You are on page 1of 10

Applied Thermal Engineering 66 (2014) 580e589

Contents lists available at ScienceDirect

Applied Thermal Engineering


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

Airside performance of oval tube heat exchangers having sine


wave ns under wet condition
Nae-Hyun Kim*, Kang-Jong Lee, Yeong-Bin Jeong
Department of Mechanical Engineering, University of Incheon, Incheon, Republic of Korea

h i g h l i g h t s

 Airside performances of oval tube heat exchangers investigated under wet condition.
 Different from round tube samples, the lowest j factor obtained for one row geometry.
 Oval tube samples yield superior performance compared with round tube samples.
 Oval geometry is more benecial under wet condition than under dry condition.

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

Article history: Experiments were conducted on sine wave n-and-tube heat exchangers having oval tubes under wet
Received 16 August 2013 condition. Twelve samples having different n pitches (2.12 mm, 2.54 mm, 3.18 mm) and tube rows (one
Accepted 18 February 2014 to four) were tested. Eight herringbone wave n-and-tube heat exchangers having round tubes were also
Available online 12 March 2014
tested. Results showed that, for oval tube samples, the effect of n pitch on j and f factor was not sig-
nicant. As for the effect of tube row, the lowest j factor was obtained for one row conguration, which is
Keywords:
clear contrast to round tube samples, where the highest j factor was obtained for one row conguration.
Oval tube
Possible reasoning is provided considering the ow and heat transfer characteristics of sine wave channel
Heat exchanger
Wet surface
combined with connecting oval tubes. Oval tube samples yielded superior airside performance than
Sine wave n round tube samples when both heat transfer and pressure drop were considered. It is also shown that
oval geometry is more benecial under wet condition than under dry condition.
2014 Elsevier Ltd. All rights reserved.

1. Introduction heat exchangers having round tubes including design correlations.


Investigations include both dry and wet conditions. Interested
Fin-and-tube heat exchangers having relatively large diameter readers are encouraged to consult monographs and review articles
tubes (12.7 mm or 15.9 mm O.D.) are commonly used as chilled on this issue [2e4]. Most of the investigations, however, have been
water cooling coils in air handling units of building air conditioning conducted for samples having small diameter tubes with residen-
system. For a n pattern, either plain or wave n is widely used. tial application in mind [4e7]. Recently Wang et al. [8,9] investi-
Wave n enhances the heat transfer with simultaneous increase of gated the heat transfer and pressure drop characteristics of
pressure drop. Depending on the operating condition (inlet air dry herringbone wave n-and-tube heat exchangers having large
bulb temperature and humidity, cooling water temperature etc.), diameter tubes (Dc 16.6 mm) under dry or wet condition. Under
cooling coils could be dry, fully wet or partially wet. To analyze the dry condition, the effect of tube row on j factor was not prominent
cooling coil properly for a given operating condition, both dry and for tube rows less than four. However, for tube rows larger than
wet surface heat transfer coefcients are needed [1]. Literature four, noticeable decrease of j factor was observed. Under wet con-
shows many investigations on airside performance of n-and-tube dition, however, the effect of tube row on j factor was not
prominent.
It is well known that usage of oval tube instead of round tube
reduces the air-side pressure drop. Low thermal performance re-
* Corresponding author. Divisionment of Mechanical System Engineering,
gion downstream of the tube is also reduced. Despite of the
Incheon National University, 12-1, Songdo-Dong, Yeonsu-Gu, Incheon 406-772,
Korea. Tel.: 82 32 835 8420; fax: 82 32 835 8410. apparent advantage of oval tube heat exchanger over round tube
E-mail address: knh0001@incheon.ac.kr (N.-H. Kim). heat exchanger, the implementation was rather slow due to the

http://dx.doi.org/10.1016/j.applthermaleng.2014.02.042
1359-4311/ 2014 Elsevier Ltd. All rights reserved.
N.-H. Kim et al. / Applied Thermal Engineering 66 (2014) 580e589 581

Nomenclature V velocity [m s1]


V_ volume [m3]
A heat transfer area [m2] Wf corrugation depth from peak to valley [m]
a major diameter [m] x distance [m]
b minor diameter [m] or slope of enthalpyetemperature xf projected n pattern length for one-half wave length
curve of saturated air [kJ kg1 K1] [m]
C mass ow rate ratio x* non-dimensional distance (xP1f Re
1
)
cp specic heat [J kg1 s1]
Dc tube diameter including n collar thickness [m] Greek notations
f friction factor thermal effectiveness
h heat transfer coefcient [W m2 K1] DP pressure loss [Pa]
j Colburn j factor h n efciency
k thermal conductivity [W m1 K1] h surface efciency
m _ mass ow rate [kg s1] r density [kg m3]
N number of tube row m dynamic viscosity [kg m1 s1]
NTU number of transfer units s contraction ratio of the cross-sectional area
p ratio of outer and inner diameter
Pd n depth, peak to valley excluding n thickness [m] Subscripts
Pf n pitch [m] a air
Pt transverse tube pitch [m] i tubeside
Pl longitudinal tube pitch [m] in inlet
Pw wave pitch [m] f n
Pr Prandtl number m mean or airesteam mixture
q ratio of major and minor diameter max maximum
rc tube radius including n collar [m] min minimum or minor
Req equivalent radius [m] o airside
ReDc Reynolds number based on Dc out outlet
t tube wall thickness [m] r tubeside
T temperature [K] t tube wall
tf n thickness [m] w water or wet
U overall heat transfer coefcient [W m2 K1] wc wet sensible

