Applied Thermal Engineering 65 (2014) 361e368

Contents lists available at ScienceDirect

Applied Thermal Engineering

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

Experimental and theoretical analysis of a heat pipe heat exchanger

operating with a low global warming potential refrigerant
Giovanni A. Longo a, *, Giulia Righetti a, Claudio Zilio a, Fabio Bertolo b
a
University of Padova, Department of Management and Engineering, Str.lla S.Nicola 3, I-36100 Vicenza, Italy
b
D.B.M. S.p.a., Via Ugo La Malfa 10, 28040 Varallo Pombia, NO, Italy

h i g h l i g h t s

This paper presents the experimental and theoretical analysis of a Heat Pipe Heat Exchanger (HPEH).

The performance of the HPEH operating with HFC134a is compared with the results obtained using HFO1234ze(E).

The HPEH is tested in operating conditions typical for air-conditioning in European countries.

The consistency of the experimental results is proved by the comparison with a semi-empirical model of the HPEH.

a r t i c l e i n f o a b s t r a c t

Article history: In the present paper the experimental and theoretical analysis of a Heat Pipe Heat Exchanger (HPEH) is
Received 23 August 2013 reported. The HPHE consists of 54 horizontal 12.7 mm micro-fin copper tubes 700 mm in length and
Accepted 16 January 2014 aluminium continuous fins 0.115 mm in thickness with 3 mm of fin spacing. The HPHE presents an
Available online 24 January 2014
evaporator section of 270 mm, a condenser section of 270 mm and adiabatic section of 160 mm. The
performance of the heat exchanger operating with the well-known hydrofluorocarbon refrigerant
Keywords:
HFC134a is compared with the results obtained using the new low global warming potential
Heat pipe
hydrofluoro-olefin HFO1234ze(E). Tests were run in a double wind channel with air operating conditions
GWP
HFO1234ze(E)
typical for air-conditioning in European countries for both summer and winter seasons. The consistency
of the experimental results is proved by the comparison with a semi-empirical model of the heat
exchanger based on heat transfer correlations available in literature.
Ó 2014 Elsevier Ltd. All rights reserved.

1. Introduction In spite of the large diffusion of this type of heat exchangers,

Heat Pipes Heat Exchangers (HPHE) usually consist of several
heat pipes that are mechanically expanded against continuous fins
(typically in aluminium), according to the classical construction
procedure of finned coils condensers and evaporators for vapour
compression refrigeration machines. HPHE are largely used for
energy recovery purposes both in civil (air conditioning) and in
industrial (air-to-air or gas-to-gas heat recovery) applications.
HPHE can promote relatively high heat transfer effectiveness, they
do not need any power input, they do not present moving parts
(thus endorsing higher reliability in comparison to “active systems”
based on vapour compression heat recovery units) and they allow
theoretically complete separation between hot and cold fluids.
there is relatively poor evidence in the open literature about several
issues related to the design and application of HPHE.
Riffat and Ma [1] and Srimuang and Amatachaya [2] presented a
review on the application of HPHE for heat recovery, aiming at
exploiting the energy saving potentials towards global warming
prevention. Yau and Ahmadzadehtalatapeh [3] focused their review
on horizontal HPHE with particular emphasis on tropical climates
applications. The common conclusion of the above mentioned re-
views is that the literature on the application of HPHE in air con-
ditioning systems is very limited and there is a clear need for
further studies to optimise the HPHE design towards a max-
imisation of the achievable energy savings. In particular, it is
possible to stress the following points:

1. The works dealing with HPHE with horizontal tubes are very
limited in number;
* Corresponding author. 2. No papers are available using tubes with internal helicoidally
E-mail address: tony@gest.unipd.it (G.A. Longo). arranged micro fins;

1359-4311/$ e see front matter Ó 2014 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.applthermaleng.2014.01.023
362 G.A. Longo et al. / Applied Thermal Engineering 65 (2014) 361e368

