You are on page 1of 11

Mathematical and Computer Modelling 57 (2013) 126–136

Contents lists available at SciVerse ScienceDirect

Mathematical and Computer Modelling


journal homepage: www.elsevier.com/locate/mcm

Mathematical model for heat transfer limitations of heat pipe


Patrik Nemec ∗ , Alexander Čaja, Milan Malcho
University of Žilina, Department of Mechanical Engineering, Univerzitná 1, 01026 Žilina, Slovakia
University of Žilina, Department of Power Engineering, Univerzitná 1, 01026 Žilina, Slovakia

article info abstract


Article history: This article is about heat transport limitation standard-used construction wick heat pipes
Received 24 March 2011 with various working fluids operating at temperatures from −30 °C to 140 °C. In this paper
Received in revised form 20 June 2011 is the proposal and verification of a mathematical model for calculating heat transport
Accepted 21 June 2011
limitations of heat pipe. The calculations of the various variants of the wick structure of heat
pipe and working fluid are presented in diagrams as a dependency of the heat transport
Keywords:
limitations on the working temperature. In the conclusion, the effect of these limits on the
Heat pipe
Heat transfer
cooling power of the heat pipe is evaluated.
Thermal performance © 2011 Elsevier Ltd. All rights reserved.
Capillary structure
Mathematical model

1. Introduction

The heat pipe is a vapor–liquid phase-change device that transfers heat from a hot reservoir to a cold reservoir using
capillary forces generated by a wick or porous material and a working fluid. The heat pipe is composed of a container lined
with a wick that is filled with liquid near its saturation temperature. The vapor–liquid interface, usually found near the inner
edge of the wick, separates the liquid in the wick from an open vapor core. Heat flowing into the evaporator is transferred
through the container to the liquid-filled wicking material, causing the liquid to evaporate and vapor to flow into the open
core portion of the evaporator. The capillary forces generated by the evaporating interface increase the pressure difference
between the vapor and liquid. The vapor in the open core flows out of the evaporator through the adiabatic region and into
the condenser. The vapor then condenses, generating capillary forces similar, although much less in magnitude, to those in
the evaporator. The heat released in the condenser passes through the wet wicking material and container out into the cold
reservoir. The condensed liquid is then pumped, by the liquid pressure difference due to the net capillary force between the
evaporator and condenser, out of the condenser back into the evaporator. Proper selection and design of the pipe container,
working fluid, and wick structure are essential to the successful operation of a heat pipe. The heat transfer limitations,
effective thermal conductivity, and axial temperature difference define the operational characteristics of the heat pipe. In
Fig. 1 a schematic of a heat pipe aligned at angle ψ relative to the vertical axis (gravity vector) is shown [1].
Due to the two-phase characteristics, the heat pipe is ideal for transferring heat over long distances with a very small
temperature drop and for creating a nearly isothermal surface for temperature stabilization. As the working fluid operates
in a thermodynamic saturated state, heat is transported using the latent heat of vaporization instead of sensible heat or
conduction where the heat pipe then operates in a nearly isothermal condition. This nearly isothermal condition offers
benefits of transporting large amounts of heat efficiently, decreasing the overall heat transfer area and saving system weight.
The amount of heat that can be transported through the use of latent heat is typically several orders of magnitude greater

∗ Corresponding author at: University of Žilina, Department of Power Engineering, Univerzitná 1, 01026 Žilina, Slovakia.
E-mail addresses: nemecpatrik@gmail.com, patrik.nemec@fstroj.uniza.sk (P. Nemec), alexander.caja@fstroj.uniza.sk (A. Čaja),
milan.malcho@fstroj.uniza.sk (M. Malcho).

0895-7177/$ – see front matter © 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.mcm.2011.06.047
P. Nemec et al. / Mathematical and Computer Modelling 57 (2013) 126–136 127

Fig. 1. Schema of the wick heat pipe [1].

than transported by sensible heat for a geometrically equivalent system. Additionally, no mechanical pumping systems
are required due to the capillary-driven working fluid. Given the wide range of operating temperatures for working fluids,
the high efficiencies, the low relative weights, and the absence of external pumps in heat pipes, these systems are seen as
attractive options in a wide range of heat transfer applications [2].