difculty of manufacturing (especially tube expansion) and the heat exchanger. Han et al. [13] also reported a similar trend. They
concern on deformation under high internal pressure. With the numerically investigated the thermal performance of two row
advancement of expansion technology, however, oval tube heat herringbone wave or louver n-and-tube heat exchanger having
exchangers have been implemented in air handling units, where oval tubes (0.43 aspect ratio), and compared the results with those
internal pressure is not a concern. Literature shows that limited of two types of round tube heat exchangers (one having same
studies are available on n-and-tube heat exchangers having oval perimeter and the other having same hydraulic diameter with the
tubes. Saboya and Saboya [10], using naphthalene sublimation oval tube). Oval tube heat exchanger yielded the lowest pressure
technique, measured mass transfer coefcients of a plate channel drop and the highest heat transfer coefcient. Additional numerical
equipped with oval tubes of 0.5 or 0.65 aspect ratio, and compared study by Tao et al. [14], revealed that, when round tubes of n-and-
the results with those of a plate channel equipped with round tube heat exchanger were replaced with oval tubes (0.6 aspect ra-
tubes. When the round tube diameter was the same as that of the tio) having the minor tube diameter same as the round tube
minor diameter of oval tube, mass transfer coefcients were diameter, signicant increase of heat transfer coefcient was
approximately the same. When the minor diameter of oval tube possible with marginal increase of pressure drop.
was smaller than that of the round tube, the mass transfer coef- Compared with numerical studies, experimental investigations
cient of the channel with oval tube was smaller. The reason was are very limited. Matos et al. [15] tested four 4 row plain n-and-
attributed to smaller air velocity around oval tubes than that tube heat exchangers having oval tubes of different aspect ratio
around round tubes at the same frontal air velocity. (from 1.0 to 0.4), all with the same minor diameter. Aluminum ns
Literature shows several numerical investigations on n-and- were attached to oval tubes, although no detailed description on
tube heat exchangers having oval tubes. Min and Webb [11] the attaching method was provided. During the test in a wind
numerically investigated the thermal performance of herringbone tunnel, heat was supplied to tubes through heaters installed inside
wave n-and-tube heat exchangers having oval tubes. The oval of oval tubes. An optimum conguration was obtained (aspect ratio
shape was progressively varied from 1.0 to 0.23 maintaining the of 0.5), which exhibited a heat transfer gain of 19% relative to the
same perimeter. At a frontal velocity of 2.0 m s1, oval tube heat round tube counterpart. Accompanying numerical investigation
exchanger of 0.33 aspect ratio yielded 6.9% lower heat transfer also yielded optimum tube and n spacing for an oval tube heat
coefcient and 45.9% lower pressure drop than those of the round exchanger. The investigated range of frontal air velocity, however,
tube heat exchanger. Similar results were reported by Leu et al. [12] was very low (from 0.1 to 0.13 m s1). Kim et al. [16,17] tested sine
through the numerical study on two row louver n-and-tube heat wave n-and-tube heat exchangers having oval tubes of 0.6 aspect
exchanger having oval tubes of 0.36 aspect ratio. They reported that ratio. Twelve samples had different n pitches and tube rows.
oval tube heat exchanger yielded 10% lower heat transfer coef- Comparison with herringbone wave n-and-tube heat exchangers
cient and 41% lower pressure drop than those of the round tube [16] or plain n-and-tube heat exchangers [17] with round tubes
582 N.-H. Kim et al. / Applied Thermal Engineering 66 (2014) 580e589

showed that oval tube heat exchangers generally yielded superior


thermal performance.
The above literature survey reveals that experimental in-
vestigations on thermal performance of n-and-tube heat ex-
changers having oval tubes are very limited. Especially,
experimental studies under wet condition are rarely available for
oval conguration. This study is a continuing effort following Kim
et al. [16], whose investigation was focused on thermal perfor-
mance of n-and-tube heat exchangers having oval tubes under dry
condition. In this study, tests were conducted using the same oval
tube heat exchangers of 0.6 aspect ratio (minor diameter 10.0 mm,
major diameter 16.5 mm) tested by Kim et al. [16], however under
wet condition. Twelve samples having different n pitches
(2.12 mm, 2.54 mm, 3.18 mm) and tube rows (one to four) were
tested. Eight herringbone wave n-and-tube heat exchangers
having round tubes of two different diameters (12.7 mm and
15.9 mm) were also tested. The round tube samples had different
tube rows (one to four) at xed n pitch (2.54 mm). The objectives
of the present study are rstly to investigate the heat transfer and
friction characteristics of oval tube heat exchangers under wet
condition, and secondly to experimentally conrm superior airside
performance of oval tube conguration over round tube
conguration.