Nomenclature Ref film Reynolds number

T temperature,  C
A heat transfer area, m2 U overall heat transfer coefficient, W m2 K1
cp specific heat capacity, J kg1 K1 w face velocity, m s1
C heat capacity rate, W K1 X humidity ratio
dext tube outside diameter, m
dh hydraulic diameter, m Greek symbols
D finned coil depth, m a heat transfer coefficient, W m2 K1
f friction factor bg inclination angle of the groove
f.s. full scale D difference
g gravity acceleration, m s2 ε sensible thermal effectiveness
G volumetric flow rate, m3 s1 l thermal conductivity, W m1 K1
hg groove depth, m m viscosity, kg m1 s1
lf fin pitch, m r density, kg m3
lr row pitch, m s surface tension, N m1
lt tube pitch, m
L tube length, m Subscripts
Lt total tube length, m atm atmospheric
L0t effective tube length, m c condensation
m mass flow rate, kg s1 e evaporation
M figure of merit, kg s3 exh exhaust air
ng number of grooves in inlet
nr number of rows l liquid
nt number of tubes per row ln logarithmic
NTU number of transfer unit out outlet
p pressure, Pa ref refrigerant
q heat flux, W m2 sat saturation
Q heat flow rate, W supply supply air
r latent heat of vaporization, J kg1 v vapour
sf fin thickness, m w wall
Re Reynolds number

