flow boiling on micropin fins entrenched inside a ... · flow boiling on micropin fins entrenched...

1

do�dtamih

sp�nhfnfltlcciCoata

ifip�h

N

cS

J

Downloa

Santosh Krishnamurthy

Yoav Peles

Department of Mechanical, Aerospace, andNuclear Engineering,

Rensselaer Polytechnic Institute,Troy, NY 12180

Flow Boiling on Micropin FinsEntrenched Inside aMicrochannel—Flow Patternsand Bubble Departure Diameterand Bubble FrequencyFlow boiling of HFE 7000 in five parallel microchannels of 222 �m hydraulic diameter,each containing a single row of 24 in-line 100 �m pin fins, was investigated. High speedphotography revealed the dominant flow patterns, namely, the bubbly flow, the multipleflow, and the wavy-annular flow. The interaction of the bubble with the pin fins duringnucleate boiling from G�350 kg /m2 s to G�827 kg /m2 s and wall heat fluxes from10 W /cm2 to 110 W /cm2 is detailed. �DOI: 10.1115/1.2994718�

Keywords: flow boiling, microchannel, micro-pin fin, bubble departure frequency, bubbledeparture diameter, surface tension

IntroductionFlow boiling in microchannels has been extensively studied

uring the past several years because of its importance in a varietyf applications, such as electronic cooling �1–13�, microrockets14�, inkjet printers �15�, and microchemical reactors �16�. As airect extension of this research endeavor, new flow configura-ions such as micropin fins entrenched inside microchannels havelso been studied for heat transfer enhancement �17–25�. Funda-ental understanding of the boiling process in these microsystems

s important to improve the design and to enhance their thermal-ydraulic performance.

Cross flow boiling in conventional scale has been extensivelytudied, and a significant scientific body of knowledge about thehysical processes governing heat transfer in these systems exists26–37�. The general consensus is that the heat transfer mecha-isms are closely linked to flow morphologies and the resultingydrodynamics. Numerous correlations and models for heat trans-er coefficient have been developed by coupling the hydrody-amic and thermal characteristics. Hwang and Yao �32� studiedow pattern interaction with tubes for three different configura-

ions, namely, single tube, single heated tube in an unheated in-ine tube bundle, and heated in-line tube bundle. For each of theseonditions, the onset of nucleate boiling �ONB� occurred at theylinders rear wake, and the cylinder spacings affected the bubblenteractions at high qualities. Using high speed photography,ornwell �29� observed turbulent bubbly flow and attributed thebserved heat transfer enhancement to the sliding bubbles. Fand etl. �38� stated that the maximum nucleation site density occurs athe frontal stagnation point and at the rear wake side for flowcross a single cylinder.

The flow characteristics around a tube/cylinder in a tube bundles strongly dependent on geometrical parameters such as tube con-guration �in-line or staggered�, longitudinal and transverseitches, and physical parameter such as the Reynolds number39�. In microscale, the studies pertaining to bubble dynamicsave been restricted to plain microchannels �40� and channels

Contributed by the Heat Transfer Division of ASME for publication in the JOUR-

AL OF HEAT TRANSFER. Manuscript received January 18, 2008; final manuscript re-eived August 12, 2008; published online February 17, 2010. Review conducted by
atish G. Kandlikar.
ournal of Heat Transfer Copyright © 20

ded 19 May 2010 to 128.113.217.15. Redistribution subject to ASM

with re-entrant cavities �41�. The growth and the dynamics of thebubbles have been observed to deviate considerably from thoseobserved in conventional scale. The small length scale of thechannel restricted the expansion of the bubble in the transversedirection and subjected it to high shear stress along the longitudi-nal direction, eventually causing it to depart at diameters compa-rable to the channel hydraulic diameter.

In the current study, flow boiling across a single row of in-linemicropin fins entrenched in a microchannel is being investigated.The microdevice consists of five 200 �m wide and 243 �m deepmicrochannels, each equipped with an inlet orifice, consisting of24 columns of 100 �m diameter circular pin fins, with pitch-to-diameter ratio of 4. With the aid of high speed photography, thedominant flow patterns in such flow configurations were revealed.Furthermore, the interaction of bubbles with these micropin finswas also investigated and pertinent bubble dynamics parameterssuch as bubble departure diameter and frequency were also mea-sured.

2 Device OverviewA computer aided design �CAD� schematic of the microdevice,

consisting of five 200 �m wide and 243 �m deep microchannelsentrenched in a 1800 �m wide channel, is shown in Fig. 1. Eachmicrochannel encompassed 24 rows of in-line 100 �m diametermicropin fins with a pitch-to-diameter ratio of 4. A micro-orifice,400 �m long and 20 �m wide, was fabricated upstream of eachmicrochannel. A heater was deposited on the back side of thechannel-pin fin section excluding the orifice to provide the requi-site heat flux. A Pyrex cover sealed the device from the top andallowed flow visualization. For additional details on the process-ing steps and experimental setup, the reader is referred to Ref.�18�.

3 Data ReductionThe voltage, current, and pressure measurements were used to

obtain the average single-phase and two-phase temperatures andmass qualities. This section gives the data reduction procedureused to obtain these parameters.

The voltage and current values were used to calculate the input
power and heater resistance. The wall heat flux was calculated as
APRIL 2010, Vol. 132 / 041002-110 by ASME

E license or copyright; see http://www.asme.org/terms/Terms_Use.cfm

Tcbb

Ts

Ttfpd

w“rwtm

ow

0

Downloa

qw� =P − Qloss

Ap�1�

he average heater temperature was obtained from the calibrationurve. Assuming 1D steady state conduction through the siliconlock, the average surface temperature of the device was obtainedy

Tsur = Theater −�P − Qloss�ts

ksAp�2�

he local quality was calculated from the known mass flow rate ashown below:

x =�P − Qloss��Lx/Lo� − mcp�Tsat − T�i

mhfg

�3�

he fluid temperature was measured in the inlet using a type Khermocouple. The local flow temperature was deduced indirectlyrom energy balances �Tmx=Tmi+ ��P−Qloss / mCp��Lx /L��. Com-arison of the experimental data with the existing models wasone through the mean absolute error �MAE�:

MAE =1

M �i=1

M��exp − �pred�

�exp� 100 �4�

here � is the measured physical quantity and the subscriptsexp” and “pred” refer to the experimental and predicted values,espectively. The uncertainty associated with the measured valuesas obtained from the manufacturers’ specification sheets while

he uncertainties associated with the derived quantities such as

Fig. 1 Device overview sh

ass flux, surface temperature, and wall heat flux were obtained

41002-2 / Vol. 132, APRIL 2010


by using the propagation of uncertainty analysis and were 3.4%,0.5°C, and 3.8%, respectively.

4 Results and Discussion

4.1 Flow Patterns. The flow patterns observed in the currentstudy were classified as bubbly flow, multiple flows, and wavy-annular flow, as schematically shown in Fig. 2. Bubbly flow com-prised of vapor bubbles, nucleating from the sidewalls and themicropin fins, dispersed in the liquid. As the bubbles moveddownstream, they expanded and formed vapor slugs, which weresheared by the pin fins resulting in regions of vapor slugs andbubbles, termed multiple flow. Eventually, wavy-annular flow wasestablished downstream, which was characterized by the presenceof a thin wavy liquid film flowing along the sidewalls and the pinfins with a vapor core �Fig. 3�. The thin liquid film on the pin finsaccelerated along its frontal side and separated from the surfaceforming a trailing liquid drop attached to the rear side where in-tense mixing was observed. As the thin liquid film on the frontalsection evaporated, causing local dry-out, a vapor cavity devel-oped, which moved axially in the upstream direction and col-lapsed intermittently due to the drag forces exerted by the incom-ing fluid. With increasing heat flux, the vapor cavity stabilized,

ing the device dimensions

Fig. 2 Schematic of flow patterns

Transactions of the ASME


ao==CflltdswbtafiTstsFvfiN

efr

Ffi

J

Downloa

nd the thin film completely evaporated while a liquid drop trailedn the rear side �q�=39 W /cm2 for G=350 kg /m2 s; q�76 W /cm2 for G=565 kg /m2 s; q�=98 W /cm2 for G827 kg /m2 s�. Similar flow patterns have also been observed byognata et al. �20� for a square micropin fin. Occasionally, the

rontal region of the pin fins was wetted by a thin liquid micro-ayer, which evaporated rapidly resulting in a nonuniform thermaloading on the pin fin. As the heat flux was further increased, therailing liquid drop shrunk, ultimately leading to a complete localry-out. For the lowest mass velocity �G=350 kg /m2 s�, theingle-phase liquid flow directly transitioned to multiple flowithout first forming bubbly flow region. The absence of isolatedubbly flow at low mass fluxes is, to some respect, in contradic-ion to the trend discussed by Revellin and Thome �42�. Thisnomaly might be a result of the significant different flow con-gurations of the current study and the study of Revellin andhome �42�. Due to the low mass velocity, the flow exhibitedlight flow reversal in which a vapor bubble �cavity� initiated athe frontal stagnation point and continued to grow conically up-tream, temporarily increasing the local void fraction, as shown inig. 4. This was accompanied by an increase in the local liquidelocity, which deformed the cavity and sheared it across the pinns causing it to break and move downstream �0� t�4.4 ms�.o such flow reversal was observed at higher mass velocities.

4.2 Heat Flux Versus Temperature. Figure 5 shows the av-

rage Tsur−qw� curves for different mass fluxes. The heat flux as aunction of the average temperature was characterized by threeegions of different slopes. The first region corresponds to single-

ig. 3 Wavy-annular flow showing the trailing liquid of the pinns and vapor cavity on the upstream side of pin fin

Fig. 4 Sequence of images showing the formati
sequent breakage „G=350 kg/m2s, qw� =33.2 W/cm
ournal of Heat Transfer


phase flow where the temperature was linearly dependent on theheat flux. The onset of boiling was accompanied by a significantdrop in the surface temperature, which was observed for all massfluxes and was attributed to boiling hysteresis. Such temperaturehysteresis at the onset of boiling was caused by the good wetta-bility of the fluid and the surface topography. The low surfacetension of HFE 7000 ��=13.8 mN /m at T=22 deg� and the sur-face roughness of the silicon surface ��0.3 �m� render majorityof the cavities inactive even at high superheated temperatures.Ultimately, boiling was triggered and vapor bursted through thechannel activating many nucleation sites and significantly reduc-ing the surface temperature. Following the onset of boiling, thetemperature continued to increase. At a certain heat flux, the tem-perature increased rapidly signifying the approaching critical heatflux �CHF� conditions. The hysteresis phenomenon clearly ap-peared when the heat flux was gradually decreased, as shown bycurve b in Fig. 5.

4.3 Bubble Nucleation and Dynamics. At the onset of boil-ing, the bubbles nucleated on the microchannel sidewalls and onthe pin fin surface. Flow visualization revealed that the bubblesfirst nucleated at the rear wake side of the first pin fin in thetwo-phase region, as shown in Fig. 6. The thick thermal boundarylayer and the resulting high local fluid temperature initially trig-gered the bubble growth in this region. Figure 7 shows the distri-bution of the active cavities along the circumference of the pin finfor different mass velocities. The majority of the nucleation siteswere distributed asymmetrically on the rear side at 90 deg��180 deg. It has been observed that for flow across tandem cyl-inders, the flow transitions from steady symmetrical flow at lowReynolds number to an oscillatory asymmetrical flow at a criticalReynolds number; for p /d=4, the transition corresponds to a criti-cal Reynolds number of Rec=68 �43�. It is very likely that suchoscillations existed in the current study �75�Red�180� distort-ing the symmetry of the nucleation sites �Fig. 7�. The nucleationof the bubbles on the rear side modified the hydrodynamics of theflow field, which resulted in bubbles nucleating along differentlocations on the downstream pin fin. Generally, bubbles from pinfins downstream the first row of the two-phase region tended to

of vapor cavity at the stagnation point and sub-2

on
, and z=4 mm…
APRIL 2010, Vol. 132 / 041002-3


nif=rnAwfi

rfa

Fodfl

0

Downloa

ucleate from the frontal stagnation point ��=0 deg�. The dynam-cs of the bubble departing from different sites along the circum-erence of the pin fins also differed from each other. At �180 deg, the bubbles grew asymmetrically due to the intensive

ecirculation of the flow and at some point the bubble ultimatelyecked close to the surface of the pin fin and eventually detached.t �=0 deg, the bubbles detached and slid along the surfacehile growing and ultimately lifted from the rear side of the pinn �Fig. 8�. Additionally, the bubbles departing from the upstream

Fig. 5 Heat flux as a function of average su

ig. 6 Images showing the bubble nucleation on the rear sidef the upstream pin fins and along the circumference on theownstream pin fins for G=565 kg/m2 s under different heat
uxes
41002-4 / Vol. 132, APRIL 2010


pin fins were often swept into the wake of the downstream pin fincausing them to circulate in the recirculation zone, expand, andcoalesce with other bubbles, and in the process agitate the liquidin the recirculation zone.

4.4 Bubble Departure Frequency and DepartureDiameter. Qualitatively, the flow patterns and bubble ebullitionprocess were similar for all channels; the quantitative results wereelucidated for channel 3 �middle channel�. Figure 9�a� shows thebubble departure diameter for �=0 deg as a function of wall heatflux for different mass velocities. The bubble departure diameterdecreased with increasing heat flux and mass velocity. At highmass velocity, the increased drag force sheared the bubbles fromtheir nucleation sites at small diameters. The departure diameterdecreased with increasing heat flux at low mass velocities, but athigh mass velocities, the growth of the bubble was suppressed bythe impingement of the flow at the stagnation point. The bubblesdeparting from �=180 deg had much larger diameter than thosedeparting from �=0 deg. However, the decreasing diametertrends were similar �Fig. 9�b��. The bubble departure frequencywas elucidated by counting the number of bubbles departing froma nucleation site over 800 frames �fps of 12,000�. Figure 10 showsthe bubble departure frequency from the two most active nucle-ation sites on the pin fins, at �=0 deg and �=180 deg, at variousmass velocities as a function of wall heat flux. The bubbles nucle-ated at the location of ONB for �=180 deg and at the immediatedownstream pin fin for �=0 deg. The frequency and bubble de-parture diameter were measured in the isolated bubble region be-

ce temperature for various mass velocities

tween z=1 mm and z=3 mm from the inlet of the channel where



tfgobhstidel

F�=

J

Downloa

he bubble coalescence was negligible. For �=0 deg, the bubblerequency increased linearly with the wall heat flux, although re-ions of two distinct slopes were observed, which was a functionf the mass velocity. At low and intermediate mass velocities, theubble frequency increased sharply with the wall heat flux, and atigh mass velocities, the bubble frequency increased with a lowerlope. During the bubble growth, interfacial tension acting alonghe contact line tended to hold the bubble in place against thenertial drag force exerted by the surrounding liquid. The bubbleeparted from the site when the net effect of the inertial dragxceeded the interfacial tension. At low mass velocities, due to theower drag force exerted by the liquid on the bubble, the bubble

Fig. 7 Circumferential nucleation site distribution along

ig. 8 Sequence of images showing the bubbles nucleating at=0 deg and sliding along the circumference of pin fins „qw�

2 2
52.6 W/cm and G=827 kg/m s…


attained a large diameter before departing, resulting in a longbubble ebullition cycle and hence low frequencies. At high massvelocities, the bubbles were subjected to higher inertial drag caus-ing them to depart at a higher rate. Such mass flux dependency onfrequency has also been observed by Lee et al. �40� for flow in arectangular 41 �m hydraulic diameter microchannel. Vigorousboiling occurred at higher heat flux, which resulted in rapidbubble growth and thus higher frequencies. Additionally with in-creasing heat flux, the surface tension decreased �due to tempera-ture increase�, causing a reduction in the holding surface tensionforce and a reduction in the departure diameter, which in turnenhanced the bubble departure frequencies. Therefore, the bubblefrequency at �=0 deg has been correlated as a function of wallheat flux by the following relations with a MAE of 18%:

f = 386.6qch� − 4612.6 �Red � 120� �Hz� �5�and

f = 124.9qch� − 1363.8 �Red � 120� �Hz� �6�

The much larger departure diameter at �=180 deg compared with�=0 deg has important implications on the bubble frequency. Ascan be seen in Fig. 10�b� for �=180 deg, the bubble departurefrequency was much smaller compared with the frequency at thefrontal stagnation point. The large bubble departure diameter,caused by the low liquid velocity in the recirculation zone wherethe drag forces exerted on the bubble, was much weaker than for�=0 deg, increased the ebullition period. The frequency in-

e first pin fin to experience boiling various mass fluxes
th
creased with mass velocity and showed a peak with respect to heat

APRIL 2010, Vol. 132 / 041002-5


fl=gcidrdibtiri�

0

Downloa

ux, which was clearly evident for G=827 kg /m2 s and G565 kg /m2 s. With increasing heat flux, the boiling front propa-ated upstream and the bubbles nucleating on the rear side wereontinuously subjected to condensation on the periphery due to anncrease in the local subcooled temperature. Because of the con-ensation on the bubble periphery and the low velocity in theecirculation zone, the bubble, after nucleating and growing to theeparture diameter, tended to be stationary before departing. Thisn turn extended the bubble ebullition cycle and decreased theubble departure frequencies. Figure 11�a� shows the variation ofhe bubble departure frequency as a function of the local subcool-ng for bubbles departing from the rear side. Such variation withespect to local subcooling was not observed for bubbles nucleat-ng from the frontal stagnation point of the downstream pin fins�=0 deg�, as shown in Fig. 11�b�. Because the bubbles were

Fig. 9 Bubble departure diameter

Fig. 10 Bubble departure frequency f

41002-6 / Vol. 132, APRIL 2010


very small at �=0 deg when they departed �Dd�10 �m�, theirperiphery was not subjected to significant condensation. There-fore, higher subcooled temperatures did not seem to slow downthe bubble departure diameter. The bubble departure frequencieson the pin fins, especially at �=0 deg, are higher than those ob-served in other microscale studies �40,41� �f �6000 Hz at highheat flux�. For example, Kuo et al. �41� reported a maximumfrequency of 600 Hz and Lee et al. �40� reported a maximum afrequency of 60 Hz for boiling in microchannel using water. Thehigher frequency observed in the current study is a result of themuch lower surface tension of the working fluid. Since the surfacetension forces are the main force resisting the bubble departure,bubbles in HFE 7000—fluid with very low surface tension—readily departed from the nucleation site. Furthermore, the lower

r „a… �=0 deg and „b… �=180 deg
fo
or „a… �=0 deg and „b… �=180 deg



lfe

twasrrmqlctnctcadt

Fw

J

Downloa

atent heat of vaporization of HFE 7100 compared with wateracilitated rapid growth of the bubble �44� hastening the bubblebullition cycle.

Traditionally, the bubble frequency has been correlated throughhe bubble departure diameter according to fnDd=const �45–47�,here n ranges from 1/2 to 2. Figure 12 compares the frequency

nd bubble departure diameter for �=0 deg with conventionalcale �45,48� and microchannel �41� correlations. None of the cor-elations agreed well with the experimental results over the entireange. Conventional scale correlations provided a better agree-ent with the experimental results at high heat fluxes �high fre-

uencies�, while at low heat fluxes, which was characterized byow bubble departure frequency, agreement with microchannelorrelation improved. The above mentioned result indicates thathere exists a partwise variation of the product fDd in microchan-els; when the bubble departure diameter is comparable to thehannel hydraulic diameter, the frequency has a different func-ional dependency on the departure diameter than for conditionsorresponding to bubbles much smaller than the channel. Kuo etl. �41� investigated bubble dynamics in water for a 222 �m hy-raulic diameter rectangular microchannel with re-entrant cavi-ies. The higher surface tension of water resulted in much larger

Fig. 11 The variation of departure frequency as a func

ig. 12 Comparison of frequency and departure diameter data
ith different correlations


bubble than for the current study ranging from 100 �m to222 �m. The large bubble departure diameter-to-channel hydrau-lic diameter ratio modified the hydrodynamics of the flow consid-erably, decreasing the product �fDd� compared with conventionalscale systems. In the current study, for low mass velocities andheat fluxes, characterized by large bubble departure diameters,microchannel results for water worked well �41�. Increasing themass velocity, resulted in bubbles considerably smaller than thechannel hydraulic diameter. The channel “loses” its identity as amicroscale channel and practically transitioned to hydrodynamicsresembling conventional scale �or miniscale� characteristics.Therefore, under these conditions, the agreement with conven-tional scale models are fairly good. While the traditional form ofthe correlation fDd=const has been successfully used in micro-scale �41�, for the current flow configuration, correlation of thisform cannot be used because the product �fDd� was observed toincrease with heat flux, as shown in Fig. 13�a�. Therefore, usingleast squares method, the following correlation for the product ofthe bubble frequency and the bubble departure diameter has beendeveloped:

fnDd = 0.000397 �7�

where n is an empirically derived constant, which was calculatedto be 0.435. 85% of the predicted data is within 25% of theexperimental data with a MAE of 14% �No. of data points=29�.Unlike for �=0 deg, for �=180 deg, the product of the bubbledeparture frequency and departure diameter did not show any spe-cific trend with respect to heat flux and mass flux, but decreasedwith increasing local subcooling �or the Jakob number� as shownin Fig. 13�b�. With increased subcooling, the bubble growth wasimpeded by continuous condensation at the periphery, resulting inlower bubble departure frequencies and departure diameters.

5 ConclusionsFlow visualization revealed the existence of bubbly flow, mul-

tiple flows, and wavy-annular flow.At the onset of boiling, the bubbles nucleated on the rear side of

the pin fins and with the propagation of the boiling front, bubblesstarted nucleating at other locations. Nucleation sites were mostlyobserved at �=0 deg and �=180 deg.

The bubble departure frequency rapidly increased with wallheat flux for �=0 deg, but this trend gradually diminished at highheat fluxes, which has been attributed to the suppression of boil-ing. For �=180 deg, the bubble departure frequency first in-

n of local subcool for „a… �=180 deg and „b… �=0 deg
tio
creased and then decreased with heat flux.

APRIL 2010, Vol. 132 / 041002-7


w

pudsdh

A

gitp0

N

ux

0

Downloa

The bubble departure diameter showed a decreasing trend withall heat flux for both nucleation sites.The product of the bubble departure frequency and bubble de-

arture diameter follows the traditional form of the correlationsed in conventional scale systems for small bubble departureiameters and tended to follow the previous studies at the micro-cale when the bubble was large compared with the channel hy-raulic diameter. The product was observed to increase with walleat flux.

cknowledgmentThis work was supported by the Office of Naval Research �Pro-

ram Officer Mark Spector�. The microfabrication was performedn part at the Cornell NanoScale Facility �a member of the Na-ional Nanotechnology Infrastructure Network� which is sup-orted by the National Science Foundation under Grant No. ECS-335765, its users, Cornell University, and industrial affiliates.

omenclatureA � dimensionless constant in Eq. �8�

Ap � planform area, m2

b � dimensionless constant in Eq. �8�Bo � boiling number, qch� /Ghfg

c � dimensionless constant in Eq. �8�cp � Specific heat, kJ kg−1 K−1

d � pin fin diameter, mDd � bubble departure diameter, m

f � frequency, HzG � mass flux, kg m−2 s−1

H � height, mhfg � latent heat of vaporization, kJ kg−1

I � current, AJa � Jakob number, lcp�Tsat−Tmx� /vhfg

ks � thermal conductivity of silicon block,W m−1 K−1

L � length, mLo � overall channel length, m2

Lx � local channel length, m2

m � mass flow rate, kg s−1

p � pitch, mP � power, W

qch� � heat flux based on channel surface area,−2

Fig. 13 Variation of fDd product with „a… heat fl

W cm

41002-8 / Vol. 132, APRIL 2010


Qloss � heat loss, Wqw� � heat flux based on planform area, W cm−2

Red � Reynolds number, Gd /�Ti � inlet fluid temperature, °C

Tmx � local mean liquid temperature, °CTsat � saturation temperature, °C

Tsur � average surface temperature, °Ct � time, ms

ts � thickness of silicon block, mx � qualityz � axial distance along the channel, m

Subscriptsl � liquid

sat � saturationsur � surface

v � vaporw � wallx � local

Greek Symbols� � measured quantity� � viscosity, Pa s� � radial angle � density, kg m−3

� � surface tension, N m−1

References�1� Garimella, S., and Singhal, V., 2004, “Single-Phase Flow and Heat Transport

and Pumping Considerations in Microchannel Heat Sinks,” Heat TransferEng., 25�1�, pp. 15–25.

�2� Muwanga, R., Hassan, I., and MacDonald, R., 2007, “Characteristics of FlowBoiling Oscillations in Silicon Microchannel Heat Sinks,” ASME J. HeatTransfer, 129, pp. 1341–1352.

�3� Saidi, M. H., and Khiabani, R. H., 2007, “Forced Convective Heat Transfer inParallel Flow Multilayer Microchannels,” ASME J. Heat Transfer, 129, pp.1230–1236.

�4� Coursey, J. S., Kim, J., and Kiger, K. T., 2007, “Spray Cooling of High AspectRatio Open Microchannels,” ASME J. Heat Transfer, 129, pp. 1052–1060.

�5� Kosar, A., and Peles, Y., 2007, “Critical Heat Flux of R-123 in Silicon-BasedMicrochannels,” ASME J. Heat Transfer, 129, pp. 844–851.

�6� Chen, T., and Garimella, S. V., 2006, “Effects of Dissolved Air on SubcooledFlow Boiling of a Dielectric Coolant in a Microchannel Heat Sink,” ASME J.Electron. Packag., 128, pp. 398–405.

�7� Li, J., Peterson, G., and Cheng, P., 2004, “Three-Dimensional Analysis of HeatTransfer in a Micro-Heat Sink With Single-Phase Flow,” Int. J. Heat Mass

at �=0 deg. „b… Jakob number at �=180 deg.

Transfer, 47, pp. 4215–4231.



J

Downloa

�8� Kosar, A., Kuo, C.-J., and Peles, Y., 2006, “Suppression of Boiling FlowOscillations in Parallel Microchannels With Inlet Restrictors,” ASME J. HeatTransfer, 128�3�, pp. 251–260.

�9� Kosar, A., and Peles, Y., 2006, “Thermal-Hydraulic Performance of MEMSBased Pin Fin Heat Sink,” ASME J. Heat Transfer, 128, pp. 121–131.

�10� Kosar, A., Kuo, C.-J., and Peles, Y., 2005, “Reduced Pressure Boiling HeatTransfer in Rectangular Microchannels With Interconnected Reentrant Cavi-ties,” ASME J. Heat Transfer, 127�10�, pp. 1106–1114.

�11� Qu, W., and Mudawar, I., 2003, “Flow Boiling Heat Transfer in Two-PhaseMicrochannel Heat Sink: I. Experimental Investigation and Assessment ofCorrelation Methods,” Int. J. Heat Mass Transfer, 46�15�, pp. 2755–2771.

�12� Qu, W., and Mudawar, I., 2003, “Flow Boiling Heat Transfer in Two-PhaseMicrochannel Heat Sink: II. Annular Two-Phase Flow Model,” Int. J. HeatMass Transfer, 46�15�, pp. 2773–2784.

�13� Qu, W., and Mudawar, I., 2005, “A Systematic Methodology for OptimalDesign of Two-Phase Micro-Channel Heat Sinks,” ASME J. Electron. Packag.,127, pp. 381–390.

�14� Hitt, D. L., Zakrzwski, C. M., and Thomas, M. A., 2001, “Mems-Based Sat-ellite Micropropulsion Via Catalyzed Hydrogen Peroxide Decomposition,”Smart Mater. Struct., 10, pp. 1163–1175.

�15� Allen, R., Meyer, J. D., and Knight, W. R., 1985, “Thermodynamics andHydrodynamics of Thermal Ink Jets,” Hewlett-Packard J., 36, pp. 21–27.

�16� Losey, M. W., Jackman, J., Firebaugh, S. L., Schmidt, M. A., and Jensen, K.,2002, “Design and Fabrication of Microfluidic Devices for Multiphase Mixingand Reaction,” J. Microelectromech. Syst., 11, pp. 709–717.

�17� Prasher, R. S., Dirner, J., Chang, J.-Y., Myers, A., Chau, D., He, D., and Prstic,S., 2007, “Nusselt Number and Friction Factor of Staggered Arrays of LowAspect Ratio Micropin-Fins Under Cross Flow for Water as Fluid,” ASME J.Heat Transfer, 129�2�, pp. 141–153.

�18� Krishnamurthy, S., and Peles, Y., 2008, “Flow Boiling of Water in a CircularStaggered Micropin Fin Heat Sink,” Int. J. Heat Mass Transfer, 51�5–6�, pp.1349–1364.

�19� Siu-Ho, A., Qu, W., and Pfefferkorn, F. E., 2007, “Experimental Study ofPressure Drop and Heat Transfer in a Single-Phase Micropin-Fin Heat Sink,”ASME J. Electron. Packag., 129�4�, pp. 479–487.

�20� Cognata, T. J., Hollingsworth, K. D., and Witte, L. C., 2007, “High-SpeedVisualization of Two-Phase Flow in a Micro-Scale Pin-Fin Heat Exchanger,”Heat Transfer Eng., 28�10�, pp. 861–869.

�21� Kosar, A., and Peles, Y., 2006, “Convective Flow of Refrigerant �R-123�Across a Bank of Micro Pin Fins,” Int. J. Heat Mass Transfer, 49, pp. 3142–3155.

�22� Krishnamurthy, S., and Peles, Y., 2007, “Gas-Liquid Two-Phase Flow Across aBank of Micro Pillars,” Phys. Fluids, 19�4�, p. 043302.

�23� Kosar, A., and Peles, Y., 2007, “Boiling Heat Transfer in a Hydrofoil-BasedMicro Pin Fin Heat Sink,” Int. J. Heat Mass Transfer, 50�5-6�, pp. 1018–1034.

�24� Kosar, A., and Peles, Y., 2007, “Micro Scale Pin Fin Heat Sinks—ParametricPerformance Evaluation Study,” IEEE Trans. Compon. Packag. Technol.,30�4�, pp. 855–865.

�25� Siu-Ho, A. M., Qu, W., and Pfefferkorn, F. E., 2006, “Pressure Drop and HeatTransfer in a Single-Phase Micro-Pin-Fin Heat Sink, ASME Paper No.IMECE2006-14777.

�26� Schrage, D. S., Hsu, J.-T., and Jensen, M. K., 1988, “Two-Phase Pressure Dropin Vertical Cross-Flow Across a Horizontal Tube Bundle,” AIChE J., 34, pp.107–115.

�27� Xu, G. P., Tou, K. W., and Tso, C. P., 1998, “Two-Phase Void Fraction andPressure Drop in Horizontal Crossflow Across a Tube Bundle,” ASME J. Flu-



ids Eng., 120, pp. 140–145.�28� Jensen, M. K., and Hsu, J.-T., 1988, “A Parametric Study of Boiling Heat

Transfer in Horizontal Tube Bundle,” ASME J. Heat Transfer., 110, pp. 976–981.

�29� Cornwell, K., 1990, “The Influence of Bubbly Flow on Boiling From a Tube ina Bundle,” Int. J. Heat Mass Transfer, 33�12�, pp. 2579–2584.

�30� Polley, G. T., Ralston, T., and Grant, I. D. R., “Forced Crossflow Boiling in anIdeal In-Line Tube Bundle,” ASME Paper No. 80-HT-46.

�31� Leong, L. S., and Cornwell, K., 1979, “Heat Transfer Coefficient in a ReboilerTube Bundle,” Chemical Engineering, 343, pp. 219–221.

�32� Hwang, T. H., and Yao, S. C., 1986, “Cross Flow Heat Transfer in TubeBundles at Low Reynolds Numbers,” ASME J. Heat Transfer, 108, pp. 697–700.

�33� Leroux, K., and Jensen, M., 1992, “Critical Heat Flux in Horizontal TubeBundles in Vertical Crossflow of r-113,” ASME J. Heat Transfer, 114, pp.179–184.

�34� Jensen, M., and Tang, H., 1994, “Correlations for the CHF Condition in Two-Phase Crossflow Through Multitube Bundles,” ASME J. Heat Transfer, 116,pp. 780–783.

�35� Mandrusiak, G., and Carey, V., 1989, “Convective Boiling in Vertical Chan-nels With Different Offset Strip Fin Geometries,” ASME J. Heat Transfer,111, pp. 697–700.

�36� McGillis, G., and Carey, V., 1993, “Subcooled Convective Boiling of BinaryMixtures Over an Array of Heated Elements,” J. Thermophys. Heat Transfer,7�2�, pp. 346–351.

�37� Roser, R., Thonon, B., and Mercier, P., 1999, “Experimental Investigations onBoiling of N-Pentane Across a Horizontal Tube Bundle: Two-Phase Flow andHeat Characteristics,” International Journal of Refrigeration, 22, pp. 536–547.

�38� Fand, R. M., Keswani, K. K., Jotwani, M. M., and Ho, R. C. C., 1976, “Si-multaneous Boiling and Forced Convection Heat Transfer From a HorizontalCylinder to Water,” ASME J. Heat Transfer, 98, pp. 395–400.

�39� Zdravkovich, M. M., 1987, “The Effect of Interference Between Circular Cyl-inders in Crossflow,” J. Fluids Struct., 1, pp. 239–261.

�40� Lee, P. C., Tseng, F. G., and Pan, C., 2004, “Bubble Dynamics in MicroChannels. Part I: Single Microchannel,” Int. J. Heat Mass Transfer, 47, pp.5575–5589.

�41� Kuo, C.-J., Kosar, A., Peles, Y., Virost, S., Mishra, C., and Jensen, M., 2006,“Bubble Dynamics During Boiling in Enhanced Surface Microchannels,” J.Microelectromech. Syst., 15�6�, pp. 1514–1527.

�42� Revellin, R., and Thome, J. R., 2007, “A New Type of Diabatic Flow PatternMap For Boiling Heat Transfer in Microchannels,” J. Micromech. Microeng.,17�4�, pp. 788–796.

�43� Mizushima, J., and Suehiro, N., 2005, “Instability and Transition of Flow PastTwo Tandem Circular Cylinders,” Phys. Fluids, 17, p. 104107.

�44� Carey, V. P., 1992, Liquid-Vapor Phase-Change Phenomena, Taylor & Francis,London.

�45� Jakob, M., and Fritz, W., 1931, “Versuche über den Verdampfungsvorgang,”Forschung im Ingenieurwesen 2, pp. 435–447.

�46� Ivey, H. J., 1967, “Relationship Between Bubble Frequency, Departure Diam-eter and Rise Velocity in Nucleate Boiling,” Int. J. Heat Mass Transfer, 10, pp.1023–1040.

�47� Cole, R., 1967, “Frequency and Departure Diameter at Sub-Atmospheric Pres-sures,” AIChE J., 13, pp. 779–783.

�48� Zuber, N., 1963, “Nucleate Boiling-the Region of Isolated Bubbles-SimilarityWith Natural Convection,” Int. J. Heat Mass Transfer, 6, pp. 53–65.

APRIL 2010, Vol. 132 / 041002-9


flow boiling on micropin fins entrenched inside a ... · flow boiling on micropin fins entrenched...

Documents