2. Experiments

2.1. Heat exchanger samples

Schematic drawings of the sine wave n and herringbone wave


n are shown in Fig. 1. Note that oval tubes are used with the sine
wave n, and round tubes are used with the herringbone wave n.
As shown in Fig. 1(a), sine wave ns have at section (with two
small grooves) between rows. The wave pitch and depth of the sine
wave n are 15.2 mm and 1.1 mm respectively. Geometric details of
the oval tube heat exchanger are listed in Table 1. The major and
minor diameter of oval tubes are 16.5 mm and 10.0 mm, transverse
tube pitch (Pt) is 35.0 mm, longitudinal tube pitch (Pl) is 30.3 mm.
The height and width of the sample are 350 mm and 440 mm
respectively. The n thickness is 0.14 mm. Oval tubes were made by
successive rolling of the 12.7 mm O.D. round tube. Oval tubes were
inserted into sine wave ns, and were expanded using specially-
made oval-shaped balls having slightly larger diameter than the
inner diameter of the oval tube. After expansion, tubes were cut at
several locations, and the contact between n and tube was visually
inspected, which turned out satisfactory.
Geometric details of the round tube heat exchangers are also
listed in Table 1, which show different dimensions depending on
the tube diameter. For samples with 15.9 mm O.D. tube, transverse
tube pitch is 38.1 mm and longitudinal tube pitch is 33.0 mm. The
wave pitch of the n is 8.25 mm and wave depth is 1.45 mm. The
height and width of the sample is 340 mm and 440 mm respec-
tively. For samples with 12.7 mm O.D. tube, transverse tube pitch is
31.8 mm and longitudinal tube pitch is 27.5 mm. The wave pitch of
the n is 6.88 mm and wave depth is 1.60 mm. The height and
width of the sample is 286 mm and 440 mm respectively. For both
geometries, n thickness is 0.12 mm. For all samples, ns were
hydrophilic treated, tube inner surface was smooth and tubes were
circuited cross-counter conguration with single inlet and outlet.

2.2. Test apparatus and procedures Fig. 1. Shape of the sine wave and herringbone n (unit: mm).

A schematic drawing of the apparatus is shown in Fig. 2. It condition of the heat exchanger is maintained by controlling the
consists of a suction-type wind tunnel, water circulation and con- chamber temperature and humidity. The inlet and outlet dry and
trol units, and a data acquisition system. The apparatus is situated wet bulb temperatures are measured by the sampling method as
in a constant temperature and humidity chamber. The airside inlet suggested in ASHRAE Standard 41.1 [18]. A diffusion bafe is
N.-H. Kim et al. / Applied Thermal Engineering 66 (2014) 580e589 583

Table 1 whose accuracy is 0.0015 L/s. The airside pressure drop across the
Geometric dimensions of the samples. heat exchanger is measured using a differential pressure trans-
Pt Pl Pw Wf Dc N Pf tf ducer. The air ow rate is measured using a nozzle pressure dif-
(mm) (mm) (mm) (mm) (mm) (mm) (mm) ference according to ASHRAE Standard 41.2 [19]. The accuracy of
Sine wave 35.0 30.3 15.2 1.10 16.5 1 3.18 0.14 the differential pressure transducers is 1.0 Pa.
35.0 30.3 15.2 1.10 16.5 2 3.18 0.14 During the experiment, the water temperature was held at
35.0 30.3 15.2 1.10 16.5 3 3.18 0.14 10  C. The chamber temperature was maintained at 32  C with 80%
35.0 30.3 15.2 1.10 16.5 4 3.18 0.14
relative humidity. Experiments were conducted varying the frontal
35.0 30.3 15.2 1.10 16.5 1 2.54 0.14
35.0 30.3 15.2 1.10 16.5 2 2.54 0.14 air velocity from 1.0 m s1 to 4.0 m s1. The energy balance between
35.0 30.3 15.2 1.10 16.5 3 2.54 0.14 the airside and the tube-side was within 3%. The discrepancy
35.0 30.3 15.2 1.10 16.5 4 2.54 0.14 increased as the air velocity decreased. All the data signals were
35.0 30.3 15.2 1.10 16.5 1 2.12 0.14
collected and converted by a data acquisition system (a hybrid
35.0 30.3 15.2 1.10 16.5 2 2.12 0.14
35.0 30.3 15.2 1.10 16.5 3 2.12 0.14
recorder). The data were then transmitted to a personal computer
35.0 30.3 15.2 1.10 16.5 4 2.12 0.14 for further manipulation. An uncertainty analysis was conducted
Herringbone 38.1 33.0 16.5 1.45 15.9 1 2.54 0.12 following ASHRAE Standard 41.5 [20], and the results are listed in
wave 38.1 33.0 16.5 1.45 15.9 2 2.54 0.12 Table 2. The major uncertainty on the friction factor was the un-
(15.9 mm O.D). 38.1 33.0 16.5 1.45 15.9 3 2.54 0.12
certainty of the differential pressure measurement (1.0 Pa), and
38.1 33.0 16.5 1.45 15.9 4 2.54 0.12
Herringbone 31.8 27.5 13.8 1.60 12.7 1 2.54 0.12 the major uncertainty on the heat transfer coefcient (or j factor)
wave 31.8 27.5 13.8 1.60 12.7 2 2.54 0.12 was that of the tube-side heat transfer coefcient (10%). The un-
(12.7 mm O.D.) 31.8 27.5 13.8 1.60 12.7 3 2.54 0.12 certainties decreased as the Reynolds number increased.
31.8 27.5 13.8 1.60 12.7 4 2.54 0.12