2. Mathematical model

A mathematical model is created in MS Office Excel Basic on computing relations of heat transfer limitations, which
define the boundary of heat pipe performance. Each of the relations express part of the total heat pipe performance, which
is influenced by hydrodynamic and thermal processes which occur in the heat pipe and are known as a heat transport
limitation. Each of these limitations exists alone and they are unique and not influenced by each other.

2.1. Heat transport limitations

Fig. 2. Heat transfer limitations of water wick heat pipe with sintered capillary structure (heat pipe inner diameter 20 mm, total length 2 m, axial
orientation 90°, spherical diameter of copper powder 0.85 mm, porosity 0.55, width of capillary structure 6 mm).

Using the analysis techniques for each limitation independently, the heat transport capacity as a function of the mean
operating temperature (the adiabatic vapor temperature) can be determined. This procedure yields a heat pipe performance
region similar to that shown in Fig. 2. As shown, the separate performance limits define an operational range represented
by the region bounded by the combination of the individual limits. In effect, this operational range defines the region or
128 P. Nemec et al. / Mathematical and Computer Modelling 57 (2013) 126–136

combination of temperatures and maximum transport capacities at which the heat pipe will function. Thus, it is possible to
ensure that the heat pipe can transport the required thermal load or to improve the design. It is important to note that in the
determination of the heat transport capacity, the mean operating temperature must be identified. However, the operating
temperature of a standard heat pipe is a function of the heat input, thus resulting in a mutually dependent case between the
heat transport and the operating temperature [2].
Heat pipes undergo various heat transfer limitations depending on the working fluid, the wick structure, the dimensions
of the heat pipe, and the heat pipe operational temperature.
Capillary limitation
The capillary limit relates to the fundamental phenomenon governing heat pipe operation which is development of
capillary pressure differences across the liquid-vapor interfaces in the evaporator and condenser. When the driving capillary
pressure is insufficient to provide adequate liquid flow from the condenser to the evaporator, dryout of the evaporator wick
will occur. The driving potential for the circulation of the working fluid is the capillary pressure difference, the maximum
capillary pressure must be greater than the sum of all pressure losses inside the heat pipe.
(1Pc )max ≥ 1Ptot . (1)
The pressure losses in heat pipes can be separated into the frictional pressure drops along the vapor and liquid paths, the
pressure drop in liquid as a result of body forces (e.g., gravity, centrifugal, electromagnetic) and the pressure drop due to
phase transition.
1Ptot = 1Pν + 1Pl + 1Pb + 1Pph . (2)
During heat pipe operation, the menisci naturally adjust the radii of curvature for the capillary pressure differential to
balance the pressure losses 1Ptot . However, the maximum radius of curvature is limited to the capillary dimension of
the wick structure such that the maximum transport occurs when (1Pc )max = 1Ptot . It is important to note that the
pressure drop associated with phase transition 1Pph is significant only under very high condensation or evaporation rates
and represents the jump condition associated with the kinetic energy change in the liquid–vapor process. Except for very
specific conditions (e.g., liquid metal heat pipes with extremely high evaporation rates), the phase transition pressure drop
is typically negligible and will not be considered further in following discussions.
If the total pressure drop exceeds the maximum capillary pressure, the return rate of liquid to the evaporator will be
insufficient and the heat pipe will experience dryout of the wick. The maximum capillary pressure 1Pc developed within
the heat pipe wick structure is given by the Laplace–Young equation.

1Pc = · cos θ (3)
reff
where, θ is contact angle fluid–wick. The contact angle is a measure of the degree of wettability of the liquid on the wick
structure, where θ = 0° relates to a perfectly wetting system. reff is the effective pores radius of the wick and can be
determined for various wick structures [2].
The capillary limitation in heat pipes occurs when the net capillary forces generated by the vapor–liquid interfaces in the
evaporator and condenser are not large enough to overcome the frictional pressure losses due to fluid motion. This causes
the heat pipe evaporator to dry out and shuts down the transfer of heat from the evaporator to the condenser. For most heat
pipes, the maximum heat transfer rate due to capillary limitation can be expressed as [3].

σl · ρl · lv K · Aw ρl · g · lt · cos Ψ
 
· 2
Qc = · · − (4)
µl leff reff σl
where, K is the wick permeability (m2 ), Aw is the wick cross-sectional area (m2 ), ρl is the liquid density (kg/m3 ), µl is
the liquid viscosity (N · s/m2 ), reff is the wick capillary radius in the evaporator (m), g is the acceleration due to gravity
(9.8 m/s2 ), and lt is the total length of the pipe (m) [1].
Viscous limitation
When the heat pipe operates at low temperatures, the available vapor (saturation) pressure in the evaporator region
may be very small and of the same magnitude as the required pressure gradient to drive the vapor from the evaporator to
the condenser. In this case, the total vapor pressure will be balanced by opposing viscous forces in the vapor channel. Thus,
the total vapor pressure within the vapor region may be insufficient to sustain an increased flow. This low-flow condition
in the vapor region is referred to as the viscous limit. As the viscous limit occurs at very low vapor pressures, the viscous
limit is most often observed in longer heat pipes when the working fluid used is near the melting temperature (or during
frozen startup conditions) as the saturation pressure of the fluid is low. Busse in [4] provided an analytical investigation of the
viscous limit. The model first assumed an isothermal ideal gas for the vapor and that the vapor pressure at the condenser end
was equal to zero, which provides the absolute limit for the condenser pressure. Using these assumptions, a one-dimensional
model of the vapor flow assuming laminar flow conditions was developed and expressed as

π · rv4 · lv · ρv · Pv
Qv = (5)
12 · µv · leff
P. Nemec et al. / Mathematical and Computer Modelling 57 (2013) 126–136 129

where, rv is the cross-sectional radius of the vapor core (m), lv is the latent heat of vaporization (J/kg), µv is the vapor
viscosity in the evaporator (N · s/m2 ), leff is the effective length of the heat pipe (m), Pv (Pa) and ρv (kg/m3 ) are the vapor
pressure and density at the evaporator end of the heat pipe. The values predicted by this expression were compared with the
results of previous experimental investigations and were shown to agree well [4]. For cases where the condenser pressure is
not selected to be zero, as could be the case when the viscous limit is reached for many conditions, the following expression
is used
 
Av · rv2 · lv · ρv · Pv
2
Pv,c
Qv = · 1− (6)
16 · µv · leff Pv2

where, Pv,c is the vapor pressure in the condenser. In [4] it is noted that the viscous limit could be reached in many cases
when Pv,c /Pv ∼ 0.3. To determine whether the viscous limit should be considered as a possible limiting condition, the vapor
pressure gradient relative to the vapor pressure in the evaporator may be evaluated. In this case, when the pressure gradient
is less than one-tenth of the vapor pressure, or 1Pv /Pv < 0.1, the viscous limit can be assumed not to be a factor. Although
this condition can be used to determine the viscous limit during normal operating conditions, during startup conditions
from a cold state, the viscous limit given by [4] will probably remain the limiting condition. As noted earlier, the viscous
limit does not represent a failure condition. In the case where the heat input exceeds the heat input determined from the
viscous limit, this results in the heat pipe operating at a higher temperature with a corresponding increase in the saturation
vapor pressure. However, this condition typically is associated with the heat pipe transitioning to being sonic limited, as
discussed in the following section [2].
Sonic limitation
The sonic limit is typically experienced in liquid metal heat pipes during startup or low-temperature operation due to
the associated very low vapor densities in this condition. This may result in choked, or sonic, vapor flow. For most heat
pipes operating at room temperature or cryogenic temperatures, the sonic limit is typically not a factor, except in the case
of very small vapor channel diameters. With the increased vapor velocities, inertial, or dynamic, pressure effects must be
included. It is important to note that in cases where the inertial effects of the vapor flow are significant, the heat pipe may
no longer operate in a nearly isothermal case, resulting in a significantly increased temperature gradient along the heat
pipe. In cases of heat pipe operation where the inertial effects of the vapor flow must be included, an analogy between heat
pipe operation and compressible flow in a converging–diverging nozzle can be made. In a converging–diverging nozzle, the
mass flow rate is constant and the vapor velocity varies due to the varying cross-sectional area. However, in heat pipes, the
area is typically constant and the vapor velocity varies due to mass addition (evaporation) and mass rejection (condensation)
along the heat pipe. As in nozzle flow, decreased outlet (back) pressure, or in the case of heat pipes, condenser temperatures,
results in a decrease in the evaporator temperature until the sonic limit is reached. Any further increase in the heat rejection
rate does not reduce the evaporator temperature or the maximum heat transfer capability but only reduces the condenser
temperature due to the existence of choked flow.
The sonic limitation actually serves as an upper bound to the axial heat transport capacity and does not necessarily
result in dryout of the evaporator wick or total heat pipe failure. Attempts to exceed the sonic limit result in increasing both
the evaporator temperature and the axial temperature gradient along the heat pipe, thus reducing further the isothermal
characteristics typically found in the vapor flow region.
Busse [4] presented an alternative approach by assuming that only inertial effects are present in one-dimensional flow.
In this case the momentum equation yields:
dP d
=− · ρ · v2 . (7)
dx dx
Integration of this expression, combining it with the continuity equation, and assuming that the vapor behaves as an ideal gas
yields an expression for the maximum heat transport capacity as a function of the thermophysical and geometric properties.
 12    12
ρ v · Pv

P P
Q = lv · · 1− . (8)
Av Pv Pv
A determination of the point where the first derivative, dQ /dP, vanishes yields a relationship for the sonic limit

Qs = 0.474 · Av · lv · (ρv · Pv )0.5 (9)


where, Av is the cross-sectional area of the vapor core (m ), ρv (kg/m ) and Pv (Pa) are the vapor density and pressure at
2 3

the evaporator exit. The greatest difficulty in determining the sonic limit is determining these two quantities along with the
inlet pressure to the condenser.
It is interesting to compare the results for the viscous limit and sonic limit where a relationship between the two exist
with respect to the quantity Pv · ρv . Inertial effects are found to vary with the product (Pv · ρv )1/2 , while the viscous effects
vary linearly with respect to Pv · ρv . As a result, when this product is small, the transport capacity is typically limited by
viscous effects but with increasing Pv · ρv , inertial effects begin to dominate and a transition occurs from the viscous to the
sonic limit. The boundary between these two limits can be determined by setting these two equations equal to each other
130 P. Nemec et al. / Mathematical and Computer Modelling 57 (2013) 126–136

and solving for the combined terms as a function of temperature [5,4]. The results indicate that the transition temperature
is dependent on the thermophysical properties of the working fluid, the geometry of the heat pipe, and the length of the
evaporator and condenser regions [2].
Entrainment limitation
Examination of the basic flow conditions in a heat pipe shows that the liquid and vapor flow in opposite directions. The
interaction between the countercurrent liquid and vapor flow results in viscous shear forces occurring at the liquid–vapor
interface, which may inhibit liquid return to the evaporator. In the most severe cases, waves may form and the interfacial
shear forces may become greater than the liquid surface tension forces, resulting in liquid droplets being picked up or
entrained in the vapor flow and carried to the condenser. In [6] doubt is expressed as to whether entrainment actually occurs
in a capillary-driven heat pipe because the capillary structure would probably retard the growth of any surface waves. In a
majority of cases studied, the wick structure of the heat pipe was flooded (i.e., excess liquid), which allowed entrainment
to occur. The most common approach to estimating the entrainment limit in heat pipes is to use a Weber number criterion.
Cotter [7] presented one of the first methods to determine the entrainment limit. This method utilized the Weber number,
defined as the ratio of the viscous shear force to the forces resulting from the surface tension.

2 · rh,v · ρv · Vh2
We = . (10)
σ
By relating the vapor velocity and the heat transport capacity to the axial heat flux as

Q
vv = (11)
Av · ρv · lv
and assuming that to prevent entrainment of liquid droplets in the vapor flow, the Weber number must be less than unity,
the maximum transport capacity based on entrainment can be written as
0.5
ρv · σl

Qe = Av · lv · (12)
2 · r c ,a v e
where, σl is the surface tension (N/m) and rc ,av e is the average capillary radius of the wick. Note that for many applications
rc ,av e is often approximated by reff . The entrainment limit refers to the case of high shear forces developed as the vapor
passes in the counter flow direction over the liquid saturated wick, where the liquid may be entrained by the vapor and
returned to the condenser. This results in insufficient liquid flow of the wick structure [2].
Boiling limitation
At higher heat fluxes, nucleate boiling may occur in the wick structure, which may allow vapor to become trapped in the
wick, thus blocking liquid return and resulting in evaporator dryout. This phenomenon, referred to as the boiling limit, differs
from the other limitations, as it depends on the radial or circumferential heat flux applied to the evaporator, as opposed to
the axial heat flux or total thermal power transported by the heat pipe. Determination of the heat flux or boiling limit is
based on nucleate boiling theory and is comprised of two separate phenomena:
• bubble formation,
• subsequent growth or collapse of the bubbles.
Bubble formation is governed by the size (and number) of nucleation sites on a solid surface and the temperature difference
between the heat pipe wall and the working fluid. The temperature difference, or superheat, governs the formation of
bubbles and can typically be defined in terms of the maximum heat flux as

λeff
Q = · 1Tc (13)
Tw
where, λeff is the effective thermal conductivity of the liquid–wick combination. The critical superheat 1Tc is defined by [8]
as

 
Tsat
1Tc = · − (1Pc )max (14)
lv · ρv rn
where, Tsat is the saturation temperature of the fluid and rn is the critical nucleation site radius, which according to [9]
ranges from 0.1 to 25.0 µm for conventional metallic heat pipe case materials. As discussed by [10], this model yields a
very conservative estimate of the amount of superheat required for bubble formation and is true even when using the
lower bound for the critical nucleation site radius. This is attributed to the absence of adsorbed gases on the surface of the
nucleation sites caused by the degassing and cleaning procedures used in the preparation and charging of heat pipes. The
growth or collapse of a given bubble once established on a flat or planar surface is dependent on the liquid temperature
and corresponding pressure difference across the liquid–vapor interface caused by the vapor pressure and surface tension
of the liquid. By performing a pressure balance on any given bubble and using the Clausius–Clapeyron equation to relate the
P. Nemec et al. / Mathematical and Computer Modelling 57 (2013) 126–136 131

temperature and pressure, an expression for the heat flux beyond which bubble growth will occur may be developed 3 [5]
and expressed as

4π · leff · λeff · Tv · σl
 
1 1
Qb = · − (15)
lv · ρv · ln
ri
rv
rn reff

where, λeff is the effective thermal conductivity of the composite wick and working fluid (W/m K), Tv is the vapor saturation
temperature (K), ri is the inner container radius (m), rv is the vapor core radius and rn is the nucleation radius.
The boiling limit occurs when the applied evaporator heat flux is sufficient to cause nucleate boiling in the evaporator
wick. This creates vapor bubbles that partially block the liquid return and can lead to evaporator wick dryout. The boiling
limit is sometimes referred to as the heat flux limit [2].

2.2. Calculation heat transport limitation of ethanol heat pipe with sintered wick structure

For calculation, heat transport limitations are needed in order to know the thermophysical properties of working fluid
in the heat pipe (Table 1 shows the thermophysical properties of ethanol in the temperature range), heat pipe parameters
(shown in Table 2), the thermal conductivity of heat pipe material (λm of copper is 393 (W · m−1 · K−1 )), the working condition
of heat pipe (t is 50 °C) and axial orientation of heat pipe (ψ is 180°).

Table 1
Thermophysical properties of ethanol.
t (°C) ρL (kg·m−3 ) σL (N · m−1 ) λL (W · m−1 · K−1 ) µl (N · s · m−2 ) lv (J · kg−1 ) pv (Pa) ρv (kg · m−3 ) µv (N · s · m−2 ) νL (m2 · s−1 )
−30 825.0 0.02760 0.177 0.0034 939 400 1 000 0.02 0.0000075 0.000004121
−10 813.0 0.02660 0.173 0.0022 928 700 2 000 0.03 0.000008 0.000002706
10 798.0 0.02570 0.170 0.0015 904 800 3 000 0.05 0.0000085 0.000001880
30 781.0 0.02440 0.168 0.00102 888 600 10 000 0.38 0.0000091 0.000001306
50 762.2 0.02310 0.166 0.00072 872 300 29 000 0.72 0.000097 0.000000945
70 743.1 0.02170 0.165 0.00051 858 300 76 000 1.32 0.0000102 0.000000686
90 725.3 0.02040 0.163 0.00037 832 100 143 000 2.59 0.0000107 0.000000510
110 704.1 0.01890 0.160 0.00028 786 600 266 000 5.17 0.0000113 0.000000398
130 678.7 0.01750 0.159 0.00021 734 400 430 000 9.25 0.0000118 0.000000309

For calculating heat transport limitations of ethanol wick heat pipe with sintered capillary structure, the main parameters
of the heat pipe such as inner container radius, cross-sectional radius of vapor core, evaporation length of the heat pipe,
adiabatic length of the heat pipe and condensation length of the heat pipe are needed. From the main parameters of the
heat pipe are derived other parameters needed in the calculation. In Table 2 are all the parameters of the heat pipe (under
the line are derived parameters from basic parameters up the line)

Table 2
Parameters of heat pipe.

Cross-sectional radius of vapor core (m) rv 0.005


Inner container radius (m) ri 0.0065
Evaporation length of heat pipe (m) le 0.15
Adiabatic length of heat pipe (m) lad 0.2
Condensation length of heat pipe (m) lc 0.15
Total length of heat pipe (m) lt 0.5
Effective length of heat pipe (m) leff 0.35
Cross-sectional area of the vapor core (m2 ) Av 0.000079
Wick cross-sectional area (m2 ) Aw 0.000054
Thermal conductivity of heat pipe material (W · m−1 · K−1 ) λm 393

lt = le + lad + lc (16)
leff = 0.5 · (le + lc ) + lad (17)

Av = π · rv 2 (18)
Aw = π · (ri − (ri − h) ).
2 2
(19)
An important part of wick heat pipe is its capillary structure. There exist three basic capillary structures: grooved
capillary structure, mesh screen capillary structure and sintered capillary structure. For calculation and verification of the
mathematical model a sintered capillary structure was chosen. The sintered capillary structure is made by sintering copper
powder on the inner surface of the heat pipe. The main parameters of a sintered capillary structure are diameter of powder
as well as porosity and width of the capillary structure. From the main parameters (up the line) are derived other parameters
132 P. Nemec et al. / Mathematical and Computer Modelling 57 (2013) 126–136

Table 3
Parameters of sintered capillary structure.

Sphere diameter (m) d 0.0001


Porosity (-) ε 0.65
Width of capillary structure (m) h 0.0015
Permeability (m2 ) K 1.4945 · 10−10
Effective radius of capillary structure (m) reff 0.000021
Effective thermal conductance (W · m−1 · K−1 ) λeff 1.69

of the capillary structure needed in the calculation and to create the mathematical model (under the line). In Table 3 are
all the parameters of the sintered capillary structure of the heat pipe (last three are derived parameters from the main first
three parameters up the line)

reff = 0.21 · ds (20)


d ·ε
2 3
K = [11] (21)
150 · (1 − ε)2
2 · λl + λm − 2 · (1 − ε) · (λl − λm )
λeff = λl [11]. (22)
2 · λl + λm + (1 − ε) · (λl − λm )

2.3. Mathematical model verification

According to the above equations of limitations and input heat pipe parameters, a calculation was created using MS
Office Excel. The result of the mathematical model is graphic dependencies of heat transport limitations from the heat pipe’s
working temperature. Mathematical model results of heat transport limitations for specific heat pipe types was compared
with results from the measurement of heat pipe performance at temperatures of 50 °C and 70 °C. These two temperatures
are average values from the working temperature range of ethanol heat pipe, which are used for the cooling of power
convertors. In Fig. 3 is a graphic comparison of heat transport limitations determining total performance of the heat pipe
from the mathematical model with measured performance of ethanol wick heat pipe with sintered capillary structure and
a spherical diameter of copper powder of 0.1 mm. The blue line creates a boundary of heat pipe performance by capillary
limitation and the red line shows boiling limitations.

Fig. 3. Verification of mathematical model by measuring heat pipe performance (ethanol wick heat pipe with sintered capillary structure and a spherical
diameter of copper powder of 0.1 mm).

According to calculations of the mathematical model, the determining limitations to this special kind of wick heat pipe
for the total working temperature range of ethanol, which was used as a working fluid in the heat pipe, can be determined.
The black line applies to results from the heat pipe’s performance at temperatures of 50 °C and 70 °C. The wick heat pipe
was made in our workplace and the manufacturing process is discussed in [12], then the heat pipe’s performance measured
on an experimental measuring device is discussed in [13]. In Fig. 4 it can be seen that the blue line and black line are in a
similar region at temperatures of 50 °C and 70 °C. The next two figures verify the mathematical model, where measuring
heat flux of the next wick heat pipes with sintered capillary structure and with grooved capillary structure is in graphs in
a similar region as the calculation results of capillary limitation, which is the determining limitation for the next two kinds
of heat pipes at temperatures of 50 °C and 70 °C. In Fig. 5 is the capillary limitation determining limitation at a full working
temperature range for ethanol grooved wick heat pipe.
P. Nemec et al. / Mathematical and Computer Modelling 57 (2013) 126–136 133

Fig. 4. Verification of mathematical model by measuring heat pipe performance (ethanol wick heat pipe with sintered capillary structure and a spherical
diameter of copper powder of 0.05 mm).

Fig. 5. Verification of mathematical model by measuring heat pipe performance (ethanol wick heat pipe with grooved capillary structure length 0.3/ width
0.2/ pitch 0.3).

3. Graphic dependencies of the mathematical model

Based upon verification of the mathematical model with the measuring of heat pipe performance, some interesting
dependencies of heat pipe performance and geometric parameters of the heat pipe’s capillary structure are obtained.
Knowledge from results may be used in optimization of sintered wick heat pipe design applications. The lines in the graphs
represent dependencies and maximal heat flux area of the heat pipe from working temperatures. In Figs. 6 and 7 the influence
of groove dimensions on total heat pipe performance is shown. Heat transport limitations of wick heat pipe with grooved
capillary structure and groove dimensions (high 0.3 mm, width 0.2 mm and pitch 0.3) created by the mathematical model.
In Fig. 6 the influence of groove height from 0.3 to 0.9 mm on total heat pipe performance is shown. It can be seen that pipe
performance rises with an increase in groove height. However, increasing the groove height from 0.7 to 0.9 is the boiling
limitation, shown as a main limitation, and at working temperatures from 80 °C to 130 °C decreases heat pipe performance.
A groove height of 0.6 is shown as a optimal height for this specific type of grooved wick heat pipe.

Fig. 6. Dependence of heat pipe performance from groove length in heat pipe with grooved capillary structure.
134 P. Nemec et al. / Mathematical and Computer Modelling 57 (2013) 126–136

In Fig. 7 the influence of groove widths from 0.2 to 0.9 mm on total heat pipe performance is shown. From the figure
it can be seen at a temperature range from 40 °C to 90 °C there is an extreme increase in heat pipe performance. In this
temperature range the capillary limitation is determined. At lower temperatures from the working temperature range,
increasing the groove width becomes the main sonic limitation. At higher temperatures the boiling limitation as a main
limitation for groove widths from 0.6 to 0.9 are values of boiling limitation equal at temperatures from 80 °C to 130 °C. The
according graphic dependencies show a groove width of 0.6 as an optimal width for this specific type of grooved wick heat
pipe.

Fig. 7. Dependence of heat pipe performance from groove width in heat pipe with grooved capillary structure.

Next, the graphic dependencies of heat pipe performance are created from a mathematical model for ethanol wick heat
pipe with sintered capillary structure and various porosities, spherical diameter of copper powder, capillary structure width
and heat pipe position. In Fig. 8 the influence of porosity on heat pipe performance is shown. The porosity of capillary
structure can be changed by adding some additive to sintered technology. There is an obvious rise in heat pipe performance
with increasing the porosity of the capillary structure. With the higher permeability of the capillary structure, the heat pipe
transfer demonstrates increased flux. The increasing permeability of the capillary structure can however cause entrainment
of liquid flow to the evaporator by vapor flow. This may cause dryout of the heat pipe’s evaporation section and decrease
total heat pipe performance.

Fig. 8. Dependence of heat pipe performance from capillary structure porosity of the sintered wick heat pipe.

In Fig. 9 the influence of the spherical diameter of copper powder on the sintered capillary structure is shown. Using the
bigger spherical diameter of copper powder to sintered technology creates a higher porosity capillary structure. It can be
P. Nemec et al. / Mathematical and Computer Modelling 57 (2013) 126–136 135

said that the increase in porosity is directly proportional to the spherical dimension of copper powder, and to make a more
porous capillary structure isn’t a required additive to sintered technology. In this case an increase in heat pipe performance
when using bigger spherical dimensions of copper powder can be seen.

Fig. 9. Dependence of heat pipe performance from the spherical diameter of the copper powder in sintered wick heat pipe.

In Fig. 10 the influence of capillary structure width on heat pipe performance is shown. Capillary structure width is
an important factor, which influences heat pipe performance. In the figure it can be seen that an increase of the capillary
structure’s width increases heat pipe performance at a temperature range from −30 °C to 60 °C, where capillary limitation as
a main limitation is shown. On the other hand an increase of the capillary structure width decreases heat pipe performance
at a temperature range from 80 °C to 130 °C. This may be caused by nucleation of bubbles in the capillary structure, when
evaporation of returned liquid from the condensation section to the evaporation section of the heat pipe occurs. Then values
of boiling limitation determine total heat pipe performance.

Fig. 10. Dependence of heat pipe performance from capillary structure width of the sintered wick heat pipe.

In Fig. 11 the influence of the position of a heat pipe on their performance is shown. Wick in heat pipe ensures liquid
returns from the condensation section to the evaporation section of heat pipe and therefore the wick heat pipe can operate
at various tilt angles, even horizontally. The deeper the tilt angle is from vertical, the greater the decrease in heat pipe
performance. It can be said that the heat pipe performance of a wick heat pipe in the horizontal position is only half that at
vertical.
136 P. Nemec et al. / Mathematical and Computer Modelling 57 (2013) 126–136

Fig. 11. Dependence of heat pipe performance from position of the sintered wick heat pipe.

4. Conclusion

Graphic dependencies expressly define which heat transfer limitations are the biggest influence on total performance
of the heat pipe. Generally, the limitation values depend on heat pipe parameters, wick structure parameters and the
thermophysical properties of the working fluid. The highest values reach viscous limitation and sonic limitation. These
limitations reach so high values, that in many cases they can be neglected. They influence the performance of heat pipes
at low temperature near the freezing temperatures of the working fluid. The critical limitations influencing heat pipe
performance are entrainment limitation, capillary limitation and boiling limitation. These limitations are dependant on
the thermophysical properties, wick and heat pipe parameters. The thermophysical properties of each working fluid are
stable in the temperature range and they can’t change. But for total heat performance, the optimization of the heat pipe can
change the dimensions of the wick structure. The dimensions of the wick structure are the most influenced by permeability,
which causes the capillary effect in wick heat pipe. However we have to be careful because increased permeability and
increased pore dimensions decrease capillary pressure, which is the biggest influence regarding circulation of the working
fluid. Generally, the capillary limit is the primary maximum heat transport limitation of heat pipe.

Acknowledgments

This article was created within the framework of project APVV-0577-10, ‘‘Cooling of power electronic systems by cycles
without mechanical drive’’.

References

[1] L.W. Swanson, in: Frank Kreith (Ed.), Heat Pipe, Heat and Mass Transfer, Mechanical Engineering Handbook, CRC Press LLC, Boca Raton, 1999.
[2] Jay M. Ochterbeck, Heat pipes, in: Heat Transfer Handbook, 1st ed., 2003.
[3] S.W. Chi, Heat Pipe Theory and Practice, Hemisphere Publishing, Washington, DC, 1976.
[4] C.A. Busse, Theory of the ultimate transfer of cylindrical heat pipes, International Journal of Heat and Mass Transfer 16 (1973) 169–186.
[5] M.N. Ivanovskii, V.P. Sorokin, I.V. Yagodkin, The Physical Properties of Heat Pipes, Clarendon Press, Oxford, 1982.
[6] C.A. Busse, J.E. Kemme, Dry-out phenomena in gravity assist heat pipes with capillary flow, International Journal of Heat and Mass Transfer 23 (1980)
643–654.
[7] T.P. Cotter, Heat pipe startup dynamics, in: Proc. SAE Thermionic Conversion Specialist Conference, Palo Alto, CA, 1967.
[8] B.D. Marcus, On the operation of heat pipes, Report 9895-6001-TU-000, TRW, Redondo Beach, State, CA, 1965.
[9] P.D. Dunn, D.A. Reay, Heat Pipes, 3rd ed., Pergamon Press, New York, 1982.
[10] P.J. Brennan, E.J. Kroliczek, Heat pipe design handbook, NASA Contract Report, NAS5–23406, 1979.
[11] V. Hlavačka, F. Polášek, P. Štulc, V. Zbořil, Tepelné Trubice v Elektrotechnice, STNL-Nakladatelství Technické Literatúry, Praha, 1990.
[12] P. Nemec, A. Čaja, R. Lenhard, Visualization of heat transport in heat pipes using a thermocamera, Archives of Thermodynamics 31 (4) (2010) str. 125.
[13] P. Nemec, A. Čaja, R. Lenhard, Thermal performance measurement of heat pipe, in: Book of Abstract on Fourth Global Conference on PCO, Kuching,
Sarawak, Malaysia, str. 14, 2010.

You might also like