V Minsk International Seminar “Heat Pipes, Heat Pumps, Refrigerators”
Minsk, Belarus , September 8-11, 2003
INVESTIGATION OF VAPOR GENERATION INTO CAPILLARY
STRUCTURES OF MINIATURE LOOP HEAT PIPES
V.M. Kiseev, A.S. Nepomnyashy, N.L. Gruzdova
Ural State University
Department of Thermophysics and Surface Phenomena
Lenin av. 51, 620083, Ekaterinburg, Russia
(3432) 616-775
(3432) 616-778 fax
Valery.Kiseev@usu.ru
Abstract
Extensive development Loop Heat Pipes (LHP) technology offers capabilities to play LHP an increasingly
important role for instrument thermal control on some space and earth applications. However, heat transport
systems on basis of miniature LHP (MLHP) are interesting in electronics packaging for high heat flux in zones
of heat input and transport. The applications could include thermal management of a various electronic devices
such as computer processors (CPU), where the heat dissipation requirement is expected to be in the range of 50
to 100 Watts. This paper presents some configurations of MLHP with flat plates of evaporators and examines the
experimental data which is investigated for MLHP due to gravity. The heat transfer coefficient by phase
exchange in the LHP capillary structures is investigated for different configurations of the vapor ducts, working
fluids and capillary structures. Experimental tests verify that the Miniature LHP operates successfully by high
heat flux in zones of heat input and transport and “loop heat pipe” heat transfer mechanism is better for these
conditions than ‘classical’ heat pipes.
KEYWORDS
Loop Heat Pipe, Steady-state behavior, heat transfer coefficient, capillary structure, flat evaporator.
INTRODUCTION
Loop heat pipes are two-phase transport devices which utilize the capillary pressure developed in a
fine capillary structure (FCS) (pore wick) to circulate the working fluid in a closed loop system
connected heat source and heat sink. The LHP are capable of operating effectively at any orientation
(top heat, horizontal heat and bottom heat) in a gravitational field over large distances. In the LHP the
liquid return to the evaporator section is along a smooth-walled tube with low frictional resistance.
The wick design for the loop scheme is small since it is only located in the evaporator section. In this
case it is the basic difference beside other heat pipes. LHP systems have the potential to transport large
amounts of heat over large distance or unfavorable accelerations with minimal temperature drops.
Due to the extensive development efforts in ground tests and flight experiments over the last few
years, the LHP have reached the state of technology readiness and commercialization. Theoretical
studies are also being conducted to better understand the operational mechanism of the LHP. It is
anticipated that the LHP will become a major player in the thermal management of space and
terrestrial systems in the near future [1].
On the other hand, the LHP are excellent heat transfer devices for electronics packaging for high
heat flux in zones of heat input and transport, when heat transport zone may connect heat source and
heat sink by the flexible tubes at any orientation. It is important for a miniaturization of electronic
components and for development of electronics as a whole [2-4].
This paper presents some configurations of MLHP with flat plates of evaporators and examines the
experimental data which is investigated for MLHP due to gravity. The experiment effort was
supplemented by extensive hardware development efforts, analytical modeling development and
ground test verification. A summary of hardware development include:
201
V.M. Kiseev et.al.
• Metal wick development status including nickel and titanium with effective pumping in the range
of 0.5 to 3 micron and permeability in the range 1 x 10-14 m2 to 4 x 10-13 m2.
• The development of MLHP with flat evaporators.
• Condensers and heat sinks development.
Analytical modeling development was focused on predicted wick material conductivity effects on
MLHP performance, heat and mass transfer between wick and compensation chamber and
design/performance characterization of wicks. Theoretical calculations of MLHP and MLHP software
are presented and compared with several experimental results.
Extensive ground test verification has been performed by Ural State University (USU) during the
past several years. Results that will be presented include:
• The heat transfer coefficient by phase exchange in the LHP capillary structures as a function of
heat input, different configurations the vapor ducts, working fluids and capillary structures.
• MLHP performance as a function of heat input/heat sink, heat agent, orientation and accelerations.
• Power cycling and sink temperature cycling behavior.
• Temperature and pressure drop characteristic.
EXPEREMENTAL APPARATUS AND PROCEDURE
In fact, in electronics packaging for high heat flux in zones of heat input where the LHP capillary
structure is a generator of vapor and capillary pump the heat transfer coefficient cannot be avoided.
The external and internal heat transfer coefficient can be defined. The external heat transfer coefficient
is provided a good contact between electronic components and LHP evaporator wall and is not of
interest of LHP performance. The internal heat transfer coefficient is very important for LHP
performance by high heat flux and defines the possibility to lead out of heat by using phase exchange
in the LHP wick. In spite of a lot the theoretical and experimental studies in these conditions it is not a
clear reply how to increase heat transfer coefficient by using capillary phenomena.
The effects of the phase exchange in the LHP wick were studied by using the open LHP system
(external pressure about 105 Pa). The schematic diagrams of experimental apparatus are shown in
Fig.1.
8 7
a) b)
L
8 6
H 4 Н2 О V
7
5
L 1
6 2
Tw in Tv
1
2 3 9
3 5
4
Q Tw ex Q Tv
Fig.1. Open LHP systems:
1 – LHP evaporator; 2 – capillary structure (wick); 3 – vapor ducts (grooves); 4 – compensation
chamber (cavity); 5 – vapor line; 6 – liquid line; 7 – condenser; 8 – water cooler; 9 – copper heater;
Tw ex, Tw in, Tv – thermocouples
202
V.M. Kiseev et.al.
The configuration and the geometrical sizes of vapor ducts in the LHP open system (Fig.1b) were
optimized. A wick is connected with the cooper heater without the wall of evaporator. Fig.2 and Fig.3
give examined configurations of vapor ducts.
a)
1 2 3 4
Concentric vapor ducts Radial vapor ducts
b)
5 6 7
Fig.2. Configurations of vapor ducts (1 – 7):
a) – only radial vapor ducts; b) – radial and concentric vapor ducts
1) a b 2) Q 3) Q
c
Q
V
V
V
L L
L
Fig.3. Geometry of concentric vapor ducts:
1 – rectangular vapor ducts; 2 – trapezoid vapor ducts; 3 – triangular vapor ducts
Metal wick development status including nickel and titanium with effective pumping in the range of
0.5 to 3 micron and permeability in the range 1 x 10-14 m2 to 4 x 10-13 m2 was based in this studying.
Table1 shows the main properties of metal porous wick.
Table1. Main properties of metal wicks (all data are experimental)
No Wick material Porosity Effective Permeability Effective thermal
ε, % pore radius K, 10-14 m2 conductivity
Rp, 10-6 m λ, W/m K
1. Titanium (Ti) 55 3.8 33.9 1.29
2. Titanium (Ti) 50 3.6 28.3 1.77
3. Nickel (Ni) 65 0.65 2.2 8.04
4. Nickel chips of porous Ni 53 1.4 2.2 9.16
203
V.M. Kiseev et.al.
Fig. 4 shows the basic configuration and geometrical size MLHP flat plates of evaporators.
31.8 40
1
30.0
V L
15 22
4 13
6 2
5.0 6
25.0
1,0
38.0
aa 2 3
Fig.4. The basic configuration and geometrical size MLHP flat plates of evaporators
1 – case of evaporator; 2 – wick (capillary structure); 3 – vapor ducts
The researching object (LHP or system of it's modeling) was located in thermostatic box where with
a help of automatic regulation system the temperature of the environment was Tenv = 25±0,5 oC or
Tenv = 40±0,5 oC. The methodic of the experiment was oriented on measuring of the temperature
fields of researching objects with different heat agent, orientations and temperatures of the
environment and heat sink. The heat input was controlled through variable transformers and measured
with wattmeter. Heat input rate was given by a resistive electric heater and measured by a calorimetric
plate. At the condenser heat out was taken by means of water in the water jacket which was installed
in cooling section. Water flow rate and the inlet/outlet temperature were measured and heat transfer
rates were calculated by calorimetric method. Temperature was measured with copper constantan
thermocouples located at appropriate points on the tube wall and surrounding insulation at the
evaporator, condenser and adiabatic sections in each. The precision of measurements of the heat flow
rate was no higher than 5%, and the precision of measurements of temperatures was no higher than
0,5%. Before the beginning of the experiment was checked the connection between a wick and a case
of evaporator by the air-bubble method. The experiments were realized for adverse 1g acceleration
(evaporator above condenser “+H”) and favorable 1g acceleration (condenser above evaporator “-H”).
EXPEREMENTAL RESULTS
At the first step of experiment was studied the influence of the wick thickness (δ) on a heat transfer
coefficient (α) with heat agent of water and acetone. This data are presented in Fig.5 and Fig.6. As can
be seen from this comparison, at other conditions being equal, there is the increasing of the heat
transfer coefficient by the thickness of the wick about δ = (6-7) mm. The dependences α = f(q) have
maximum that it is evidence of a competing phenomena into capillary structure between liquid and
vapor flow. The internal heat transfer coefficient α (W/m2 K) is defined as
q Q tr
α = , q =
T w in − T v S inp
(1)
where q is heat flux (W/m2), Qtr is heat transfer rates calculated by calorimetric method in condenser,
Sinp is the square of the surface of heat input, Tw in is temperature on the internal surface of a
204
V.M. Kiseev et.al.
evaporator wall (the surface of the contact between a case wall and a wick) and Tv is temperature of
vapor.
1,4 1
1 Water 1
1,2 Acetone
4 4
0,8
a , 10 W/м К
1 3 3
2
0,8 2 0,6 2
4
0,6
0,4
0,4
0,2
0,2
0 0
0 5 10 15 20 0 5 10 15
q, 104 W/м2 q, 104 W/м2
Fig.5. Dependence the heat transfer coefficient α on the heat flux q for Nickel wick No 3 (Table 1)
1 – δ = 7 mm; 2 – δ = 3 mm; 3 – δ = 11 mm; 4 - δ = 6 mm. Adverse 1g acceleration H = + 0.03 m.
Open LHP system by external pressure about 105 Pa
3 1,6
Water 1,4 Acetone
a,10 W/м K
1,2
2
2 3
2 1
1
4
4
0,8
1
0,6 3
2
0,4
1
0
0,2 4
0 20 40 60
4
q,10 W/м
2 0
0 10 20 30
0 100 200 300 Q, Вт 4 2
q, 10 W/м
Fig.6. Dependence the heat transfer coefficient α on the heat flux q for Titanium wick No 1 (Table 1)
1 – δ = 7 mm; 2 – δ = 3 mm; 3 – δ = 11 mm; 4 - δ = 6 mm. Adverse 1g acceleration H = + 0.03 m
Open LHP system by external pressure about 105 Pa
The next series of the experiments solved a problem of optimization of a configuration and the sizes
of the vapor ducts. The models of the capillary structure (No 3, Table 1) had the optimal thickness δ =
6 mm. The vapor ducts (radial and concentric, Fig.2) had quadratic profile (1 x 1 mm) on a surface of
the wick and a different configuration. The effective diameter of the vapor duct Def and their total
205
V.M. Kiseev et.al.
relative section ξ are defined as
4S Svd
Def = , ξ=
Π Sinp (2)
where S is the square of vapor duct section, П is vapor duct perimeter; Svd is the total square of the
section of the vapor ducts on the surface of the contact between the wick and the wall of heat input.
This data is presented in Fig.7.
0,9 1
Aceton Acetone
0,8 0,9
0,7 0,8
a, 104 W/м2 К
0,6 1 0,7
0,5 3 0,6
4 0,5
0,4
7 0,4
0,3 6
0,3
0,2 5
0,2
2
0,1 Р = 100 0,1
0 Р = (60 - 500)
0
0 5 10 15 0 5 10 15
4 2
q, 10 W/м q, 10 W/м
4 2
Fig.7. Dependence the heat transfer coefficient α on the heat flux q for Nickel wick No 3 (Table 1)
1 – No 1 (Fig.2), ξ = 0; 2 – No 2 (Fig.2), ξ = 0.09; 3 – No 3 (Fig.2), ξ = 0.16; 4 - No 4 (Fig.2),
ξ = 0.23; 5 - No 5 (Fig.2), ξ = 0.33; 6 - No 5 (Fig.2), ξ = 0.42; 7 - No 7 (Fig.2), ξ = 0.61;
Adverse 1g acceleration H = + 0.03 m; left: open LHP system; right: close LHP system
As can be seen from Fig.7 the maximal heat transfer coefficient by the maximum of heat flux is
achieved by ξ = 0.4 -0.45 (No 5, Fig.2) as for open LHP system and for close LHP system. Existence
of an optimum in surface configuration of vapor grooves gives the evidence about the competition
between hydraulic and thermal resistances in zone of vapor grooves.
After optimization of the coefficient ξ that is determined practically of geometry of heat and mass
input to surface of evaporation into capillary structure was made several series of experiments. It
optimized the effective diameter of the concentric vapor ducts Def. The radial vapor ducts (Conf. No
5, Fig.2) had only quadratic profile (1 x 1 mm) on a surface of the wick. The concentric vapor ducts
had a different profile (Fig.3) and Def. By this study we didn’t observe the visible influence of
concentric vapor ducts profile on heat transfer process. Therefore we used triangular concentric vapor
ducts surface as more technological. They were located on the wick or on an internal surface of the
wall of heat input. These data are shown in Fig.8.
As can be seen in Fig.8 the decrease of the sizes and distance between concentric vapor ducts goes
to intensification of heat transfer and the increasing of heat transfer coefficient. The placement of
concentric vapor grooves on an internal surface of the wall increase the heat transfer coefficient more
than (10 – 30) %. Apparently, the decrease of distance between concentric vapor ducts decreases the
distance of vapor exit from vapor pores to vapor grooves improves the vapor collection from pores and
accordingly intensify heat transfer in fine porous structure.
206
V.M. Kiseev et.al.
1,2 1,8
Acetone
Acetone 1,6
1 Р = 100 kPa
1,4 3
Р = 100 kPa
a , 10 W/м К
4
0,8 1,2
2
5
1 6
0,6
4
1 0,8
0,4 2 0,6
3 0,4
0,2 4
0,2
0 0
0 5 10 15 0 5 10 15 20
4 2
q, 10 W/м 4
q, 10 W/м
2
Fig.8. Dependence the heat transfer coefficient α on the heat flux q for Nickel wick No 3 (Table 1)
Left: concentric vapor ducts are located on the wick.
1 – Def = 2 mm; 2 – Def = 1.5 mm; 3 – Def = 1 mm; 4 - Def = 0.5 mm;
Right: concentric vapor ducts are located on a internal surface of the wall of heat input.
5 - Def = 0.3 mm; 6 - Def = 0.2 mm;
Adverse 1g acceleration H = + 0.03 m; open LHP system
CONCLUSIONS
The results obtained in heat transfer experiments in MLHP with flat plates of evaporators are
summarized as follows:
(1) There is an optimal thickness of the fine porous wick (about 5-7 mm) for concrete conditions of
MLHP operation.
(2) It is showed that for the intensification of heat transfer in zone of heat input of MLHP with flat
plates of evaporators a special system of vapor ducts is necessary. There is the optimal surface square
of vapor grooves not more than 0.4 – 0.5 from heat input square.
(3) The decreasing of the sizes and distance between concentric vapor ducts goes to intensification of
heat transfer and the increase of heat transfer coefficient. The placement of concentric vapor grooves
on an internal surface of the wall increase the heat transfer coefficient more than (10 – 30) %.
Apparently, the decrease of distance between concentric vapor ducts decreases the distance of vapor
exit from vapor pores to vapor grooves improves the vapor collection from pores and accordingly
intensify heat transfer in fine porous structure.
References
1. Faghri A. Heat pipe science and technology, Taylor & Francis, Washington, 1995, pp. 578-624.
2. Kiseev V., Belonogov A. Miniature heat transport systems with loop heat pipes, Proc. of 4th Minsk
Int. Seminar ‘Heat Pipes, Heat Pumps, Refrigerators”, Minsk, Belarus, 2000, pp. 15-22.
3. Chernysheva M.A., Vershinin S.V., Maidanik Yu.F. Development and test results of loop heat pipes
with a flat evaporator, Proc. of 12th Int. Heat Pipe Conference, Moscow, Russia, 2002, A2-3.
4. Pastukhov V.G., Maidanik Yu.F., Vershinin S.V., Korukov M.A. Miniature loop heat pipes for
electronic cooling, Proc. of 12th Int. Heat Pipe Conference, Moscow, Russia, 2002, F2.
207
V.M. Kiseev et.al.