2.3. Data reduction

installed behind the test sample to mix the outlet air. The waterside The data reduction details are provided by Mirth and Ram-
inlet condition is maintained by regulating the ow rate and inlet adhyani [1], Pirompugd et al. [4] and Kim et al. [21] and short
temperature of the constant temperature bath situated outside of summary is provided here. For the cross-counter conguration of
the chamber. Both the air and the water temperatures are measured the present study, appropriate equations for the heat exchanger
by pre-calibrated RTDs (Pt-100 U sensors). Their accuracies are analysis are given by ESDU 98005 [22], and are summarized in
0.1 K. The water ow rate is measured by a mass ow meter, Table 3. The UA value is obtained from the following equations.

Fig. 2. Schematic drawing of the test setup.


584 N.-H. Kim et al. / Applied Thermal Engineering 66 (2014) 580e589

Table 2 where
Experimental uncertainties.
s
Parameter Max. Uncertainties 2hw
m (7)
Temperature 0.1 K kf tf
Differential pressure 1.0 Pa
Water ow rate 1.5  106 m3 s1    
2.4% Req Req
ReDmin
7.6%
f 1 1 0:35In (8)
f rc rc
j 11.1%

Req p
h i
e2:41 0:7 e2q ln0:7 q 3 1 row (9)
rc
UA Cmin NTU (1)
Req p
h i
e2:31 0:855 e:215q ln0:72 q 3 over 2 row
_ min =m
C m _ max (2) rc
(10)
For the one row conguration, a cross-ow eNTU equation of
mixed-unmixed conguration (Holman [23]) was used. The airside  0:5
Pl
heat transfer coefcient under wet condition (ho) is obtained from Req 0:64Pt  0:2 1 row (11)
the following equations. In Eq. (3), actual surface area considering Pt
the corrugation was used as airside heat transfer area (Ao).
 0:5
Pl
bw;m 1 br bt t Req 0:635Pt  0:3 over 2 row (12)
  (3) Pt
ho hw Ao UA hi Ai kAt
h pi
hw cpm p 2Req = 1:5a b  ab (13)
ho (4)
bw;m
q a=b (14)
Here, br,bt,bw,m are the slope of saturated air enthalpy at the
water, tube wall, and water lm temperature respectively. The here, q is the ratio of major and minor diameter of the oval tube,
tube-side heat transfer coefcient is obtained from the Gnielinski and p is the ratio of outer and inner diameter of equivalent annular
[24] correlation. For an accurate assessment of the airside heat n. Min et al. [25] have shown that the n efciency calculated
transfer coefcient, it is important to minimize the tube-side using above correlation is within 4% error when compared with
thermal resistance. Throughout the experiment, the tube-side that calculated using exact sector method. For round tube samples,
thermal resistance was less than 10% of the total thermal resis- same equations from (6) to (8) are applicable with Req/rc obtained
tance. The surface efciency ho for use in Eq. (3) is obtained from Eq. from the well-known Schmidt equation [26].
(5).
0q 10:5
2
Pt @ Pt =2 Pl
2
A Req
ho 1  f 1  h (5) 0:64  0:2A 1 row (15)
Ao rc rc Pt
Fin efciency correlation for a n-and-tube heat exchanger
having oval tube has been provided by Min et al. [25]. 0q 10:5
2
Pt @ Pt =2 Pl
2
Req
tanhmrc f 0:635  0:3A over 2 row (16)
h (6) rc rc Pt
mrc f

Table 3
eNTU relationship for cross-counter conguration with single inlet and outlet.