3. Hydrofluorocarbon fluids (HFC) are widely used, nevertheless
they are not widely investigated in literature.
Reay and Harvey [4] recently highlighted the role of heat pipes
in intensified unit operations, whereas MacGregor et al. [5] evi-
denced the importance of using Low Global Warming potential
working fluids for closed two-phase thermosyphons stressing the
application of HFO1234ze(E).
Abd El-Baky and Mohamed [6] experimentally analysed a HPHE
consisting of 25 horizontal copper pipes with a length of 500 mm,
inside diameter 10.2 mm, outside diameter 12.7 mm, evaporator
section length 200 mm, condenser section 200 mm and adiabatic
section 100 mm (inner diameter/total tube length ratio R ¼ 49.0).
They inserted four layers of 100 mesh brass screen wick. The heat
pipe working fluid was CFC11. They used 0.5 mm thick aluminium
fins (fin spacing 3.5 mm). No information about heat pipes fluid
charge was supplied.
Guo et al. [7] studied a commercially available HPHE, with 12
tubes, arranged in two rows, 33 mm rows spacing, 37 mm height
spacing, staggered arrangement. The outside diameter of the tubes
was 11.3 mm, aluminium plain fins with 0.2 mm thickness were
used (fin spacing 1.586 mm). The length of the tubes was
495.3 mm. The Authors report that 5 well-known refrigerants
were used. They investigated different inclination angles. Refer-
ring to horizontal installation, they found that the heat transfer
effectiveness was almost insensible to inclinations for angles from
the horizontal up to 9 when the cold side of the heat exchanger is
elevated, whereas the effectiveness suddenly drops even for very
small inclination angles when the hot side is elevated.
Unfortunately, the Authors didn’t report the type of tube inner
surface and the presence or not of a wick, the type of fluids and
the filling ratio.
The use of longitudinal (axial) open fins (or grooves) inside heat
pipes is well documented in literature (see for example, Ref. [8]).
Recently, Yong et al. [9] investigated the performance of heat pipes
with axially oriented micro-grooves produced by means of an
extrusion-ploughing process. They studied one single horizontal
heat pipe, operating with acetone and having the inner diameter of
5.2 mm (210 mm long). Other types of internal micro-grooves have
been investigated in the open literature like, for example, helicoidal
3-D grooves [10], axial “U” shaped [11], but none of the above
mentioned works refers to HPHE.
It is well known [8] that the choice of heat pipe working fluid
markedly affects the heat pipe effectiveness. HFC134a is widely
adopted as working fluids for HPHE since it is quite widespread as a
refrigerant for refrigeration and air conditioning units including
finned coil condensers and evaporators. HFC134a can be used for
working temperatures down to 25  C or so and up to 90  C or so
(being 101.06  C the critical temperature). Hassan [12] investigated
the performance of a HPHE operating with HFC134a, HCFC22,
HFC410A and HFC407C. The major characteristics of the analysed
HPHE were: 640 mm tube length, (310 mm evaporator length,
310 mm condenser length, 20 mm adiabatic sector length), 7 rows,
10 tubes each, tube outside diameter 20 mm, staggered arrange-
ment, (100-mesh brass wick, fin spacing 4 mm, plain fins with
0.5 mm thickness). He found that HFC407C displays the worst
performance: this is not surprising if one considers that it is a
zeotropic mixture and the additional mass transfer resistance
 G.A. Longo et al. / Applied Thermal Engineering 65 (2014) 361e368
decreases the heat transfer effectiveness. The other three investi-
gated fluids were found to outperform HFC407C.
One of the major constraints for future use of HFC134a is the
relatively high global warming potential (GWP ¼ 1430 over a 100-
years scenario) that could bring about possible future restrictions in
the use, as already ongoing for automotive air-conditioning in
Europe. One of the possible solutions is represented by HFOs
(hydrofluoro-olefins). In particular HFO1234yf (proposed for auto-
motive air conditioning) having GWP ¼ 4 and HFO1234ze(E)
(proposed for stationary air conditioning and large water re-
frigerators) having GWP ¼ 6. These fluids are not toxic and are
considered mildly-flammable by ASHRAE and so they are prefer-
able to hydrocarbons ad ammonia in terms of safety.
Fluids suitable for heat pipe operation should display latent heat
of vaporisation as large as possible, high surface tension, low liquid
viscosity.
In Table 1, the most relevant properties of HFC134a, HFO1234yf
and HFO1234ze(E) are reported at saturation temperatures of 10
and 50  C, as calculated with Refprop [13].
In the mentioned table, also the figure of merit number is
evaluated according to Reay and Kew [8]:
M ¼ rl sr=ml (1)
where rl and ml are the density and the dynamic viscosity of the
liquid phase, s is the surface tension and r the latent heat of
vaporisation of the refrigerant. From the results in Table 1,
HFO1234ze(E) is surely a possible candidate to substitute HFC134a.
Basing on the above literature survey, in the present paper the
experimental and theoretical analyses of a 6 rows HPHE are re-
ported. The heat pipes are horizontal copper tubes with internal
helicoidal continuous micro-fins. The performance of the heat
exchanger operating with HFC134a is compared with the results
obtained using HFO1234ze(E). Tests were run in a double wind
channel with air operating conditions typical for air-conditioning in
European countries for both summer and winter seasons. The
consistency of the experimental results is proved by the compari-
son with a semi-empirical model of the heat exchanger based on
heat transfer correlations available in literature.
2. Experimental set-up, procedures and data reduction
The experimental rig, shown in Fig. 1, consists of two air lines,
the exhaust air and the supply air, linked by a HPHE. In each line
ambient air supplied by a variable speed fan can be cooled and
dehumidified in the fin and coil heat exchanger, heated by electric
heating elements and humidified by steam injection to achieve the
set conditions at the inlet of the tested HPHE. The exhaust and the
supply air streams go through the tested heat exchanger where the
heat transfer takes place and then they are discharged. Both the air
lines contain two measurement sections located at the inlet and at
the outlet of the regenerative heat exchanger to measure temper-
ature and humidity ratio. Each measuring section consists of four T-
type thermocouples (uncertainty (k ¼ 2) within 0.1 K) and a
Table 1
Relevant thermodynamic and thermophysical properties of HFC134a, HFO1234yf and HFO1234ze(E).
Fluid T ( C) ps (bar) rl (kg m3) ll (W m2 K1)
HFC134a 10 4.15 1261.0 0.088
HFO1234yf 10 4.38 1144.0 0.072
HFO1234ze(E) 10 3.08 1210.4 0.083
HFC134a 50 13.18 1102.3 0.070
HFO1234yf 50 13.02 990.4 0.062
HFO1234ze(E) 50 9.97 1073.8 0.071
363
humidity tap connected to a dew point temperature probe (un-
certainty (k ¼ 2) within 0.2 K). The pressure drops on both the
sides of the heat exchanger are measured by differential strain-
gauge pressure transducers (uncertainty (k ¼ 2) within 0.1% f.s.),
whereas the air flow rates are measured by a vortex-shedding flow
meters (uncertainty (k ¼ 2) within 1.0% of the reading.). A barom-
eter (uncertainty (k ¼ 2) within 0.08% f.s.) measures the absolute
atmospheric pressure. The refrigerant pressure inside the third row
of heat pipes is measured by an absolute strain-gauge pressure
transducer (uncertainty (k ¼ 2) within 0.075% f.s.). All the readings
are scanned and measured by a data acquisition system consisting
of a 20 channel switch unit and a 6½ digit multimeter (uncertainty
(k ¼ 2) within 2.7 mV). Table 2 gives the specification of the
different measuring devices in the experimental rig.
The HPHE consists of 54 horizontal 12.7 mm micro-fin copper
tubes 700 mm in length and aluminium continuous fins 0.115 mm
in thickness with 3 mm of fin spacing. The HPHE presents an
evaporator section of 270 mm, a condenser section of 270 mm and
adiabatic section of 160 mm with a 40 mm polyurethane insulation
panel which separates completely the exhaust air flow from the
supply air flow. All the tubes are filled up with a two-phase fluid: in
present experimentation HFC134a and HFO1234ze(E) refrigerants
were tested. The refrigerant charge was set in accordance with the
recommendation by Reay and Kew [8] for a proper charge of the
heat pipes. The heat pipes are perfectly horizontal and only the
capillary forces acting along the spiral grooves of the micro-fin in-
tube surface govern the back flow of the condensate. Fig. 2 and
Table 3 provides the main geometrical characteristics of the HPHE
tested.
Temperature, humidity ratio and flow rate of the exhaust and
supply air at the inlet of the regenerative heat pipe heat exchanger
are adjusted to obtain the set inlet conditions. Once steady state
conditions in temperature, humidity ratio and flow rate at the heat
pipe heat exchanger inlet and outlet on both the supply and
exhaust air sides are reached, all values are recorded for a set time
after which an average value is computed for each parameter
recorded. From the average values of the measurements recorded
during the steady state conditions, it is possible to compute the
following characteristic parameters:
2.1. Sensible heat recovery Q






Qsupply ¼ msupply cp$supply
Tsupply$out eTsupply$in
(2)
Qexh ¼ mexh cp$exh jTexh$out eTexh$in j (3)
 .
Q ¼ Qsupply þ Qexh 2 (4)
where msupply and mexh are the supply and the exhaust air mass
flow rates measured by the vortex-shedding flow meter, cp$supply
and cp$exh are the supply and the exhaust air specific heat capacities
ml (Pa s) s (N m1) r (kJ kg1) M (N m2)
2.35E-04 1.01E-02 190.74 1.04Eþ10
1.86E-04 8.04E-03 156.60 7.76Eþ09
2.38E-04 1.08E-02 177.53 9.76Eþ09
1.42E-04 4.89E-03 151.82 5.77Eþ09
1.15E-04 3.35E-03 122.13 3.53Eþ09
1.48E-04 5.74E-03 145.55 6.06Eþ09
364 G.A. Longo et al. / Applied Thermal Engineering 65 (2014) 361e368

Fig. 1. Schematic view of the experimental rig.

and Tsupply$in and Tsupply$out and Texh$in and Texh$out are the supply h i
and the exhaust air temperatures at the inlet and outlet of the Cmax ¼ MAX msupply cp$supply ; mexh cp$exh (7)
regenerative heat exchanger measured by thermocouples. During
the experimental tests the average absolute misbalance between and the number of transfer units
the supply and the exhaust air side is around 3.5%.
NTU ¼ UA=Cmin (8)
2.2. Sensible thermal effectiveness ε
UA ¼ Q =DTln (9)
h

i.


h   i.
ε ¼ MAX
Tsupply$out eTsupply$in
; jTexh$out eTexh$in j


(5) DTln ¼ Tsupply$in eTexh$out  Tsupply$out eTexh$in


h . i (10)

Tsupply$in eTexh$in

ln Tsupply$in eTexh$out Tsupply$out eTexh$in
which depends on the ratio between the heat capacity rates
h i
Cmin ¼ MIN msupply cp$supply ; mexh cp$exh (6)
2.3. Fanning friction factor f

Table 2 h i
Specification of the different measuring devices. fsupply=exh ¼ Dpsupply=exh = 4ðD=dh Þrsupply=exh w2supply=exh =2
Devices Type Uncertainty Range
(11)
(k ¼ 2)

Thermometers - T-type thermocouple 0.1 K 20/80  C where Dpsupply and Dpexh are the pressure drop measured by the
Hygrometer - Chilled mirror sensor 0.2 K 50/50  C
differential pressure transducers, D ¼ 156 mm is the depth and
Air side differential - Strain-gauge 0.10% f.s. 0e40 mbar
pressure transd.
dh ¼ 3.561 mm the hydraulic diameter of the fin and coil, rsupply and
Air flow meter - Vortex-shedding 1.0% 400e1000 m3 h1 rexh the average density and wsupply and wexh the incoming face
reading. velocity of the supply and exhaust air. The Fanning friction factor f
Barometer - Strain-gauge 0.1% f.s. 800e2000 mbar depends on the Reynolds number:
Refrigerant absolute - Strain-gauge 0.05% f.s. 0/1.6 MPa
pressure
transducer Resupply=exh ¼ rsupply=exh wsupply=exh dh =msupply=exh (12)
Acquisition system - 20 channel switch 2.70 mV 0/100 mV
unit where msupply and mexh are the average dynamic viscosity of the
- 6½ digit multimeter
supply or the exhaust air.
 G.A. Longo et al. / Applied Thermal Engineering 65 (2014) 361e368 365

Fig. 2. Heat pipe heat exchanger (HPHE) (measures in mm).

2.4. Average saturation temperature of the refrigerant within the Standard 410-2001 [14] and the ANSI/ASHRAE Standard 33-2000
pipes [15] for testing air-to-air heat exchangers.

 
Tref ¼ Tsat pref (13)
3. Analysis of the experimental results and comparison with
theoretical models
where pref is the refrigerant pressure measured by the absolute
pressure transducer installed on the third row of pipes. The
Two different sets of experimental tests are carried out on the
refrigerant properties are evaluated by Refprop 9.0 [13].
HPHE with both refrigerant HFC134a and refrigerant HFO1234ze(E)
This work refers directly to the sensible heat recovery and the
as two-phase fluid within the pipes.
sensible thermal effectiveness, as, in all the experimental tests
The first set includes 26 runs with HFC134a and 26 runs with
carried out, the operating conditions on both the supply and the
HFO1234ze(E) in which the exhaust air inlet is at 25 C  0.1  C with
exhaust air side are always far from saturation conditions.
a relative humidity around 55%, whereas the supply air inlet is at
The instrumentation, the procedures and the data reduction
35 C  0.1  C and 40 C  0.1  C with a relative humidity between
techniques used in present work are consistent with the AHRI
26 and 30%. Both the experimental tests are carried out with the
same air volumetric flow rates on the exhaust and supply line,
Table 3
within the range 400e1000 m3 h1. Further tests were run with
Geometrical characteristics of the heat pipe heat exchanger.
different volumetric flow rates on the “hot” and “cold” side of the
Enhanced tube type Micro-fin HPHE. This set of tests is aimed to investigate the summer mode
Two-phase fluid HFC134a, HFO1234ze(E)
operating conditions.
Tube material Copper
Tube outside diameter dext (mm) 12.70 The second set includes 24 runs with HFC134a and 24 runs with
Number of grooves ng 65 HFO1234ze(E) in which the exhaust air inlet is at 20  C  0.1  C with
Groove depth hg (mm) 0.20 a relative humidity around 60%, whereas the supply air inlet is at
Inclination angle of the groove bg ( ) 18 7  C  0.1  C and 10  C  0.1  C with a relative humidity between 80
Total tube length Lt (mm) 700
Effective tube length Lt0 (mm) 540
and 90%. This set of tests is aimed to investigate the winter mode
Number of rows nr 6 operating conditions. The investigated air flow rates are the same as
Number of tubes per row nt 9 in “summer” tests.
Tube pitch lt (mm) 30 Table 4 shows the operating conditions during experimental
Row pitch lr (mm) 26
tests: exhaust air inlet temperature Texh$in, humidity ratio Xexh$in,
Fin type Corrugated
Fin material Aluminium flow rate Gexh, supply air inlet temperature Tsupply$in, humidity ratio
Fin thickness sf (mm) 0.115 Xsupply$in, flow rate Gsupply, heat recovery Q and average refrigerant
Fin pitch lf (mm) 3 temperature Tref.
Supply air side dimension (mm) 270  270 A detailed error analysis, performed in accordance with [16],
Exhaust air side dimension (mm) 270  270
indicates an overall uncertainty within 4.4% for the heat recovery
366 G.A. Longo et al. / Applied Thermal Engineering 65 (2014) 361e368

Table 4
Operating conditions during experimental tests.

Set Fluid Mode Runs Texh$in ( C) Xexh$in Vexh (m3 h1) Tsupply$in ( C) Xsupply$in Vsupply Q (W) Tref ( C)
(g kg1) (g kg1) (m3 h1)

1st HFC134a Summer 26 25 4.6e9.8 418e1007 35 and 40 4.5e9.1 421e1004 565e1615 28.7e32.6
1st HFO1234ze(E) Summer 26 25 9.4e12.7 433e1002 35 and 40 9.5e13.0 423e1004 586e1593 28.5e32.5
2nd HFC134a Winter 24 20 4.3e10.8 406e1000 7 and 10 4.3e9.1 419e997 592e1557 12.6e15.8
2nd HFO123ze(E) Winter 24 20 7.3e9.7 434e1008 7 and 10 5.0e7.3 430e1005 620e1548 12.6e15.9

measurement and within 8.0% for air side pressure drop
measurement.
Fig. 3 shows experimental thermal effectiveness vs. the number
of transfer units and the ratio between the heat capacity rates: for
comparison the theoretical trends for a counter-flow heat
exchanger with a capacity rate ratio of 0.5 and 1.0. are also reported.
There is no difference in thermal effectiveness between the
experimental tests with refrigerant HFC134a and those with
refrigerant HFO1234ze(E). So HFO1234ze(E) can be surely consid-
ered an environmental friendly substitute for HFC134a in HPHE.
Moreover all the experimental results are consistent with the
theoretical trends confirming that a fin and coil with more than 4
rows of tubes is equivalent to a counter-flow heat exchanger [17].
Fig. 4 shows the experimental Fanning friction factor vs. the
Reynolds number together with the trends predicted by Kays &
London [17] and Wang et al. [18] model. The data points with a
Reynolds number from 700 to 1700 are included in the interval
between the Kays & London [17] model and the Wang et al. [18]
model and probably refer to turbulent flow, whereas the data
points with a Reynolds number lower than 700 exhibit a different
trend compatible with a laminar or a transition flow.
Present heat transfer data has been compared against a
computational procedure in which, the heat exchanger is divided
into a series of ranks.
The enhancement in two-phase heat transfer inside micro-fin
tubes is partly due to the simple increase in the effective heat
transfer area, and also to the turbulence and wavy effects induced
in the condensate film by the micro-fins and to the surface tension
effects on condensate drainage. The first effect is effective both for
condensation and vaporisation, whereas the second and third ef-
fects are relevant only for condensation.
Within the tested operating conditions, the condensate film
Reynolds number
Ref ¼ 2mc =ml Lc (14)
was found to be always lower than 100. Accordingly, the conden-
sation heat transfer coefficient was calculated with reference to the
classical Nusselt theory [19] for laminar condensation on vertical
surface
nh i. o1=4
ac ¼ 0:943 rl ðrl erv Þg r0 l31 ½ml Lc ðTsat  Twall Þ (15)
r 0 ¼ r þ 3=8cp$l ðTsat eTwall Þ (16)
where rv is the density of the vapour phase, ll and cp$l the thermal
conductivity and the specific heat capacity of the liquid phase of the
refrigerant, Tsat and Twall the saturation temperature of the refrig-
erant and the wall temperature of the tube and Lc the length of the
condensation section of the tube. The constant was further incre-
mented by 30% to keep into account the wavy contribution created
by the tube fins (similar approach, with 20% increment was pro-
posed by McAdams [20], for smooth tubes). The actual heat transfer
area was considered as reference (i.e. keeping into account the
whole fins area and the base tube).
The heat transfer coefficient on the evaporation part was eval-
uated according to Imura et al. equation [21].

Fig. 3. Experimental thermal effectiveness vs. number of transfer units and the ratio of the heat capacity rates.
 G.A. Longo et al. / Applied Thermal Engineering 65 (2014) 361e368 367

Fig. 4. Experimental Fanning friction factor vs. Reynolds number.

coil heat exchangers having the same geometrical characteristics

h . i
produced from the same manufacturer.
ae ¼ 0:32 l0:3
1 r1
0:65 0:7 0:2 0:4
cp$l g qe r0:25
v r 0:4 m0:1
l ðpsat =patm Þ
Air temperature, relative humidity and flow rate at the inlet of
(17) the condenser and the evaporator sections are input data. The
actual geometry of the heat exchanger (including among others the
where qe is the heat flux in the evaporation section of the tube, psat tube diameter and length, fin spacing, thickness, length and height)
and patm the saturation pressure of the refrigerant and the absolute is an input as well.
atmospheric pressure. This approach was used also by Noie [22] to The simulation proceeds by iterating on the saturation pressure
fit his data inside a thermosyphon. The actual heat transfer area of the fluid inside the heat pipes; given the subdivision of the HPHE
(including fins) was considered for heat flow rate calculations. in ranks, the heat flow rate is assumed to be the same in each heat
The air-side heat transfer coefficients are assumed to be uniform pipe on the same rank. A further iteration is on the overall heat
and are evaluated through the Gray and Webb [23] correlation. The exchanger: it starts by giving a guess value for the cold air at the
air side pressure drops are evaluated according to Wang et al. [18] heat exchanger exit. Each single rank is then simulated starting
correlation. The models of Gray and Webb [23] and of Wang et al. form the first one on the hot air direction. The output values for the
[18] were chosen because they best fit experimental data of finned first rank are then the input for the second one etc. The iteration

Fig. 5. Comparison between calculated and experimental heat flow rate. Fig. 6. Comparison between calculated and experimental refrigerant temperatures.
368 G.A. Longo et al. / Applied Thermal Engineering 65 (2014) 361e368
Fig. 7. Comparison between calculated and experimental air side pressure drops.
ceases when the saturation temperature inside each rank allows
the calculated exit temperature of the coil air to suitably approxi-
mate the initial guess value.
Fig. 5 shows the comparison between the experimental heat
flow rate exchanged and the values calculated by the above
computational procedure: the mean absolute percentage deviation
between calculated and experimental heat flow rate is around 3.3%.
It should be noted that the thermal resistance is mainly concen-
trated on the air side, therefore the accuracy in the prediction of the
heat flow rate is mainly due to the accuracy in the estimation of the
air side heat transfer coefficient.
Fig. 6 shows the comparison between the experimental average
saturation temperature of the refrigerant inside the pipes derived
from the refrigerant pressure measurement on the third row and
the calculated values by the above computational procedure: the
mean absolute percentage deviation between calculated and
experimental saturation temperature is around 0.4  C. The good
agreement between experimental and calculated values confirms
the ability of the model in simulating the HPHE considered in this
paper both with refrigerant HFC134a and HFO1234ze(E).
Present pressure drop data in turbulent flow regime (Re > 700)
has been compared against Wang et al. [18] model. Fig. 7 shows the
comparison between experimental and calculated pressure drop:
the mean absolute percentage deviation between calculated and
experimental values is around 7.1% and Wang et al. [18] equation in
general over-predicts the data.
4. Conclusion
The experimental data obtained during the tests inside a double
wind channel of a HPHE indicates that it is possible to achieve good
heat transfer effectiveness even in horizontal installation of the
heat pipes when the internal surface of the pipes presents heli-
coidal micro-fins.
A numerical approach based on the Nusselt [19] model for film
condensation, on the Imura et al. [21] model for pool boiling in
thermosyphons and using the Gray and Webb [23] correlation for
air side heat transfer coefficient was able to satisfactorily capture
the experimental measurements.
When the low global warming potential refrigerant HFO1234-
ze(E) is used, the performance of the heat exchanger is equivalent
to the one obtained under the same operating conditions with the
traditional HFC134a refrigerant. So HFO1234ze(E) can be consid-
ered a viable environmentally friendly solution as two-phase fluid
in heat pipe heat exchangers.
References
[1] S. Riffat, X. Ma, Recent developments in heat pipe technology and applica-
tions: a review, Int. J. Low Carbon Technol. 2 (2007) 162e177.
[2] W. Srimuang, P. Amatachaya, A review of the applications of heat pipe heat ex-
changers for heat recovery, Renew. Sustain. Energy Rev. 16 (2012) 4303e4315.
[3] Y.H. Yau, M. Ahmadzadehtalatapeh, A review on the application of horizontal
heat pipe heat exchangers in air conditioning systems in the tropics, Appl.
Therm. Eng. 30 (2010) 77e84.
[4] D.A. Reay, A. Harvey, The role of heat pipes in intensified unit operations, Appl.
Therm. Eng. 57 (2013) 147e153.
[5] R.W. MacGregor, P.A. Kew, D.A. Reay, Investigation of low global warming
potential working fluids for a closed two-phase thermosyphon, Appl. Therm.
Eng. 51 (2013) 917e925.
[6] M.A. Abd El-Baky, M.M. Mohamed, Heat pipe heat exchanger for heat recovery
in air conditioning, Appl. Therm. Eng. 27 (2007) 795e801.
[7] P. Guo, D.L. Ciepliski, R.W. Besant, A testing and HVAC design methodology for
air-to-air heat pipe heat exchangers, HVAC&R Res. 4 (1998) 3e26.
[8] D.A. Reay, P.A. Kew, Heat Pipes, fifth ed., ButterwortheHeinemann, Oxford,
2006.
[9] T. Yong, C. Peng, W. Xiaowu, Experimental investigation into the performance
of heat pipe with micro grooves fabricated by extrusion-ploughing process,
Energy Convers. Manag. 51 (2010) 1849e1854.
[10] Q. Liao, T.C. Jen, Q. Chen, L. Li, W. Cui, Heat transfer performance in 3D
internally finned heat pipe, Int. J. Heat. Mass Transf. 50 (2007) 1231e1237.
[11] Y. Chen, C. Zhang, M. Shi, J. Wu, G.P. Peterson, Study on flow and heat transfer
characteristics of heat pipe with axial ‘‘U”-shaped microgrooves, Int. J. Heat.
Mass Transf. 52 (2009) 636e643.
[12] M.A.M. Hassan, Investigation of performance of heat pipe as heat exchanger
using alternative refrigerants, J. Energy Eng. 139 (2013) 18e24.
[13] E.W. Lemmon, M.L. Huber, M.O. McLinden, NIST Standard Reference Database
23: Reference Fluid Thermodynamic and Transport Properties-REFPROP,
Version 9.0, NIST, Gaithersburg, MD, 2010.
[14] AHRI Standard 410-2001, Forced-circulation Air-cooling and Air-heating Coils,
AHRI, Arlington, VA, USA, 2001.
[15] ANSI/ASHRAE Standard 33-2000, Method of Testing Forced Circulation Air
Cooling and Air Heating Coils, ASHRAE, Atlanta, GA, USA, 2000.
[16] S.J. Kline, F.A. McClintock, Describing uncertainties in single-sample experi-
ments, Mech. Eng. 75 (1954) 3e8.
[17] W.M. Kays, A.L. London, Compact Heat Exchangers, third ed., Krieger Pub-
lishing Company, Malabar, FL, USA, 1984.
[18] C.C. Wang, K.Y. Chi, C.J. Chang, Heat transfer and friction characteristics of
plain fin-and-tube heat exchangers, part II: correlation, Int. J. Heat. Mass
Transf. 43 (2000) 2693e2700.
[19] W. Nusselt, Die Oberflachenkondensation des Wasserdampfes, Z. Ver. Dt. Ing.
60 (1916) 541e575.
[20] W.H. McAdams, Heat Transmission, third ed., McGraw-Hill, New York, 1954.
[21] H. Imura, H. Kusuda, J. Ogata, T. Miyazaki, N. Sakmoto, Heat transfer in two-
phase closed-type thermosyphons, Heat Transf. Jpn. Res. 8 (1979) 41e53.
[22] S.H. Noie, Heat transfer characteristics of a two-phase closed thermosyphon,
Appl. Therm. Eng. 25 (2005) 495e506.
[23] D.L. Gray, R.L. Webb, Heat transfer and friction correlations for plate finned-
tube heat exchangers having plain fins, in: Proceeding 8th International
Journal of Heat Transfer Conference IHTC-8 6, 1986, pp. 2745e2750.