junction level heat transfer is not considered).
The thermal resistances encountered are due to conduction in
the silicon, conduction through the thermal interface material
(TIM), conduction in the base material, and convection and fluid
temperature rise in the heat exchanger. A brief discussion on the
individual resistances and their impact on the cooling system
design are presented next.
Address correspondence to Professor Satish G. Kandlikar, Mechanical
Engineering Department, Rochester Institute of Technology, James E. Gleason Building, 76 Lomb Memorial Drive, Rochester, NY 14623-5603. E-mail:
[email protected]
It is clear that the thermal resistance of the TIM, the conduction resistance of the substrate, and the convective resistance
inside the microchannels are all crucial in determining the maximum heat that can be transferred for a given coolant fluid flow
rate and chip surface and fluid inlet temperatures. A separate
chip, called thermal chip, may be used for cooling purposes.
The thermal chip may incorporate microchannels or other cooling techniques and may be attached to the IC chip with a suitable
TIM.
Table 1 shows the temperature drops across the silicon chip
(thickness 0.7 mm), TIM, 1-mm thick copper heat spreader base,
and heat exchanger. The thermal resistance values used in Table
1 are based on currently available technologies for TIM and
heat exchangers. Given an allowable chip surface temperature
of 80◦ C and a coolant design inlet temperature of 35◦ C, the total
available temperature difference is only 45◦ C. As seen from
Table 1, a temperature difference of 67.2◦ C is needed at a heat
flux of 1 MW/m2 (100 W/cm2 ), rising to 672◦ C at 10 MW/m2
(1 kW/cm2 ).
To meet the cooling demand at higher heat fluxes, each thermal resistance listed needs to be reduced significantly. Such reductions may be attainable only through dramatic improvements
in the individual technologies.
At 10 MW/m2 (1 kW/cm2 ), the temperature drop across the
silicon substrate itself consumes the entire available temperature potential. Thinning the chip down to 0.2 mm will reduce
the temperature drop from 47.6◦ C to 13.6◦ C. Structural and
911
Downloaded By: [Ingenta Content Distribution TandF titles] At: 19:10 10 August 2009
912
S. G. KANDLIKAR AND A. V. BAPAT
Figure 1 Projected cooling demands in comparison to other man-made thermal systems, modified from Oktay et al. [1].
thermal expansion issues need to be addressed in making a final
decision.
Thermal interface materials have been studied extensively in
the past decade. A recent report by Samson et al. [2], summarized
in Table 2, indicates that the best available TIM in 2004 was a
Phase Change Metallic alloy that yielded a minimum thermal
resistance of 5 mm2 ◦ C/W. This is half the value from Table
1, but the resulting temperature drops across the TIM are still
unacceptably large at higher heat fluxes.
Recent advances have been reported on metallic solders with
indium and silver. One of the major problems with a solder is
void generation and lack of compliance between the two mating
interfaces. Colgan et al. [3] report the development of an indium
solder to overcome these problems; in the future, a value lower
than 3 mm2 ◦ C/W is expected. Research is also continuing regarding the mechanical properties of indium solder, to extend its
lower operating temperatures to below 0◦ C. Another major development in this field is related to the use of carbon nanotubes
that are grown on the substrates and connect the two mating interfaces. The high thermal conductivity of carbon nanotubes (3000
W/m ◦ C) coupled with direct growth on the substrate results in
a low interface thermal resistance. Xu and Fisher [4] report their
Figure 2 Schematic representation of thermal resistance elements in heat
dissipation from a silicon chip.
heat transfer engineering
work on CNT-copper interfaces and present models to represent
various thermal resistances in the composite material. Park and
Taya [5] report a new TIM composed of carbon nanotubes, silicon thermal grease, and chloroform. They report a measured
value between 2.65 and 3.99 mm2 ◦ C/W. A detailed survey of
this field is provided by Prasher [6]. With the developments in
indium solders and CNT- and nanoparticle-based composites, it
is reasonable to expect that future developments in this field will
lead to a value of 1 mm2◦ C/W in the next 3–5 years.
The heat exchanger substrate or heat spreader thickness of
1 mm in copper certainly poses a problem. A reduction in the
substrate thickness to 0.2 mm or lower is possible and may
be readily implemented. The resulting temperature drop will be
only 1.5◦ C for a heat flux of 300 W/cm2 . Direct cooling of silicon
chips is a preferred option, but it may not be acceptable due to
leakage and manufacturing concerns. Incorporating a thermal
chip that has enhancement features is another option.
The heat exchanger system used in chip cooling presents a
major challenge, as seen from Table 1. The conventional systems using compact heat exchangers are barely able to meet
the cooling requirement of 1 MW/m2 (100 W/cm2 ). It is clear
from Table 1 that at 10 MW/m2 (1 kW/cm2 ), major innovations
are needed to dramatically reduce the thermal resistance in the
heat exchanger. The three competing cooling technologies are
reviewed in the following sections.
JET IMPINGEMENT COOLING
In a jet impingement cooling system, high-speed jets issue
from nozzles and impinge on the target plate kept at some distance away from the nozzle tip. A very thin boundary layer is
formed under the jet, and very high heat transfer coefficients are
achieved in this zone. As fluid starts flowing radially outward, the
boundary layer thickens, and heat transfer is adversely affected.
This is an inherent disadvantage of using a single jet for cooling.
Uniform cooling is not possible with a single jet, and hence
multiple jets are employed. The jet streams from adjacent nozzles, however, mix with each other and interact with the exiting
fluid. The design of outlet ports becomes an important issue.
Figure 3 shows the nozzle plate arrangement of a cooling system
with multiple jets.
Figure 3
jets.
Schematic of a jet impingement cooling scheme employing multiple
vol. 28 no. 11 2007
S. G. KANDLIKAR AND A. V. BAPAT
913
Table 1 Temperature drop across various thermal resistances
q = 1 MW/m2 (100 W/cm2 )
Downloaded By: [Ingenta Content Distribution TandF titles] At: 19:10 10 August 2009
Silicon wafer, 0.7 mm
TIM, 10 mm2 ◦ C/W
Copper, 1 mm
Convective, h =
20,000 W/m2 K
Total T needed
q = 3 MW/m2 (300 W/cm2 )
Q = 10 MW/m2 (1000 W/cm2 )
4.7◦ C
10.0◦ C
2.5◦ C
14.3◦ C
30.0◦ C
7.5◦ C
47.6◦ C
100.0◦ C
25.0◦ C
50.0◦ C
67.2◦ C
150.0◦ C
186.8◦ C
500.0◦ C
672.6◦ C
Jet impingement can be further classified as submerged and
free surface jets. Submerged jets have a surrounding liquid
through which the jet flows, whereas free surface jets are surrounded by a gas. In general, free surface jets have a lower
heat transfer capability, unless evaporation is employed at the
interface.
In recent years, a number of investigators have proposed different enhancement techniques to improve the performance of
jet impingement cooling, particularly in the field of electronics
packaging. Increasing the turbulence of the jet stream, modifying the target surface by coatings, and enhancing the surface
structure are some of the commonly employed enhancement
techniques. Table 3 outlines recent investigations on jet impingement in chronological order [7–20].
Garimella and Rice [21] investigated the distribution of the
heat transfer coefficient for a submerged jet in confined and
unconfined spaces for dielectric fluid FC-72. Two peaks were
observed, one at the stagnation zone and another at some radial
distance from the center line. For smaller diameter jets, both
peaks were more pronounced than for larger diameter jets. The
secondary peak was associated with probable turbulence and
recirculation caused by the confinement due to the upper wall.
As the jet diameter increased, the peak became flatter, indicating
a broader stagnation zone. Further increase in nozzle diameter
caused the heat transfer coefficient to increase further due to the
recirculation.
Oh et al. [7] developed liquid jet cooling modules for very
high heat fluxes, those encountered in applications such as plasmas, optical beams, and semiconductor laser arrays. At very high
heat fluxes, the mechanical failure of the target plate becomes an
important issue due to the production of large thermal stresses.
Induced mechanical stress was also evaluated in this study. Dissipations as high as 17 MW/m2 (1.7 kW/cm2 ) were achieved
with single-phase water. The heater area was 1000 mm2 , and
the surface temperature was greater than 600◦ C. This dissipation rate was associated with a very high Reynolds number of
200,000 and an inlet pressure of 580 kPa.
Wen and Jang [9] investigated heat transfer between constant
heat flux plates and air jets with and without swirl inserts over
a Reynolds number range between 500 and 270,000. Figure 4
shows the helical inserts used to give a swirling pattern to the
Table 2 Thermal resistances (mm2 -C/W) for the best available TIM, adapted
and modified from Samson et al. [2]
Material type
1995
1998
2001
2004
Elastomer
Grease
PCM
PCMA
Gel
Indium solders and
CNT/nanoparticle composites
100
30
—
—
—
70
20
25
—
30
60
20
20
10
20
50
10
10
5
8
2006
2.65–4
Abbreviations: PCM = phase change material, PCMA = phase change metallic
alloy, CNT = carbon nanotubes.
heat transfer engineering
Figure 4 Nozzles illustrated with one flute and two flute inserts, redrawn
from Wen and Jang [9].
vol. 28 no. 11 2007
Downloaded By: [Ingenta Content Distribution TandF titles] At: 19:10 10 August 2009
914
S. G. KANDLIKAR AND A. V. BAPAT
flow. Symmetric vortices were formed, and their strength decreased as the spacing to jet diameter ratio increased. The vortices were yawed as a result of complex interactions between
the centrifugal and buoyancy forces in the swirling streams. The
vortices produced after the impingement were recorded. The
heat transfer on the impinged surface was highest at the radial
location where the jet flow cone impinged the surface. However,
a more uniform radial distribution of Nu was obtained without
the swirl. An important finding in this study was that the swirl
enhanced the heat transfer as a result of air-induced turbulence.
The performance enhancement over the plain nozzles was found
to be relatively modest, between 9 and 14%.
Fabbri et al. [11] carried out single-phase experiments using
water and FC-40 with jet diameters of 50–250 μm. The jet arrays
had 61–397 jets and were capable of removing more than 2.5
MW/m2 (250 W/cm2 ) from a 19.5 mm diameter copper surface,
which represented the back side of the chip. However, the surface
temperatures reached values as high as 350◦ C.
Later, Wang et al. [13] used two-phase, multiple microjet
arrays of diameters 40–76 μm to cool VLSI chips. Although
the experiments were performed in an open loop, the study was
directed as a potential technology for IC chip cooling. The wall
temperature rise was around 100◦ C.
Kanokjaruvijit and Martinez-botas [14] used an eight-byeight array of jets on a dimpled surface. The dimples were used
as turbulence promoters and tests were carried out at a Reynolds
number of 11,500. The dimples acted as tools to thin the boundary layer and enhance the heat transfer over the flat portions between dimples. The effect of cross-flow was also investigated,
with air as the working fluid. They found that the dimpled surface did not always enhance the heat transfer, and the cross-flow
of the spent air played an important role. For the minimum crossflow arrangement, impingement phenomena dominate, and the
recirculation inside the dimples was not carried away by the flow.
In case of maximum cross-flow, the impact of impingement was
carried downstream as a result of the channel flow effect, and
higher heat transfer coefficients were obtained.
Liu and Qiu [17] studied the CHF in the stagnation zone
when using a superhydrophilic surface obtained by spraying
TiO2 and irradiating it with UV rays. It gradually decreased
the contact angle with water over time and eventually reached
zero. They found that boiling incipience was greatly delayed, but
the heat transfer in the nucleate boiling region was enhanced.
A superhydrophilic surface caused an increase in the CHF by
as much as 50%. These tests were performed at a very high
Reynolds number of 40,000, and a subcooling of 80 K and heat
fluxes greater than 10 MW/m2 (1 kW/cm2 ) were realized under
certain conditions.
Natarajan and Bezama [20] developed a ceramic submerged
microjet cooling device that demonstrated a cooling capacity
of 2.5 MW/m2 (250 W/cm2 ) with a pressure drop of less than
70 kPa. The low pressure drops are vital in electronics cooling
from system reliability considerations. This device had 1,600
jets and 1,681 interstitial returns that required a very complex
design of outlet manifolds. Figure 5 gives the plain view of inlet
heat transfer engineering
Figure 5 Schematic of microjet jet arrays developed with distributed returns
redrawn from Natarajan et al. [20].
and return jets integrated with the chip. This device was tested
with a heated silicon chip. Water was used in the single-phase
mode and operated at Reynolds numbers less than 500 to limit
the pressure drop. Multilayer ceramic technology is used for
making intricate microchannels and interlayer connecting vias
inside a multilayer ceramic substrate.
Pavlova and Amitay [19] investigated synthetic jet heat transfer mechanisms. Synthetic jets are produced by applying controlled excitation to the jets by the flapping motion of a piezo
disk attached above the orifice. The synthetic jets require less
complicated setup than jet impingement. These jets are excited
by a piezo-electric sensor at 400 and 1200 Hz. Although the heat
fluxes dissipated were quite small, the investigators showed that
synthetic jets are more efficient than continuous jets in specific
localized cooling applications.
Summary of Jet Impingement Research
and Future Directions
Jet impingement heat transfer relies on high velocity jets to
remove large quantities of heat in the forced convection regime.
Under nucleate boiling mode, high flow velocities imply increased bubble departure frequency and higher heat transfer coefficients. Jets offer a good solution for localized cooling. However, for chip cooling, the surface temperature variation needs
to be limited to just a few degrees. This necessitates the use of
multiple jets. With multiple jets, the interaction of jets becomes
crucial. Hence the nozzle arrangement is important. At the microprocessor cooling level, management of the inlet and outlet
jets and manifolds becomes an even more critical issue. From
Table 3, it is evident that many investigators have focused on the
arrangements of different types of nozzles to optimize the cooling based on the surface temperature distribution. Spent fluid
management also seems to be linked with the spatial distribution of the jets. If jet impingement is to be considered as a future
cooling option for microelectronics purposes, then modeling of
the flow distribution will prove to be very useful. Because jet
vol. 28 no. 11 2007
915
2006, Mozumder et al.
[18]
2006, Pavlova and
Amitay [19]
2006, Natarajan and
Bezama [20]
2004, Silverman and
Nagler [12]
2004, Wang et al. [13]
2005, Kanokjaruvijit and
Martinez-Botas [14]
2006, Sleiti and Kapat
[15]
2006, Sung and Mudawar
[16]
2006, Liu and Qiu [17]
2003, Fabbri et al. [11]
Single phase, closed loop,
submerged jet
—, —
Water
250, 52000
1, —
Water
Two phase
Liquid jet-array cooling modules for very high fluxes
Enhancement technique/major findings
—, —, —
53, —, 1.6
<85, —, 324
—, —, 15
—, —, 8
70, —, —
100, —, —
117 (drop), 8, —
—, —, —
—, 3333, 50
Nucleate boiling characteristics on
super-hydrophilic surfaces
Investigate parameters controlling maximum heat
flux during quenching process
Investigation of mechanism of synthetic jet heat
transfer removal
Liquid microjet cooler with 1600 microjets and 1681
interstitial returns using glass ceramic material
Micromachined jets for cooling of VLSI chips
8×8 array jet impinging on staggered array of
dimples with different crossflow arrangements
Triangular and rectangular array cooling for
simulated conditions of stator surface
Hybrid device of slot JI into a microchannel
JI cooling for high-energy accelerator targets
172 (drop), 3.5, 30 Micromachined single jets on simulated silicon chip
—, —
Investigation of heat transfer with swirl inserts
—, 288, —
Practical cooling implementation with triple
technologies, Rankin cycle, JI and minichannel
heat exchanger
—, —, —
61–271 jet array effects on single-phase heat transfer
580, —, 47
p (kPa),
Q (ml/min),
V ( m/s)
—, —, 126
400, 80, 6939
—, 80, 13
—, —, 200
∼ 90,—
Air
PF-5052
Single phase, closed loop
170 (rise), —, 22528
100 (rise), —, 100
—, —, —
200, —, 1000
—, 292.5, —
80 (rise), —, 100
—, —, —
95, —, 113
600 (rise), —, 1000
T S (◦ C),
T Sub (◦ C),
A (mm2 )
—, 1200
750, —
Poly-alpha-olefin
Single phase, closed loop
90, 30000
—, —
1000, —
250, —
45, 2200
—, —
177, 72950
1700, —
q max
(W/cm2 ),
h (W/m2 K)
Water
Water
Air
Two phase, open loop
Single phase, open loop
Two phase, closed loop, free
jet
Single phase, synthetic jet
Water
Water, FC-40
Water
Air
Water
Water
Fluid
Single phase, free surface,
closed loop
Closed loop
Single-phase, submerged jet,
closed loop
Two-phase, closed loop
Single phase, open loop
Two phase, closed loop
1998, Oh et al. [7]
2001, Zhang et al. [8]
2003, Wen and Jang [9]
2003, Bintoro et al. [10]
Operating conditions
Year, author
Table 3 Recent experimental studies on jet impingement cooling
Downloaded By: [Ingenta Content Distribution TandF titles] At: 19:10 10 August 2009
916
S. G. KANDLIKAR AND A. V. BAPAT
Downloaded By: [Ingenta Content Distribution TandF titles] At: 19:10 10 August 2009
Figure 6 Illustration of spray cooling application to chip cooling, the Laboratory of Physical Sciences, University of Maryland.
impingement largely relies on high velocity flow over a heated
surface, structured surfaces are not effective in heat transfer enhancement. Swirl pattern and jet geometry, configuration, and
velocity are the parameters that need to be optimized in enhancing the performance of a jet impingement system under given
pressure drop constraints.
SPRAY COOLING
Spray cooling is being pursued as a promising technology
for electronics cooling applications. Figure 6 shows a schematic
representation of a spray cooling system using multiple sprays
to cool an array of IC chips.
Spray cooling relies on liquid droplets impinging on the
heated surface. When this happens, a very thin liquid film is
formed, and through the evaporation of this film, a large amount
of heat is transported from the surface. The evaporation of the
thin film and secondary nucleation in the film reduces the wall
superheat requirement, and hence spray cooling has the potential to remove heat at much lower surface temperatures as compared to jet impingement. When sprays are used with singlephase liquid (without evaporation, liquid film is drained away),
droplet impingement induces agitation in the liquid, causing the
enhancement.
Although spray cooling is being pursued aggressively, there
are still many unknowns due to the complex heat transfer mechanism. Many investigators in recent years have tried to understand
the underlying mechanism of heat transfer in spray cooling systems in an effort to enhance their performance. Table 4 gives a
brief summary of recent investigations [22–38].
Tilton et al. [22] investigated direct liquid spray cooling of
high heat flux electronics. Miniature nozzles were developed
that could be incorporated in the packaging. They simulated a
high heat flux multi-chip module and used FC 72/87 as working
fluids. A heat flux of 144 W/cm2 was achieved with a flow rate
of 1366 ml/min over a 645 mm2 area. High inlet pressures of up
to 930 kPa were required. They found that higher mass flow rate
result in an increased heat transfer rate.
Ortiz and Gonzalez [24] studied the effects of mass flow rate,
surface roughness, and degree of subcooling on spray cooling
heat transfer engineering
potential. In an open loop, they removed more than 5 MW/m2
(500 W/cm2 ) at 48.5 ml/min over a 122 mm2 heat transfer area.
An air-assisted spray was used. Air helps in the atomization
of droplets and also induces secondary nucleation. Inlet pressures of 861 kPa were required. It was found that heat flux
increased with higher mass flow rates as well as the surface
roughness; however, higher inlet subcooling yielded a decrease
in flux values. Subcooling was found to help in lowering the
surface temperature, but it adversely affected the heat transfer
coefficient.
Lin and Ponnappan [28] simulated the cooling of high-power
laser diodes with a multi-nozzle spray in a confined chamber
within a closed loop. Pressure drops of 70–315 kPa were measured across the nozzle. Eight miniature nozzles were used with
a swirl insert in each nozzle. Swirl caused wider spray angle and
also intensified liquid break-up. They achieved a CHF of more
than 500 W/cm2 with water as the working fluid at a volumetric flux of 0.0249 m3 /m2 -s. They were probably the first ones to
study the effects of non-condensable gases on spray cooling system performance. It was found that the non-condensable gases
adversely affect the overall heat transfer of the closed loop spray
cooling systems.
Horacek et al. [33] investigated single nozzle spray cooling
with the effects of dissolved gasses. The primary effect of noncondensable gases was an increase in saturation temperature
and hence the liquid enters in a subcooled state. They obtained
a design flow rate of 32 ml/min, with a pressure drop of 371 kPa
across the nozzle. The authors specified two types of subcooling:
one using a chiller to reduce the temperature of the liquid spray
below saturation, called thermal subcooling (TS), and the other
by gassing the chamber to increase the saturation temperature,
called gas subcooling (GS). It was seen that, in agreement with
previous research, TS increased the heat transfer as a result of
sensible heating, and there was no change in the temperature at
which CHF occurred. The result of more dissolved gases and
hence a higher GS was to increase the CHF while at the same
time increasing the temperature at which CHF occurs. It was
found that in the cases of predominant GS, even with lesser
total subcooling (TS + GS), greater heat transfer was observed.
The presence of gases causes additional single-phase convection
over the heater areas not covered by the droplets and may also
contribute to additional evaporation. Authors also suggest that
gas may cause bubbles to nucleate within the drops, spreading
the liquid over a larger area. The effect of non-condensable gases
on condenser performance degradation and the overall system
was not investigated.
Bash et al. [29] demonstrated how thermal inkjet technology
can be effectively utilized to cool a heat source with non-uniform
power density. They achieved CHF of 2.7 MW/m2 (270 W/cm2 ).
The advantage cited by the authors is that better spatial flow
variation can be achieved to match the local cooling requirement. With the thermal inkjet printing technology, an array of
nozzles is assembled with individual operable resistive heaters,
and these nozzles can be arranged and fired independently.
Also, fast response in flow rate changes is possible due to high
vol. 28 no. 11 2007
S. G. KANDLIKAR AND A. V. BAPAT
917
Table 4 Recent experimental studies on spray cooling
Year, author
System details
Fluid
q max (W/cm2 ),
h (W/m2◦ C)
1992, Tilton et al.
[22]
1996, Yang et al.
[23]
Two phase,
closed loop
Air-assisted
FC-87, FC-72
144, 2200
Water
>800–20000
∼55, 5–25, 645
∼150, 121
Air-assisted, two Water
phase, open
loop
2002, Rini et al. [25] Two phase,
FC-72
closed loop
∼500, —
120, 76, 122
80, —
68, —, 100
2002, Xia [26]
Two phase
Water
920, —
40 (rise), —, 100
2003, Cabrera and
Gonzalez [27]
Water
∼500, —
2003, Lin and
Ponnappan [28]
Two phase,
air-assisted,
open loop
Two phase,
closed loop
FC-87, FC-72,
methanol, water
2003, Bash et al.
[29]
Two phase, open Water
loop
500/490 (water/
methanol),
97800/64500
(water/
methanol)
270, —
1999, Ortiz and
Gonzalez [24]
Downloaded By: [Ingenta Content Distribution TandF titles] At: 19:10 10 August 2009
T S (◦ C),
p (kPa), Q (ml/min), Enhancement technique
V (m/s)
/principal investigation
TSub (◦ C), A(mm2 )
500, 25–78, 38
∼120,
3–10, 200
275 (inlet), —, —
240 (drop), —,
1–2
180, 34–60, 9–15
Simulated chip module
cooling
Effect of liquid and
secondary gas flow on
boiling investigated
Effects of subcooling, impact
angle, and surface
roughness
Study of bubble behavior in
saturated spray cooling of
FC-72
Piezo-electric vibrators to
atomize the droplets. Micro
nozzle arrangement.
Correlation for heat flux in
nucleate boiling regime
69–310 (drop), —,
—
8 miniature nozzles with swirl
inserts used for closed
loop, high heat flux cooling
—, 4.56, —
2005, Horacek et al. Two phase,
[33]
closed loop
2005, Coursey et al. Two phase,
[34]
closed loop
FC-72
80, —
90, 39.6, 49
—, —, —
PF-5060
131, —
80, —, 198.8
365, 96, —
2005, Fabbri et al.
[35]
Water
93, 15,000
—, —, 43.4
843 (drop), 81.56,
—
Use of thermal inkjet printer
technology for
microprocessor cooling
Microporous coating of ABM
on the heater surface
2×2 nozzle array in
conjunction with structured
surface: cubic pin fins,
pyramids, and straight fins
Investigation of effect of
efficiency of liquid usage
on CHF
Effect of dissolved gases on
single nozzle spray cooling
Copper heat sinks with
straight fins of varying fin
lengths
Comparative study of
microjets and spray nozzles
FC-72
77.8, ∼14,000
103–301 (inlet),
23.4–289, —
Detail parametric study of
nozzle array pattern effects
Polyalphaolefin
—, 1100
2004, Kim et al. [30] Two phase,
Water
air-assisted
2004, Silk et al. [31] Degassed/gassed PF-5060
chamber, two
phase, closed
loop
2004, Chen et al.
Two phase
Water
[32]
2006, Sleiti and
Kapat [15]
Single phase,
closed loop,
free jets
Two phase,
closed loop,
degassed
Single phase,
closed loop
2006, Hsieh and
Yao [38]
Two phase, open Water
loop
2005, Pautsch and
Shedd [36,37]
—, —, 281.25
930 (inlet), 1366,
—
446 (inlet),
33/17400
(water/air), —
517–861 (drop),
24–48.5, —
2.1, 500
—, —, 2500
48 (inlet), 2.4, —
156, 23956
(gassed
chamber)
86.7, —, 200
379 (drop), 200,
—
708.1, —
132, —, 100
551 (inlet), —,
13.6
60, 4800
resistor cycling frequency. At low volumetric flux, the spray is
affected by the vapor rising from the heater surface because of
the lower spray momentum. However, sustained operation with
higher volumetric fluxes was achieved at higher flow rates.
Silk et al. [31, 39] investigated spray cooling on enhanced
structures. Copper surfaces were enhanced with either straight
copper channels or cubic pin fins and pyramid structures. For
straight-finned surfaces, they found that enhancement decreased
heat transfer engineering
—, —
170 (rise), —, 22528 758 (inlet), —, —
—, 77–82, 635.04
507 (inlet), —,
31.9
Triangular and rectangular
array cooling for simulated
conditions of stator surface
Micro structured silicon
surface at low mass fluxes
with increasing fin length. The finned surface promoted nucleation at lower wall temperatures than flat surfaces. The effect was
more pronounced under two-phase conditions. The second investigation with cubic fins and pyramids also studied the effects
of degassing. The highest heat flux was obtained for gassy condition (N2 dissolved at 101kPa) with straight fins at 1.56 MW/m2
(156 W/cm2 ) and a surface temperature of 86◦ C. Dielectric liquid PF-5060 was used as the working fluid. It is interesting to
vol. 28 no. 11 2007
Downloaded By: [Ingenta Content Distribution TandF titles] At: 19:10 10 August 2009
918
S. G. KANDLIKAR AND A. V. BAPAT
Figure 7 Micro structures fabricated on silicon surfaces, redrawn from Hsieh and Yao [38].
note that the dissolved gases were responsible for enhancing the
heat transfer.
Amon et al. [40] discussed a MEMS-based droplet impingement embedded cooling device to remove 1 MW/m2 (100
W/cm2 ) from the IC chips. The project, called EDIFICE, is
designed with micronozzles that incorporate swirl nozzles and
micro-injectors for jet breakup. Also included in EDIFICE are
enhanced silicon surfaces for better heat transfer, and electrostatic microvalves for on-demand control of dielectric coolant
flow rates. Dielectric fluid HFE-7200 was used because it is expected to give better atomization, which is crucial in this case as
the operating space is very small. The cooling surface area was
25.4 × 25.4 mm2 . Irregularly shaped nozzles with hydraulic
diameters of 150 μm were used with swirl inserts and injection pressures of 30–105 kPa. Tests carried out on a prototype
removed 45 W/cm2 of heat at 33.3 g/cm2 -min.
Hsieh and Yao [38] studied the evaporation of water spray
on microstructures at low mass fluxes. The authors found
that the low Bond number of the microstructures is primarily
responsible for the heat transfer enhancement. The three
grooved surfaces, etched using DRIE technique on silicon, are
illustrated in Figure 7.
The grooves on the heater surface help to retain the liquid
film as well as spread the liquid to make it thinner and improve
the evaporative heat transfer. The effect of dissolved nitrogen
gas was found to have a negligible effect on heat transfer. It was
found that the liquid film on a microtextured silicon surface
holds the original contact line (after the drop impingement)
during most of the evaporation period. That is, the surface
tension-induced contraction is balanced by the generated
capillary forces.
Hsieh and Yao [38] observed four stages for liquid film
distribution, as shown in Figure 8. Heat transfer was found
to remain unaffected by surface structure in the flooded and
dryout regimes, as the surfaces have liquid and vapor coverings,
heat transfer engineering
respectively. However, in the thin film and partial dryout
regions, the micro-textures enhanced the heat transfer by
spreading the liquid film over a larger surface area. This is a
result of capillary forces and surface wetting for a longer time
period. It was observed that heat transfer deteriorated after
the liquid film breakup and was related to the Bond number.
Smaller Bond numbers produced better performance. Using
water as the test fluid, heat transfer around 50 W/cm2 was
achieved, with more than 80◦ C of subcooling. Heat transfer
coefficients of 4800 W/m2 ◦ C were reported.
Summary of Spray Cooling Research and Future Directions
Unlike jet impingement, spray cooling offers a much more
uniform cooling and is a more attractive option for electronics
Figure 8
Four stages of liquid distribution, redrawn from Hsieh and Yao [38].
vol. 28 no. 11 2007
Downloaded By: [Ingenta Content Distribution TandF titles] At: 19:10 10 August 2009
S. G. KANDLIKAR AND A. V. BAPAT
919
cooling. Spray cooling mainly relies on evaporative heat transfer.
Secondary nucleation, which is caused by the entrainment of air
in the thin film, also enhances the heat transfer. Sehmbey et al.
[41] gave a detailed explanation for spray cooling mechanisms.
As evaporation is the key to heat transfer, capillary forces might
play an important role in spray cooling by maintaining a layer of
liquid on the heater surface. It is not a coincidence that in Table
4, many investigators can be seen trying different surface structures. The grooves in the surface help to retain the thin layer of
fluid, and hence more heat is transferred. However, gas-assisted
cooling may not be very conducive to electronics applications.
Another way of atomizing droplets is to use pressure-assist nozzles. However, the high inlet pressures required to produce microdroplets remain a concern for the application of spray cooling
in the thermal management of electronic systems.
Figure 9 Representation of a portion of the microchannel cooler geometry
employed by Colgan et al. [3].
MICROCHANNEL COOLING
Single-Phase Flow
Microchannels are identified as channels with a minimum
channel dimension in the range of 10–200 μm [42]. The small
hydraulic diameter leads to a high heat transfer coefficient even
with laminar flow. At the same time, it leads to a very large
pressure gradient for the circulating fluid. Microchannels can be
used with single-phase flow of a liquid or with two-phase flow
under flow boiling conditions.
Microchannels provide a high heat transfer coefficient due to
the small passage dimensions employed. In addition, the area enhancement resulting from channels with high aspect ratio is also
important. For example, utilizing a microchannel heated from
three sides, 100 μm wide, 450 μm deep, with a wall thickness
of 50 μm, yields a hydraulic diameter of 163.6 μm and an area
enhancement factor of 6.67. This area enhancement is crucial in
achieving a projected surface area-based heat transfer coefficient
on the order of 500,000 W/m2 ◦ C in meeting the high heat flux
challenges of 1 kW/cm2 or higher for chip cooling application.
The large pressure gradient experienced in the small passages
of microchannels can be effectively reduced by reducing the
flow length. This is accomplished by designing the header with
multiple inlets and outlets.
Table 5 shows some of the recent publications on both singlephase flow and flow boiling studies in microchannels published
in the last two years. An earlier survey reported in [42] provides
details of the work done until 2004.
A comprehensive survey of single-phase cooling technology
with microchannels was provided by Kandlikar and Grande [43].
The main identifiable issues were high pressure gradients and
fouling concerns. The high pressure gradients can be overcome
by reducing the flow length. Multiple inlet/outlet headers will
be needed at shorter distances along the flow length. This leads
to heat transfer enhancement in the entry region.
Although the heat transfer coefficients in microchannels are
very high, still higher heat transfer coefficients are needed
heat transfer engineering
to meet the heat flux challenges of 5–10 MW/m2 (500–
1000 W/cm2 ). Significant heat transfer enhancement can be
achieved by employing the offset strip-fin geometry that is successfully used in compact heat exchangers. The enhancement
potential of this technique is truly outstanding as the entry region
lengths could be limited to only a few hundred micrometers.
Colgan et al. [3] employed the offset strip-fin geometry with
a fin length of 500 μm and a channel width between 50 and
200 μm. The heat transfer coefficient for this arrangement was
shown to be on the order of 200,000 W/m2 ◦ C [3, 44]. Further
enhancement is possible through the effective usage of the offset
strip-fin configuration. Figure 9 is a rendition of a section of the
microchannel cooler geometry employed by Colgan et al. [3].
Kosar and Peles [45] studied the effect of 243 μm pin fins
staggered in a rectangular channel with an effective hydraulic diameter of 99.5 μm. Their results indicated that the performance
of this geometry was predictable using the conventional compact heat exchanger correlations for similar geometries above a
Reynolds number of 50. Below Re = 50, some variations were
noted, but it is not clear whether these are due to higher uncertainties at lower ranges or a specific phenomenon.
Header design remains an important consideration in a microchannel cooling system. The ability to provide and remove
the cooling liquid at short distances along the microchannel flow
length poses design challenges for manifolds. The objective is
to provide uniform flow with a low pressure drop penalty in the
distributor.
Flow Boiling in Microchannels
The high heat transfer coefficient resulting from the small hydraulic diameter in microchannels, combined with the inherently
more efficient process of flow boiling, has long been expected to
provide significant performance enhancement. The early experimental work indicated unstable flow behavior in microchannel
vol. 28 no. 11 2007
920
S. G. KANDLIKAR AND A. V. BAPAT
Table 5 Recent investigations after 2004 on microchannels
Year, author
2005, Lee et al. [50]
2005, Colgan et al. [3]
System flow
conditions
Rectangular
microchannel,
single phase
Offset strip fins,
single phase
Downloaded By: [Ingenta Content Distribution TandF titles] At: 19:10 10 August 2009
2006, Steinke and Kandlikar [44] Offset strip fins,
single phase
2006, Kosar and Peles [45]
Micro pin fins,
single phase
2005, Lee and Mudawar [51, 52] Rectangular
microchannel,
two phase
2005, Ghiu and Joshi [53]
Enhanced
structured
microchannel,
two phase
2006, Lee and Mudawar [54]
Rectangular
microchannel,
two phase
2006, Kosar et al. [49]
Rectangular
microchannel,
two phase
2006, Kosar and Peles [55]
Micro pin fins,
boiling
incipience
Fluid
q max (W/cm2),
TS (◦ C),
Enhancement technique/
h (W/cm2K) TSub (◦ C), Dh (μm) p (kPa), Re principal investigation
Water
—, 23,000
—, —, 318–903
Water
300, 130000
—, —, ∼100
Water
—, —
—, —, 76–98
Water
167, 55000
—, —, 99.5
R134a
93.8, 50000
—, —, —
45, —
—, —, —
R134a
100, 40000
55, —, 200–500
Water
614, —
—, —, 227
R123
65.5, —
90, —, 99.5
PF 5060
flow boiling [49–51]. However, these issues seem to be now
resolved [44, 52].
Table 5 summarizes some of the recent work published on microchannel flow boiling [3, 44, 45, 49–55]. It is seen that there is
an increased focus on obtaining the heat transfer characteristics
as well as CHF limits. Also, further work on understanding the
fundamental mechanisms and developing new models is needed
[56, 57]. This is an open area where more experimental data are
needed. The application of enhanced geometries is also desirable, and some work in that direction has already begun in a
number of research labs all over the world.
Critical heat flux is another important consideration. The experimental data available so far cover only the lower range of
mass fluxes. Additional experimental work on CHF is needed
with plain as well as enhanced microchannels.
It is conceivable that the IC chip may be cooled directly
using any of these cooling techniques. Microchannels may be
especially suited for this application, as the fluid can be easily
contained within the cooling passages. The fabrication issues
involved, however, with integrating large-scale material removal
processes in microchannel production with the sub-micrometer
scale IC circuits may prove to be difficult to resolve from a
yield consideration. Electronics engineers may be reluctant to
subject the expensive finished IC chips to further processing. If
a non-dielectric fluid such as water is used for cooling the IC
directly, the danger of leakage and accompanying catastrophic
failure may pose an unacceptable risk. However, these options
heat transfer engineering
—, 300–3500 Investigate validity of large scale
correlations for predicting thermal
behavior at microscale
35, 26–282 Practical implementation of single
phase silicon microchannel on to a
high power chip
180, 100–200 Single phase heat transfer study of
offset strip fins
∼25, 14–112 Thermo-hydraulic analysis of flow of
water over bank of staggered pin
fins
660 (inlet), — High heat flux micro channel cooling
for refrigeration cooling
applications
—, —
Boiling performance from enhanced
single layered structure
∼120
Microchannel evaporator for
maintaining low surface
temperatures
—, —
Suppression of flow boiling
oscillations through inlet
restrictors
20, 134–314 Investigation of boiling incipience of
R123 over pin fins
need to be carefully reevaluated in light of the future high heat
flux dissipation requirements.
COMPARISON OF COOLING OPTIONS
Thermohydraulic Performance Comparison
Jet impingement technology relies mainly on the high velocity jet to provide a high heat transfer rate. A large number of
jets are needed to provide a uniform heat transfer rate over a
surface. Current performance needs to be further improved for
meeting the 10 MW/m2 (1 kW/cm2 ) target. The enhancement
methods that are being experimented with JI are swirl and jet
turbulence. The enhancement potential of these techniques is
seen to be modest as per the above review.
Spray cooling and microchannels both offer high heat transfer
performance, which can be further increased by incorporating
specific methods. In the case of spray cooling, structured surfaces are seen to be extremely promising. Further research is
expected to identify specific configurations that provide significant improvements. However, the pressures required in the spray
systems are very high. Unless effective spray patterns (droplet
size and distribution) can be obtained at relatively modest pressures (less than one atmosphere), it is rather difficult to implement spray technology in electronics cooling applications. This
is a major challenge that needs to be addressed by the research
community.
vol. 28 no. 11 2007
Downloaded By: [Ingenta Content Distribution TandF titles] At: 19:10 10 August 2009
S. G. KANDLIKAR AND A. V. BAPAT
Microchannels in single-phase mode provide a very high level
of performance by using enhanced surfaces such as offset-strip
fins. Multiple inlet/outlet headers effectively reduce the pressure drop requirement to less than 100 kPa. The optimization
of microchannel geometry is expected to provide an effective
solution to the high flux cooling requirement, with heat transfer
coefficients in excess of 500,000 W/m2 ◦ C.
Microchannels with flow boiling are expected to provide even
better performance than the single-phase mode, although this is
yet to be proven. A major obstacle is the critical heat flux limitation; working fluid poses another challenge. Water systems
are able to provide very high level of performance, but need to
be operating under vacuum conditions [58]. Refrigerant performance is generally a factor of 5–10 lower than water systems,
both in heat transfer coefficient and CHF limits. This is an open
area where considerable research efforts are needed.
Direct Cooling of IC Chips with Boiling Refrigerants
An option that has long been overlooked is providing the
refrigerant under flow boiling mode directly over the IC chips.
Such a technique, using a microgap above the IC chip, was illustrated by Kandlikar [42]. The advantages of this configuration
are the elimination of the thermal resistances of the silicon layer,
TIM, and heat spreader substrate. Although the heat transfer coefficients of such a device may be below a microchannel device,
the effective performance will be comparable when the overall
thermal resistances are taken into account. The concern regarding damage to the ICs could be easily addressed by employing
developments in inkjet printer technology such as protecting the
heater surface with a titanium oxide layer. Further research is
needed to establish the heat transfer coefficient and CHF limits
for this configuration. Once thermal performance is verified, the
system needs to be evaluated from a packaging perspective.
Structural and Leakage Considerations
Packaging issues are often overlooked in evaluating cooling
options. The pressures required by spray cooling systems are unacceptably high. System reliability becomes a major issue when
high pressures—often exceeding several atmospheres—are employed. Another issue that needs to be considered is the leakage
concern. Providing a large unsupported area for accommodating
a spray or a jet system will cause large deflections into the silicon chips. Confining the working fluid in this space has been a
major challenge. The direct cooling option with a microgap will
also face the same problems, although the pressures may not be
as high as in jet or spray systems. In this regard, microchannels provide an effective solution, as the multiple channel walls
provide more contacting points to add to the strength of the unit.
Leakproof operation has been demonstrated in such coolers
[3]. The structural redesign of the spray cooling option may
overcome some of these problems by providing multiple intermediate support structures.
heat transfer engineering
921
CONCLUDING REMARKS
A critical review of jet impingement and spray and microchannel cooling options is presented in this paper, with a
special emphasis on developments in this field over the last five
years. The thermohydraulic performance comparison indicates
that the single-phase microchannel option is most viable at this
time, and additional research on enhanced surfaces and flow
boiling systems is needed to make them more effective. Further
research in spray systems is expected to make them an attractive
alternative as well. However, pressures in spray systems need
to be reduced in order to address reliability issues. Structural
and leakage considerations again make the microchannels most
attractive at the current time. The direct cooling of IC chips
with microgaps provides an interesting alternative, but significant research efforts are needed to make it a viable option. Further research is needed in all these systems from a packaging
perspective.
NOMENCLATURE
A
h
Nu
p
q
Q
Re
Ts
Tsub
V
surface area, mm2
heat transfer coefficient, W/m2 , ◦ C
Nusselt number
pressure, kPa
heat flux, W/cm2
flow rate, ml/min
Reynolds number
surface temperature, ◦ C
subcooling degree, ◦ C
flow velocity, m/s
REFERENCES
[1] Oktay, S., Hannemann, R., and Bar-Cohen, A., High Heat From
a Small Package, Mechanical Engineering, vol. 108, no. 3,
pp. 36–42, 1986.
[2] Samson, S. C., Machiroutu, S. V., Chang, J-Y., Santos, I., Hermerding, J., Dani, A., Prasher, R., and Song, D., Interface Material Selection and a Thermal Management Technique in SecondGeneration Platforms Built on Intel Centrino Mobile Technology,
Intel Technology Journal, vol. 9, no. 1, pp. 75–86, 2005.
[3] Colgan, E. G., Furman, B., Gaynes, M., LaBianca, N., Magerlein, J. H., Polastre, R., Bezama, R., Marston, K, and Schmidt,
R., High Performance and Sub-Ambient Silicon Microchannel
Cooling, Proc. ICNMM2006 Fourth International Conference on
Nanochannels, Microchannels and Minichannels, Limerick, Ireland, June 2006.
[4] Xu, J., and Fisher, T. S., Enhanced Thermal Contact Conductance
Using Carbon Nanotube Array Interfaces, IEEE Transactions on
Components and Packing Technologies, vol. 29, no. 2, pp. 261–
267, 2006.
[5] Park, J.-J., and Taya, M., Design of Thermal Interface Material
with High Thermal Conductivity and Measurement Apparatus,
Journal of Electronic Packaging, vol. 128, pp. 46–52, 2006.
vol. 28 no. 11 2007
Downloaded By: [Ingenta Content Distribution TandF titles] At: 19:10 10 August 2009
922
S. G. KANDLIKAR AND A. V. BAPAT
[6] Prasher, R., Thermal Interface Materials: Historical Perspective,
Status and Future Directions, Proc. IEEE, vol. 94, no. 8, pp. 1571–
1586, 2006.
[7] Oh, C. H., Lienhard, J. H. V., and Younis, H. F., Liquid Jet-Array
Cooling Modules for High Heat Fluxes, AICHE Journal, vol. 44,
no. 4, pp. 769–779, 1998.
[8] Zhang, L., Wang, E. N., and Koch, J. D., Microscale Liquid Impingement Cooling, 2001 ASME International Mechanical Engineering Congress and Exposition, November 11–16, 2001, vol. 3,
pp. 135–140, 2001.
[9] Wen, M., and Jang, K., An Impingement Cooling on a Flat Surface by Using Circular Jet with Longitudinal Swirling Strips, International Journal of Heat and Mass Transfer, vol. 46, no. 24,
pp. 4657–4667, 2003.
[10] Bintoro, J. S., Akbarzadeh, A., and Mochizuki, M., A Closed
Loop Impinging Jet Cooling for Computer Chip, 2003 ASME International Mechanical Engineering Congress, November 15–21,
2003, vol. 3, pp. 349–355, 2003.
[11] Fabbri, M., Jiang, S., and Dhir, V. K., Experimental Investigation of Single-Phase Micro Jets Impingement Cooling for Electronic Applications, 2003 ASME Summer Heat Transfer Conference, New York, New York, and Las Vegas, Nevada, United States,
vol. 2003, pp. 461–468, July 21–23, 2003.
[12] Silverman, I., and Nagler, A., High Heat Flux Cooling with Water
Jet Impingement, Proc. ASME Heat Transfer/Fluids Engineering
Summer Conference 2004, HT/FED 2004, July 11–15, 2004, vol.
1, pp. 277–282, 2004.
[13] Wang, E. N., Zhang, L., and Jiang, L., Micromachined Jets
for Liquid Impingement Cooling of VLSI Chips, Journal of
Microelectromechanical Systems, vol. 13, no. 5, pp. 833–842,
2004.
[14] Kanokjaruvijit, K., and Martinez-Botas, R. F., Jet Impingement
on a Dimpled Surface with Different Crossflow Schemes, International Journal of Heat and Mass Transfer, vol. 48, no. 1, pp.
161–170, 2005.
[15] Sleiti, A. K., and Kapat, J. S., An Experimental Investigation
of Liquid Jet Impingement and Single-Phase Spray Cooling Using Polyalphaolefin, Experimental Heat Transfer, vol. 19, no. 2,
pp. 149–163, 2006.
[16] Sung, M. K., and Mudawar, I., Experimental and Numerical
Investigation of Single-Phase Heat Transfer Using a Hybrid
Jet-impingement/micro-Channel Cooling Scheme, International
Journal of Heat and Mass Transfer, vol. 49, no. 3–4, pp. 682–694,
2006.
[17] Liu, Z., and Qiu, Y., Critical Heat Flux of Steady Boiling for Water
Jet Impingement in Flat Stagnation Zone on Superhydrophilic
Surface, Journal of Heat Transfer, vol. 128, no. 7, pp. 726–729,
2006.
[18] Mozumder, A. K., Monde, M., and Woodfield, P. L., Maximum Heat Flux in Relation to Quenching of a High Temperature Surface with Liquid Jet Impingement, International Journal
of Heat and Mass Transfer, vol. 49, no. 17–18, pp. 2877–2888,
2006.
[19] Pavlova, A., and Amitay, M., Electronic Cooling Using Synthetic
Jet Impingement, Trans. ASME Journal of Heat Transfer, vol. 128,
no. 9, pp. 897–907, 2006.
[20] Natarajan, G., and Bezama, R. J., Microjet Cooler with Distributed
Returns, Heat Transfer Engineering, vol. 28, no. 8, pp. 779–787,
2007.
heat transfer engineering
[21] Garimella, S. V., and Rice, R. A., Confined and Submerged Liquid
Jet Impingement Heat Transfer, Journal of Heat Transfer, vol. 117,
no. 4, pp. 871–877, November 1995.
[22] Tilton, D. E., Tilton, C. L., and Pais, M. R., High-Flux Spray
Cooling in a Simulated Multichip Module, 28th National Heat
Transfer Conference and Exhibition, August 9–12 1992, vol. 2062, pp. 73–79, 1992.
[23] Yang, J., Chow, L. C., and Pais, M. R., Nucleate Boiling Heat
Transfer in Spray Cooling, Journal of Heat Transfer, vol. 118, no.
3, pp. 668–71, 1996.
[24] Ortiz, L., and Gonzalez, J. E., Experiments on Steady-State High
Heat Fluxes Using Spray Cooling, Experimental Heat Transfer,
vol. 12, no. 3, pp. 215–233, 1999.
[25] Rini, D. P., Chen, R., and Chow, L. C., Bubble Behavior and
Nucleate Boiling Heat Transfer in Saturated FC-72 Spray Cooling,
Journal of Heat Transfer, vol. 124, no. 1, pp. 63–72, 2002.
[26] Xia, C., Spray/Jet Cooling for Heat Flux High to 1 kw/cm2 , Proc.
2002 18th Annual IEEE Semiconductor Thermal Measurement
and Management Symposium, March 12–14, 2002, pp. 159–163,
2002.
[27] Cabrera, E., and Gonzalez, J. E., Heat Flux Correlation for Spray
Cooling in the Nucleate Boiling Regime, Experimental Heat
Transfer, vol. 16, no. 1, pp. 19–44, 2003.
[28] Lin, L., and Ponnappan, R., Heat Transfer Characteristics of Spray
Cooling in a Closed Loop, International Journal of Heat and Mass
Transfer, vol. 46, no. 20, pp. 3737–3746, 2003.
[29] Bash, C. E., Patel, C. D., and Sharma, R. K., Inkjet Assisted Spray
Cooling of Electronics, 2003 International Electronic Packaging
Technical Conference and Exhibition, July 6–11, 2003, vol. 2,
pp. 119–127, 2003.
[30] Kim, J. H., You, S. M., and Choi, S. U. S., Evaporative Spray
Cooling of Plain and Microporous Coated Surfaces, International
Journal of Heat and Mass Transfer, vol. 47, no. 14–16, pp. 3307–
3315, 2004.
[31] Silk, E. A., Kim, J., and Kiger, K., Investigation of Enhanced Surface Spray Cooling, 2004 ASME International Mechanical Engineering Congress and Exposition, IMECE, November 13–19,
2004, vol. 375, pp. 685–690, 2004.
[32] Chen, R., Chow, L. C., and Navedo, J. E., Optimal Spray Characteristics in Water Spray Cooling, International Journal of Heat
and Mass Transfer, vol. 47, no. 23, pp. 5095–5099, 2004.
[33] Horacek, B., Kiger, K. T., and Kim, J., Single Nozzle Spray Cooling Heat Transfer Mechanisms, International Journal of Heat and
Mass Transfer, vol. 48, no. 8, pp. 1425–1438, 2005.
[34] Coursey, J. S., Kim, J., and Kiger, K. T., Spray Cooling of SmallPitched, Straight-Finned, Copper Heat Sinks, 2005 ASME International Mechanical Engineering Congress and Exposition,
IMECE 2005, November 5–11, 2005, vol. 376 HTD, pp. 271–277,
2005.
[35] Fabbri, M., Jiang, S., and Dhir, V. K., A Comparative Study of
Cooling of High Power Density Electronics Using Sprays and
Microjets, Journal of Heat Transfer, vol. 127, no. 1, pp. 38–48,
2005.
[36] Pautsch, A. G., and Shedd, T. A., Spray Impingement Cooling with
Single- and Multiple-Nozzle Arrays, Part I: Heat Transfer Data
Using FC-72, International Journal of Heat and Mass Transfer,
vol. 48, no. 15, pp. 3167–3175, 2005.
[37] Pautsch, A. G., and Shedd, T. A., Spray Impingement Cooling
with Single- and Multiple-Nozzle Arrays, Part II: Visualization
vol. 28 no. 11 2007
S. G. KANDLIKAR AND A. V. BAPAT
[38]
[39]
Downloaded By: [Ingenta Content Distribution TandF titles] At: 19:10 10 August 2009
[40]
[41]
[42]
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]
and Empirical Models, International Journal of Heat and Mass
Transfer, vol. 48, no. 15, pp. 3176–3184, 2005.
Hsieh, C., and Yao, S., Evaporative Heat Transfer Characteristics
of a Water Spray on Micro-Structured Silicon Surfaces, International Journal of Heat and Mass Transfer, vol. 49, no. 5–6, pp.
962–974, 2006.
Silk, E. A., Kim, J., and Kiger, K., Spray Cooling Trajectory Angle
Impact upon Heat Flux Using a Straight Finned Enhanced Surface,
2005 ASME Summer Heat Transfer Conference, HT 2005, July
17–22, 2005, vol. 4, pp. 743–751, 2005.
Amon, C. H., Murthy, J., Yao, S. C., Narumanchi, S., Wu, C.-F.,
and Hsieh, C.-C., MEMS-Enabled Thermal Management of HighHeat-Flux Devices EDIFICE: Embedded Droplet Impingement
for Integrated Cooling of Electronics, Experimental Thermal and
Fluid Science, vol. 25, no. 5, pp. 231–242, November 2001.
Sehmbey, M. S., Chow, L. C., and Pais, M. R., High Heat Flux
Spray Cooling: A Review, Proc. 1994 International Mechanical
Engineering Congress and Exposition, November 6–11, 1994, vol.
301, pp. 39–46, 1994.
Kandlikar, S. G., High Flux Heat Removal with Microchannels—
A Roadmap of Challenges and Opportunities, Heat Transfer Engineering, vol. 26, no. 8, pp. 5–14, 2005.
Kandlikar, S. G., and Grande, W. J., Evaluation of Single
Phase Flow in Microchannels for High Flux Chip Cooling—
Thermohydraulic Performance Enhancement and Fabrication
Technology, Heat Transfer Engineering, vol. 24, no. 1, pp. 3–17,
2003.
Steinke, M., and Kandlikar, S. G., Single-Phase Liquid Heat Transfer in Plain and Enhanced Microchannels, Proc. ICNMM2006
Fourth International Conference on Nanochannels, Microchannels and Minichannels, Limerick, Ireland, June 2006.
Kosar, A., and Peles, Y., Thermal-Hydraulic Performance of
MEMS-Based Pin Fin Heat Sink, Journal of Heat Transfer,
vol. 128, no. 2, pp. 121–131, 2006.
Cornwell, K., and Kew, P. A., Boiling in Small Parallel Channels,
Proc. CEC Conference on Energy Efficiency in Process Technology, Athens, Paper 22, Elsevier Applied Sciences, pp. 624–638,
October 1992.
Hetsroni, G., Segal, Z., and Mosyak, A., Nonuniform Temperature
Distribution in Electronic Devices Cooled by Flow in Parallel Microchannels, IEEE Transactions on Components and Packaging
Technologies, vol. 24, no. 1, pp. 16–23, March 2001.
Kandlikar, S. G., Fundamental Issues Related to Flow Boiling
in Minichannels and Microchannels, Experimental Thermal and
Fluid Science, vol. 26, no. 2–4, pp. 389–407, 2002.
Kosar, A., Kuo, C., and Peles, Y., Suppression of Boiling Flow Oscillations in Parallel Microchannels by Inlet Restrictors, Journal
of Heat Transfer, vol. 128, no. 3, pp. 251–260, 2006.
Lee, P., Garimella, S. V., and Liu, D., Investigation of Heat Transfer
in Rectangular Microchannels, International Journal of Heat and
Mass Transfer, vol. 48, no. 9, pp. 1688–1704, 2005.
heat transfer engineering
923
[51] Lee, J., and Mudawar, I., Two-Phase Flow in High-Heat-Flux
Micro-Channel Heat Sink for Refrigeration Cooling Applications,
Part I: Pressure Drop Characteristics, International Journal of
Heat and Mass Transfer, vol. 48, no. 5, pp. 928–940, 2005.
[52] Lee, J., and Mudawar, I., Two-Phase Flow in High-Heat-Flux
Micro-Channel Heat Sink for Refrigeration Cooling Applications,
Part II: Heat Transfer Characteristics, International Journal of
Heat and Mass Transfer, vol. 48, no. 5, pp. 941–955, 2005.
[53] Ghiu, C., and Joshi, Y. K., Boiling Performance of Single-Layered
Enhanced Structures, Journal of Heat Transfer, vol. 127, no. 7,
pp. 675–683, 2005.
[54] Lee, J., and Mudawar, I., Implementation of Microchannel Evaporator for High-Heat-Flux Refrigeration Cooling Applications,
Trans. ASME Journal of Electronic Packaging, vol. 128, no. 1,
pp. 30–37, 2006.
[55] Kosar, A., and Peles, Y., Convective Flow of Refrigerant (R-123)
across a Bank of Micro Pin Fins, International Journal of Heat
and Mass Transfer, vol. 49, no. 17–18, pp. 3142–3155, 2006.
[56] Kandlikar, S. G., Heat Transfer Mechanisms during Flow Boiling in Microchannels, Journal of Heat Transfer, vol. 126, no. 1,
pp. 8–16, 2004.
[57] Jacobi, A. M., and Thome, J. R., Heat Transfer Model for Evaporation of Elongated Bubble Flows in Microchannels, Journal of
Heat Transfer, vol. 124, no. 6, pp. 1131–1136, 2002.
[58] Pokharna, H., Masahiro, K., DiStefanio, E., Mongia, R., Crowley,
B. J., Chen, W., and Izenson, M., Microchannel Cooling in
Computing Platforms: Performance Needs and Challenges in
Implementation, Keynote Paper, ASME Paper No. ICMM20042325, Second International Conference on Microchannels and
Minichannels, June 17–19, 2004, Rochester, New York, pp. 109–
118, 2004.
Satish Kandlikar is the Gleason Professor of Mechanical Engineering at RIT. He received his Ph.D.
from the Indian Institute of Technology in Bombay
in 1975 and has been a faculty member there before
coming to RIT in 1980. His research is mainly focused
in the area of flow boiling, microchannels, roughness
effects, and water management in fuel cells. He has
published more than 175 journal and conference papers. He is a fellow member of ASME and has been
the organizer of the international conferences on microchannels and minichannels sponsored by ASME.
Akhilesh Bapat is a M.S. student at the Rochester
Institute of Technology in New York. He completed
his Bachelor’s degree in mechanical engineering from
Government College of Engineering, Pune, India, in
2004, after which he worked for a year in Thermax India Ltd as a graduate engineer. He joined the Thermal
Analysis Lab at RIT in 2005 and is currently working
on single- and two-phase chip cooling.
vol. 28 no. 11 2007