Row

Cmin (air)
 
2row 1 K 1  exp(NTU/2)
1  K 1K 1exp2KC
2 2
C
 
3row K 1  exp(NTU/3)
C1 1  1
K 2
K 1 4 CK 1 2 expKC1 2 exp3CK
K K

 


3
4row 2
A K2 1  K2 K4 K 1  K2 1  2KC 1  K2 exp2KC 1  K2 exp4KC
1/C(1  1/A) K 1  exp(NTU/4)
Cmin (water)
 
2row 1 K 1  exp(NTU$C/2)
1  K
2 1K2 exp2KC
 
3row K 1  exp(NTU,C/3)
1 1
K 1K4 KC 1K2 expKC 1K2 exp3K
C
2

 

2K

4row A K 1  K2 K4
2
K 1  K2 1  2K K K 3 exp 4K
2 C 1  2 exp C 1  2 C
(1  1/A) K 1  exp(NTU,C/4)
N.-H. Kim et al. / Applied Thermal Engineering 66 (2014) 580e589 585

The heat transfer coefcient is expressed as the j factor with


minor diameter (Dmin) for oval tube or tube outer diameter (Dc) for
round tube as characteristic length [27].

ra Vmax Dmin
ReDmin (17)
ma

ho
j Pra (18)
ra Vmax cpa
The friction factor is obtained from Eq. (19).
" #
Ac rm 2DP rin   r
f  1s2 in
1 (19)
Ao rin rm Vmax
2 rout
In Eq. (19), the entrance and the exit loss coefcients are
neglected following the suggestion by Wang et al. [28].

3. Results and discussions

Fig. 3 shows the effect of n pitch for oval tube samples. As


mentioned previously, minor tube diameter (Dmin) was used as a
characteristic dimension of Reynolds number. As a representative
example, error bars on j and f factor are drawn for one row, 2.12 mm
n pitch sample [Fig. 3(a)]. The effect of n pitch on j and f factor is
not signicant. Torikoshi et al. [29] performed a three dimensional
unsteady numerical computation for one row plate n-and-tube
heat exchangers having various n pitches (from 1.7 mm to
3.0 mm for Dc 10 mm). As n pitch increased, downstream ow
eld became more unsteady. However, the ow eld in the region
between ns remained steady even at the largest n pitch. The heat
transfer on the n surface was also unaffected by the n pitch.
Although the numerical study was limited to a plain n congu-
ration, similar arguments may apply to other n geometries. The
independency of j factor with n pitch has also been reported by
other investigators [4,30,31]. The effect of n pitch on f factor from
the literature survey is inconclusive. Rich [30] reported that f factor
increased with the increase of n pitch for a plain n conguration.
Wavy n data of Wang et al. [31] yielded a rather complex trend.
Above a certain Reynolds number, f factor increased with the in-
crease of n pitch. Under that Reynolds number, however, f factor
decreased with the increase of n pitch. Liu et al. [9] reported that
the effect of n pitch on f factor was only marginal for a plain n-
and-tube heat exchanger under dehumidifying condition, which
is in line with the present results.
The effect of tube row on j and f factor of oval tube samples along
with those of round tube samples are shown in Fig. 4. For all
samples, n pitch is 2.54 mm. For round tube samples, j factor
decreases as number of tube row increases, although the effect is
not prominent for 15.9 mm O.D. samples for the samples having
tube rows larger than two. For n-and-tube heat exchangers, air
ows through channels formed by narrow spaced ns and con-
necting tubes. For channel ow, the heat transfer is the largest at
inlet of the channel, and decreases along the ow path. Then, j
factor decreases as number of tube row increases [4,30,32]. As
noted earlier, Wang et al. [8,9] investigated the heat transfer and
pressure drop characteristics of herringbone wave n-and-tube
heat exchangers (Dc 16.6 mm) having large number of tube
rows under dry or wet condition. Under dry condition, the effect of
tube row on j factor was not prominent for tube row less than four.
However, for tube row larger than four, noticeable decrease of j
factor was observed. Wang et al. [8] attributed the trend to the
developing characteristics of channel ow. Under wet condition, j
factor decreased as number of tube row increased. However, the
Fig. 3. Effect of n pitch on j and f factor for oval tube samples.
effect of tube row was not prominent, which agrees with the
586 N.-H. Kim et al. / Applied Thermal Engineering 66 (2014) 580e589

ns govern the momentum transfer process, and the effect of tube


row on f factor diminishes [4,30,31]. For one row conguration,
however, the aforementioned argument may not apply, because
tubes are no longer in staggered fashion. Then, the channel effect
may govern the momentum transfer process, and increases the f
factor. For oval tube or 12.7 mm O.D. round tube samples, f factor
decreases as number of tube row increases.
Fig. 4 shows that the row effect on j factors of oval tube samples
is signicantly different from that of round tube samples. The
lowest j factor is obtained for one row conguration. For other tube
rows, the difference is not signicant. Close examination of the
results, however, reveals that highest j factor is obtained for two
row conguration, and the j factor of four row conguration is
larger than that of three row conguration. The same trend has
been reported under dry condition for the same oval tube samples
[16]. As discussed previously, it is commonly observed that j factor
decreases as the number of tube row increases. The present results
are very peculiar, not reported before. The reason may be attributed
to present sine wave channel combined with connecting oval tubes.
The ow and heat transfer characteristics of sine wave channel has
been investigated by Rush et al. [33]. They performed visualization
tests in sine wave channels having different channel pitch and
depth. At Reynolds numbers (based on hydraulic diameter) less
than 200, the ow was smooth and no mixing was observed. At
Reynolds number around 200, ow mixing started at far end of the
channel, which propagated upstream as Reynolds number
increased. In Fig. 5, local Nusselt numbers measured by Rush et al.
[33] for the wave channel (Wf/Pf 0.48 and Pw/Pf 5.4) at Reynolds
number of 333 are replotted. The present oval tube sample has
similar conguration (Wf/Pf 0.43 and Pw/Pf 6.0) at Pf 2.54 mm.
Original Rush et al. [33] data, which were presented as a function of
x* (x/PfRe), was converted to x (length from channel inlet) in Fig. 5.
In addition, the present wave geometry (Pw 15.2 mm and
Wf 1.1 mm) is also drawn at x axis. Fig. 5 shows that local Nusselt
number is approximately constant upto second wave, and increases
from there on. As illustrated in Fig. 1(a), one row sample encloses
two sine waves, and two row sample encloses four sine waves.
Thus, we may expect higher j factor for two row conguration as
compared with one row conguration.
The foregoing discussion suggests that heat transfer coefcient
of two row sample may be larger than that of one row sample due
to the characteristics of sine wave channel. This argument may be
strengthened by the usage of oval tube for the present oval tube
heat exchanger. For an oval tube, the inuential region downstream
of the tube will be smaller than that of the round tube. Then, the
heat transfer characteristics of n-and-tube heat exchanger will be

Fig. 4. Effect of tube row on j and f factor for oval tube and round tube samples.

present results. The reason was attributed to the suction effect of


condensing water vapor near valley of the wave n, which even-
tually alleviates the deterioration of j factor for deep row samples
[9]. Fig. 4 shows that, for 15.9 mm O.D. round tube sample, f factors
are not affected by number of tube row, except for one row
conguration. From two row conguration, tubes are arranged in Fig. 5. Local Nusselt numbers of sine wave channel for Wf/Pf 0.48 and Pw/Pf 5.4
staggered fashion. Then, the turbulence generated by the tubes and (Rush et al. [33]).
N.-H. Kim et al. / Applied Thermal Engineering 66 (2014) 580e589 587

Fig. 7. Ratios of ho ho Ao =V_ and DP/L between oval and round samples.
Fig. 6. ho ho Ao =V_ and DP/L for round and oval tube samples (dry data from Ref. [16]).

governed by the channel characteristics, which may be the case of 30.1e34.9%, 32.2e34.5% for oval, 15.9 mm and 12.7 mm round tube
present oval tube heat exchanger. Similar argument may apply to sample respectively) than those under dry condition. The DP/L
explain the higher j factor for four row sample compare with that of values under wet condition are, however, larger than those under
three row sample. For the present oval tube heat exchanger, dry condition. In addition, the increase is larger for round tube
transverse tube pitch (Pt) is 35.0 mm and tube minor diameter samples (37.4e45.5% for 15.9 mm and 34.9e49.7% for 12.7 mm
(Dmin) is 10.0 mm, resulting the distance between tubes to be sample) than for oval tube sample (17.1e22.4%). At present, the
25.0 mm. This distance is wider than those (22.2 mm and 19.1 mm) reason why oval tube sample yields smaller pressure drop increase
of 15.9 mm and 12.7 mm O.D. round tube samples. In addition, the compared with round tube samples is not clear. Size of condensate
inuential region downstream of the oval tube will be smaller hanging underneath the tube may be smaller for oval tube than for
compared with that of the round tube. In such case, third row will round tube. Sine wave pattern may be more benecial for
mostly be affected by rst row, and fourth row will be affected by condensate drainage than herringbone wave pattern. In Table 4,
second row. Then, four row sample may yield higher j factor than ho ho Ao =V_ and DP/L values of the samples under wet or dry condi-
three row sample. tion are listed at frontal velocity of 2.5 m s1.
In general, thermal performance comparison between n-and- In Fig. 7, ratios of ho ho Ao =V_ between oval and round tube for
tube heat exchangers is made by comparing j and f factor of each different tube row are plotted as a function of ratios of DP/L using
heat exchanger. However, when the denitions of Reynolds num- the data listed in Table 4. Fig. 7 shows that ratios of ho ho Ao =V_ are
ber of each heat exchanger are different as is the case of present larger than ratios of DP/L irrespective of tube row. This implies that
samples, direct comparison of j and f factor may lead to erroneous airside performance (considering both heat transfer and friction) of
conclusion. In such a case, comparison of heat transfer rate per unit oval tube heat exchanger is superior to that of round tube heat
volume ho ho Ao =V_ and consumed power per unit volume (or exchangers. The effect is more pronounced when oval tube samples
pressure drop per unit length, DP/L) will be more appropriate [2]. are compared with 15.9 mm O.D. samples than when compared
Fig. 6 shows ho ho Ao =V_ and DP/L for oval and round tube samples with 12.7 mm O.D. samples. Also shown in Fig. 7 are the ratios
having two row and 2.54 mm n pitch. This gure shows that heat under dry condition. The ratios of ho ho Ao =V_ under wet condition
transfer rates of oval tube sample are slightly (13.2e15.0% and 1.1e are larger than those under dry condition, which implies that oval
8.0%) larger than those of 15.9 mm or 12.7 mm O.D. round tube tube geometry is more benecial under wet condition.
sample respectively. On the other hand, pressure drops are signif- To further investigate the effect of individual parameters on
icantly (33.1e38.7% and 35.4e44.5%) smaller than those of 15.9 mm _ heat transfer coefcient (ho) and n efciency (ho) of two
ho ho Ao =V,
or 12.7 mm O.D. round tube sample. Also shown in Fig. 6 are row oval or round tube samples are shown in Figs. 8 and 9. Fig. 8
ho ho Ao =V_ and DP/L obtained under dry condition [16]. The shows that heat transfer coefcients of round tube samples are
ho ho Ao =V_ values under wet condition are smaller (23.3e30.1%, larger (9.8e14.1% for 15.9 mm and 17.6e29.7% for 12.7 mm) than

Table 4
Heat transfer and pressure drop data of round and oval tube samples at frontal velocity 2.5 m s1.
ho ho Ao
Row $ kW m3 K1 DP/L (Pa m1) ho (W m2 K1) ho
V
a
Oval 1row 21.30(22.14) 604.3(533.9) 38.26(35.9) 0.71(0.86)
2row 26.24(32.07) 644.2(440.6) 49.01(50.9) 0.69(0.84)
3row 24.81(28.59) 593.8(372.5) 45.29(45.7) 0.70(0.87)
4row 25.41(30.21) 515.1(358.3) 46.96(48.3) 0.69(0.85)
Round (15.9 mm O.D.) 1row 20.60(31.17) 1279(752.4) 63.11(62.6) 0.46(0.70)
2row 22.78(33.20) 1051(580.02) 56.55(60.9) 0.56(0.76)
3row 22.07(32.33) 964.7(565.2) 56.10(59.8) 0.56(0.77)
4row 21.92(33.35) 936.1(534.4) 55.38(62.3) 0.56(0.76)
Round (12.7 mm O.D.) 1row 23.41(34.31) 1238(793.2) 71.39(65.9) 0.44(0.71)
2row 25.17(37.69) 1099(592.3) 63.94(66.9) 0.53(0.76)
3row 24.66(37.96) 944.3(590.3) 63.32(67.7) 0.52(0.75)
4row 22.27(37.67) 832.3(550.6) 54.74(68.2) 0.56(0.75)
a
Values in parenthesis denote those obtained under dry condition [16].
588 N.-H. Kim et al. / Applied Thermal Engineering 66 (2014) 580e589

those of round tube samples (21.0e48.7% for 15.9 mm and 20.1e


54.7% for 12.7 mm). Fin efciencies are, however, larger (18.8e
35.2% for 15.9 mm and 18.8e38.0% for 12.7 mm). The pressure
drops of oval tube samples are smaller than those of round tube
samples (38.4e52.8% for 15.9 mm and 38.1e51.2% for 12.7 mm).
One thing to be noted is that heat transfer area per unit volume
_ for an oval tube sample is larger than that of round tube
Ao =V
sample (8.0% for 15.9 mm and 4.6% for 12.7 mm). As mentioned
previously, actual surface area considering the corrugation is used
as heat transfer area.

4. Conclusions

In this study, experiments were conducted on sine wave n-


and-tube heat exchangers having oval tubes of 0.6 aspect ratio
Fig. 8. ho and DP/L for round and oval tube samples (2row, Pf 2.54 mm) (dry data (minor diameter 10.0 mm, major diameter 16.5 mm) under wet
from Ref. [16]). condition. Twelve samples having different n pitches (2.12 mm,
2.54 mm, 3.18 mm) and tube rows (one to four) were tested. Eight
herringbone wave n-and-tube heat exchangers having round
those of oval tube sample. As noted earlier, pressure drops of round tubes of two different diameter (12.7 mm and 15.9 mm) were also
tube samples are also larger (31.9e38.7% for 15.9 mm and 35.4e tested. The round tube samples had different tube rows (one to
44.5% for 12.7 mm) than those of oval tube sample. Also shown in four) with xed n pitch (2.54 mm). Data are also compared with
Fig. 8 are ho and DP/L obtained under dry condition. The heat those obtained under dry condition [16]. Listed below are major
transfer coefcients under wet condition are smaller than those ndings.
under dry condition, with the difference larger for oval tube sample
than for round tube samples (15.6e21.3%, 7.8e11.9%, 2.4e6.7% for (1) For oval tube samples, the effect of n pitch on j and f factor is
oval, 15.9 mm and 12.7 mm sample respectively). The DP/L values not signicant.
under wet condition are larger than those under dry condition as (2) For round tube samples, the highest j factor is obtained for
noted previously. one row conguration. For oval tube samples, however, the
Fig. 9 shows that the n efciency of oval tube sample is lowest j factor is obtained for one row conguration. Possible
signicantly larger than those of round tube samples with the effect reasoning is provided considering the ow and heat transfer
more pronounced under wet condition (10.8e19.4% for 15.9 mm characteristics of sine wave channel combined with con-
and 13.4e27.1% for 12.7 mm). The n efciency of an oval tube is necting oval tubes. The friction factor decreases as number of
larger than that of round tube because heat ow path in the n is tube row increases.
more perpendicular to the ow direction. In addition, the n (3) Heat transfer coefcients of the oval tube samples are
thickness of oval tube sample (0.14 mm) is larger than that of round smaller than those of round tube samples. However, pressure
tube sample (0.12 mm). Also shown in Fig. 9 are n efciencies drops of oval tube samples are even smaller than those of
under dry condition. Wet n efciencies are in general smaller than round tube samples, which (along with higher n efciency
dry n efciencies. Under wet condition, steeper temperature of the oval tube sample) yield superior airside performance
gradient is manifested along the n due to moisture transfer in for oval tube samples.
addition to heat transfer, which resulted in lower n efciency than (4) Under wet condition compared with under dry condition,
under dry condition. heat transfer coefcient and n efciency decrease, and
In Table 4, heat transfer coefcients and n efciencies of oval pressure drop increases. The differences in n efciency and
and round tube samples are listed at frontal velocity of 2.5 m s1. pressure drop are larger for round tube samples than for oval
Heat transfer coefcients of oval tube samples are smaller than tube samples. Resultantly, it is shown that oval geometry is
more benecial under wet condition than under dry
condition

Acknowledgements

This research was supported by Basic Science Research Program


through the National Research Foundation of Korea by the Ministry
of Education, Science and Technology in 2011.

References

[1] D.R. Mirth, S. Ramadhyani, Prediction of cooling coil performance under


condensing conditions, Int. J. Heat Fluid Flow 14 (1993) 391e400.
[2] R.L. Webb, N.H. Kim, Principles of Enhaced Heat Transfer, second ed., Taylor
and Francis Pub., 2005.
[3] C.C. Wang, Recent advances in n-and-tube heat exchangers, Int. J. Air-Cond.
Refrig. 19 (2011) 291e301.
[4] W. Pirompugd, C.C. Wang, S. Wongwises, A review on reduction method for
heat and mass transfer characteristics of n-and-tube heat exchangers under
Fig. 9. Fin efciency of round and oval tube samples (2row, Pf 2.54 mm) (dry data dehumidifying conditions, Int. J. Heat Mass Trans. 52 (9e10) (2009) 2370e
from Ref. [16]). 2378.
N.-H. Kim et al. / Applied Thermal Engineering 66 (2014) 580e589 589

[5] C.C. Wang, C.T. Chang, Heat and mass transfer for plate n-and-tube heat [18] ASHRAE Standard 41.1, Standard Method for Temperature Measurement,
exchangers, with and without hydrophilic coating, Int. J. Heat Mass Trans. 41 ASHRAE, 1986.
(1998) 3109e3120. [19] ASHRAE Standard 41.2, Standard Method for Laboratory Air-ow Measure-
[6] C.C. Wang, W.S. Lee, W.T. Sheu, Y.T. Chang, A comparison of the airside per- ment, ASHRAE, 1987.
formance of n-and-tube heat exchangers in wet conditions; with and [20] ASHRAE Standard 41.5, Standard Measurement Guide, Engineering Analysis
without hydrophilic coating, Appl. Therm. Eng. 22 (2002) 267e278. and Experimental Data, 1975.
[7] J. Min, X. Wu, L. Shen, F. Gao, Hydrophilic treatment and performance eval- [21] N.H. Kim, W.K. Oh, J.P. Cho, W.Y. Park, Y. Baek, Data reduction on the air-side
uation of copper nned tube evaporator, