8 Journal of Donghua University(Eng.Ed.)Vo!.23.No.3(2006)
Parametric Influence on Thermal Performance of Flat Plate Closed
Loop Pulsating Heat Pipes
YANG Hong·hai(杨洪海) ,KHANDEKAR Sameerz,GROLL Manfred
1 College of Environmental Science and Engineering.Dongh∞ University.Shanghai 201 620
2 Institute of Nuclear Technology and Energy System,University of Stuttgart,70569 Stuttgart,Germany
TIlis paper presents all experimental study on a flat plate
closed loop pulsating heat pipes.It consisted of total 40
channels with square cross sectioa (2× 2 nirll2.165 mm
long)machined directly on an aluminum plate(180 X 120 X 3
mJ ),which was covered by a transparent plate.1n1c
working fluid employed was ethano1.As the results,the
influence parameters of thermal performance were
investigated,such as filling ratiot hea t load and operational
orientations etc. Filling ratio was found to be a critical
parameter,and its effect was rather complicated.According
to its values the PHP plate could have four distinct working
zoues with differem operational characteristics and heat
trailsfer performance.The effect of heat load 0n thermal
performance was found to be positive,and .m general,
increa sing the heat loa d would improve heat transfer
performance.In order to analyze the effect of gravity on
thermal performance,three different heat modes and total
seven tilt angles were tested and compared. Successful
operation at all orientations wi th respect to gravity was also
achieved .
Keywords:fZat plate closed loop pulsating heat pipes,
parametric infZUeTIces,heat transfer characteristics.
Meandering tube pulsating heat pipes(PHPs),which
are very attractive entrants in the family of heat pipes,
have alrea dy found successful applications in cooling power
and microelectronicsE due to their simple structures,cost
effectiveness and excellent thermal performance.They also
have a potential for thermal control application for space
and avionics. Since its introduction in 1990’s research
activity in this area has steady increased,including various
experimental studies and attempts of mathematical
modelingE2-4].Yet reliable experimental data for PHPs is
very limited, especially in the wake of the ongoing
miniaturization process and efforts to combine electronic
compo nents and heat spreaders on the chip leve1.A logical
next step for further applications of PHPs in
microelectronics coo ling is to design integral struc tures ,
i.e. PHPs as an integral part of thermal spreader/
substrate,having typical dimensions as applicable for multi·
chip modules or PCBsL .Previous studies suggest some
intercsting thermal behavior of plate structures Iike inter-
channel heat transfer.contact angle hysteresis,capillary
effects due to sharp corners etd- . These and other
ope rational aspects need to be further studied and verified .
Experimental Setup
Fig.1 is the photography of the experimental setup.It
basically consists of vacuum device, electrical power
supply,cooling water cryostat,data logging systemt the
visualization part,etc.
Fig.1 Photograph of the experimental setup
The details of the PHP test section are shown
schematically in Fig.2.The specimen was made with an
aluminum plate(180× 120× 3 alia0)with 40 parallel
rectangular channels milled on it directly,forming a looped
serpentine structure.Each channel has the dimension of
1 65×2×2 mm3.A po lycarbonate plate with a transparent
silicon gasket was covered tightly on the top of the PHP
plate for visualization.A copper heater bloc k (100× 30×
15 mm0)and a water。cooled copper block (100× 6O×
12 mm3)were attached on the back face of the PliP plate.
The area be tween the heater and the cold plate formed the
Received Nov.9,2005
Supported by the German National Science Foundation(DFG)(No.GR一412/33)
Correspondence should be addressed to YANG hong-hai,associate prof.,E-mail:yhh@dhu.edu.cn
维普资讯 http://www.cqvip.com
Joorrml of Oon,:lm Universi (的 .Ed.)Yo1.23,No.3(2006) 9
adiabatic section, where a vacuum/filling tube was
provided near the center point. Axial temperature
distribution was measured by standard Type.K thermo-
couples(ID/OD = 0.5/1.5 ram). which were factory
calibrated and programmed with claimed net accuracy of±
0.5"C.The“Ahlborn-Almemo”programmable universal
data logging system(TYPE 5590—1,up t0 50 channe1)
was employed to collect the ternpe rature data.The data
reporting frequency was always 1 Hz.with a resolution of
0.I oC for temperature.During the expe riment。the total
PHP specimen was insulated well except the front face for
visualization.Thus the heat loss is kept less than 5% .Here
the value of the heat input loa d is considered nearly equal
to the electrical power.
Photograph of the plate
Fig.2 Schematic of the PH_P test section
HP
late
Ethanol was used as the working fluid.before charging.
the p"late was vacuumed until about 6×10一 mbar. fining
ration(is defined as the ration of the liquid volume enclosed in
the loops to the total loop volume)ranged from dry structure
t0 90% . In this study the thermal resistance of the dry
structure was also measured as reference values.In"addition.
three-heat modes including total seven-tilt angles could be
tested :they were bottom heat mode(30。,60。and 90。),
horizontal heat mode(0。)and top heat ilx~e(一30。,一6o’,
一 90。)respectively.
Results and Discussion
1 Division of working zone and operational
principle
According to the different value of the filling ratio,
the PHP plate showed different operational characteristics.
Four working zones were divided as shown in Table 1 and
Fig.3:
(1)Working zone I:when the filling ratio was quite
low (1ess than 2O%).the specimen operated in bottom heat
mode.but could not operate in horizontal or anti·gravity
top heat mode.There was almost no pulsating action.The
specimen worked mainly as two-phase gravity assisted
thermosyphon rather than PHPs.
(2)Working zonesⅡ:when the filling ratio was
be tween 25% and 40% ,the specimen ope rated both in
bo ttom and horizo ntal heat mode。but it still could not
operate in anti·gravity top heat mode.That means in this
zone pulsating action played role together with gravity
action.
(3)Working zone 111:when the filling ratio was in the
range of 45% to 75% ,the specimen operated rather well
in all global orientations. Th at means pulsating action
predominated in this working zone,and the specim en
worked as true PHP.
(4)Working zones Ⅳ :when the filling ratio was
greater than 75% ,the specimen could operate in bo ttom
heat mode and occasionally in horizontal heat mode(it was
not always guaranteed)。but it could never operate in anti·
gravity top heat mode. Th at means pulsating action
decreased again even disappeared in this zone.
Table 1 Filling ration and possible operation mode
Filling Bottom Horizontal Top
rati0 heat nl0de heat In0de heat In0de Zone
O%
5%
2O%
3O%
40%
5O%
65%
8O%
9O%
100%
Empty plate,heat transfer by conductivity
√
√
Single phase liquid,
co nvective flow
x/
\/
I
Ⅱ
Ⅲ
Ⅳ
heat transfer througll nature
2 Comparison of thermal performance,
analyses of the influence parameters
In this paper,thermal performance was evaluated by
average evapo rate tempe rature,thermal resistance and the
maximum lim itation of heat input load.Here,thermal
resistance was defined as ternpe rature difference be tween
evapo rator and condenser divided by heat input load.
2.1 Eff酏t offilling ratio Ollthermal performance
As abo ve mentioned , filling ratio was a critical
parameter.For different filling ratio,the aluminum P臁
showed different operational characteristics. Hence
thermal pe rformance changed greatly with filling ratio as
shown in Fig.3.It co uld be found that:
(1)In working zone I (filling ratio.5%一20%),the
specim en showed relative smaller thermal resistance and
维普资讯 http://www.cqvip.com
1O Journal of Donghua University(Eng.Ed.)Vo/.23,No.3(2006)
: 一-o-IⅢ(X )W[ ‘。m 。 ‘mde
一
(90
一
~)50W 2 一 · a)
+ I 一 【lI】 一 -一 一
三 三 //
/厂’\/ //
\ ., 一
一
~ :
三;==一 ---4"SS一‘
一 1 —● E= —:
0
1.20
1.00
0 80
0.60
20 40 60 80
Filling ratio(%)
(a)Bottom heat mode
+50w Horizontal mode(0。) (b)
..-iImw 、 ,一 ’
+I50W I ./ +2tX)W
,250W \'/,
_..1IXIW
一 35oW :一 / 一
一 4cmW
-_=
0
0.40
兰 0.20
20 40 60 80
Filling ratio( )
(b)Horizontal mod e
+ 1I)0w Top heatmode(--90。1 (c)
+ I50w
一 200w
一 250W
‘ 30(1W ●一 一 一
一 350W
=:=
especially when the filling ratio was in the between of 50%
and 65% . The PHPS showed excellent thermal
performance in all heat modes and tilt angles (Fig.4).
Comparing the case for filling ratio 50% and 65% , the
former showed a little be tter thermal pe rformance.
However.their differences were very small except in the
case of bottom heat mod e with lower heat input(1ess than
50W ).
(4)In working zone IV (filling ratio> 75%),the
specimen showed the poorest thermal pe rformance.There
were two probable reasons.One was that for too higher
filling ratio,there was no enough free space for bubbles to
expand and to play the pumping action.Another one was
that the capillary action of sharp angled edges decreased
greatly as the filling ratio was too high.Both of them
hindered the fluid flow.
(5)In the case of bottom heat mod e there existed two
optimum filling ratio,they were 15% and 40% ,showing
relative be tter thermal pe rformance. The former
corresponded to the maximum thermosyphon action of the
structure:the latter was probable the mutual results of
thermosphon action and pulsating action.
(6)In the case of horizontal mod e for lower heat load
(1ess than 150 W ).40% was an optimum—filling ratio;but
for higher heat load(more than 200 W).50% filling ratio
showed better thermal performance. It was tentatively
analyzed as follows:with the increase of filling ratio,on
one hand,the capillary action of sharp corners decreased
gradually,but on the contrary,pulsating action increased
at first and achieved maximum at 50% filling ratio,and
then decreased again.An other fact was that higher heat
input generally enhanced the pulsating action.So there
should be an optimum—filling ratio and it changed with heat
input and heat mod e.
(7)Increase of heat load mitigated the difference of
thermal pe rformance caused by the different filling ratio
(Fig.4(a)).
As abo ve analysis,effect of filling ratio on thermal
pe rformance was complicated,it is be cause thermosyphon
action,pulsating action,capillary action of sharp corners,
which decided the thermal performance were themselves
greatly depe nded on the filling ratio.
2.2 Effect of heat input on thermal performance
Fig.4 depicts the effect of heat input on the thermal
resistance.It was found that increasing heat load markedly
improves the thermal performance.Till abo ut 200 W input
powers the performance improvement was quite drastic
while thereafter it was mild.This tendency was the same
for all filling ratios and inclination angles.
In genera1.the heat input was the “pump’’for the
thermo-fluidic action. Thus increasing the “pumping
po wer” increased the pe rformance.At low filling ratios
如 ∞ 舳 ∞ ∞ 如 ∞
●
O O O O O
l(芝 一0u写 一s3二duI矗LI
如 ∞ 舳 ∞ ∞ 如 ∞
n n
l(芝 一03u皇∽【s0三一 上| LI卜 l(≥~ 一03udlSl∞u_l
维普资讯 http://www.cqvip.com
J~ rm l of Doo0/~a University(Eng.Ed.)Vo1.23,No.3(2006) 11
(Zone I—Thermosyphon mode and Zone II—Transition
zone),increasing the heat input made the liquid layer in
the counter—current flow thinner. The effective fluid
velocity also increased thus enhancing the local wall heat
transfer coefficient. In Zo ne 111 operation, bubble
pumping action got enhanced due to rapid bubble growth.
Also,after a certain heat input,the flow changed from
purely capillary slug flow to churn or semi—annular and
sometimes to fully annular flow in individual channels.This
greatly enhanced the heat transfer coe fficient. Thus ,
increasing heat load enhanced the pe rformance till a certain
type of dry—out occurs.In Zo ne I,the dry—out occurred as
a combination of counter.current flow limitation and
insufficient fluid inventory. starving some channels
completely In other zones,no dry.out could be observed.
Experiments were terminated for safety reasons beyond
120℃ loeal temperatures.
1 20
Heatload(v
(a)Bottom heat mode
(b). Horizontal mode(0。) —.-} =27
{ —.。l =舯% FR=∞%
、
—.。l =6s%
- {
: ==
. .
0
1.oo
0 8O
0.6O
0.40
O 2O
O oo
0 2oo 3oo 400 5oo
Heatload(v
(b)Horizontal mode
(c)、 T。p 。 m。 。(一90。 1+rR 。。
、 I二:.!==:::
\
、
i
、
、
、
:
...⋯
O 2oo 3oo 400 500
Heat1oad rW3
(c)Top heat mode
Fig.4 Effect of hea t flux on the thermal resistance
of the tested structure
Fig.5 records the maximum heat input load achievable
in three different heat modes for reaching the average
eva19orator temperature 100℃ .it was found that in the
case of bottom heat mode,when filling ratio was in a wide
range from 10% to 70% ,the PHP spe cimen worked rather
well and pe rmitted higher heat loads in the case of
horizontal and anti—gravity top heat mod es, only when
filling ratio was in a relative narrow range from 45% to
70% ,the PHP specimen showed higher heat transfer
pe rformance.
≥
一
勺
旦
工
量
E
矗
— — ._一 Bottom heat
mode(90。)
— — Horizontal heat
mode(0。)
— — , Top heat mode
(-90。)
Filling ratio
Fig.5 Maximum permit heat load(Te~100"C)
2.3 Eff t of heat mode and tilt angle on thermal
performance
Heat mode evidently affected the operation performance
of the aluminum P胁 .In the case of bo ttom heat mode,
the PHP plate ope rated at all filling ratios~in horizontal
heat mod e,PHP plate could only ope rate in working zones
11 andⅢ (filling ratio:25%一75%);in anti—gravity top
heat mod e.the PHP plate co uld only ope rate in working
zoneⅢ (filling ratio:45%一75%).
Fig.6 quantitatively analyzes the effect of tilt angle
on thermal performance. It was found:in the case of
vertical bottom heat mode(+90。),the aluminum PHPs
always showed the minimum thermal resistance. which
increased as tilt angle changed to horizontal(0。) and
then to anti—gravity top heat mode (一90。). This
tendency was as the same as for all filling ratios and heat
inputs.The results indicated that the gravity vector really
affected the fluid flow and heat transfer characteristics of
一 9O 一 60 —3O 0 30 60 90
Tilt angle(。)
(a)Heat input 150W
≥^ \p — uI|g∞ 暑E LI卜
加 ∞ 舳 ∞ ∞ 加 ∞
● ● 0 0 0 0 0
≥^ \p— u g∞ aJ JJ LI卜
≥^、p一0u LI曼∽ aJ jLlJuL1 LIJ
≥^ \p一0u LI≈ 3J互三uc_l_
维普资讯 http://www.cqvip.com
Joorrm/of D哪 』a University(Eng.Ed.)Vo/.23,No.3(2006)
Tilt angle(。)
(b)Heatinput 250W
I
I
—
p
三
芒
&
兰
V-
(c) "lop heat mode(--90。) L
1 f -
.,v I - 一 ~
I
/ h_ '
,。。 150 2001:∞ 300 350 ∞
f HeatinputinW att
I—Ave rag~T-
(c)Top heat mode
Fig 7 Variation of average evaporator
temperature with time
Fig.6 Effect of heat mode and tilt angle
。“ h。 m 。p。 m “ 。 Conclusion
PHPs:in the bottom heat mode,gravity played positive
action on the fluid flow thus enhanced heat transfer
performance;but in horizontal and anti·gravity top heat
mod,it played no action or negative action on the fluid
flow thus hindered heat transfer.
Fig.7 depicts the typical tempe rature-time history for
the average evaporator temperature at filling ratio about
5o% under three different operating conditions.
The fluctuations in the evaporator were the minimum
f0r the bottom heat mode.The amplitude increased as the
inclination angle changed to the horizontal and then to the
top heat mode.OveralI smoother operation was obtained at
higher heat loads.Unacceptable variations were recorded
at relatively low heat input(1ess than 50 W)in top heat
m ooe. . ■
p
三
芒
&
量
p
三
芒
&
三
(a) Bottom heat mode 90。 I
,~ }
, 、..-
r、、
厂 一——
f ∞ IOO 150 2oo 25o 30o 3∞ 40o 450
f H∞I_m Llin蝴
J
— A~fage
Time(s)
(a)Bottom heat mode
00
【b) Horizontal mode(0‘)
一
I . .^ ^
^‘ 山
} I
j
/
5o 1oo 15O 2oo 250 300 350 400
HeatmpL—m w |Il
『--A坩 9e
Time(s)
(b)Horizontal mode
Experimental study was pe rformed on an alum inum
flat plate closed loop pulsating heat pipes,thus the effect
of filling ratio,heat Ioad and operational orientation on
thermal performance were quantitatively analyzed.Here
are the brief summaries:
(1) Filling ratio was a critical parameter。and its
effect on thermal pe rformance was complicated.According
t0 its value,four working zones were divided .Each one
showed the different ope rational characteristics and heat
transfer pe rformance.Both the ca pillary action of sharp
angles,gravity assisted thermosyphon action and pu lsating
action changed with the variation of filling ratio,thus
there exist some optimum filling ratios leading to the
optimum performance, which also depended on other
parameters,e.g.heat load and operational orientations,
etc.
(2)Heat input Ioad was coilsidered as the source of
pumping action,and it played the positive role on the
operation of PHPS.In general·increasing the heat input
would enhance the fluid flow and heat transfer
performance.However,there shou ld exist perform ance
lim itations to avoid the occurrence of dry·out, which
depended on the mechanism of P唧 s.
(3)The gravity vector really affected the fluid flow
and heat transfer performance.It played positive action in
the case of bottom heat m(x~e (+ 0 。),and played no
action or negative action when the PHPs operated in ttle
horizontal(O。)and anti·gravity top heat mode(一90 ).
Therefore,the best thermaI pe rformance always occurred
in verticaI bottom heat mode (+ 9O。),then decreases
gradually as the tilt angle decreases from +90。to 一90。.
References
[1]
[2]
Akachi H .and Pohi~ck F.,Preprints 10 th Int
CD .,Stuttgart,Germany,1997,3:8—12
GrollM . and Khandekar S,, Proc. 3rd hit
Heat Pj
Conf.otl
维普资讯 http://www.cqvip.com
Joorrm/of£)【 ghl^a University(En9.Ed.)Vo1.23,No.3(2006) 13
Transport Phenomena in Muhiphase跏 fP, ,Baranow
Sandomierski,Pbland,2002:35—43
[3] Khandekar S.,Groll M.and Luckchoura V.,Electronics
Cooling·2003·9(2):38—41
[4] Groll M.and Khandekar S.,Proc.3rd,}lf. l,.On Etwrgy
apulF~tdromnent,Shangh~,China,2003.1:723—730
[5] Lee W.,Jung H.,Kim J.,Ct a1.,Proc.11th lnt.Heat
Plpe Conf.。Tokyo,Japan-1999=355—360
[6] Khandekar S.,Schneider M., Schiller P., et a1.,
Microscale Thermophysical Engi,,eeri,,g·2002,6(1)
维普资讯 http://www.cqvip.com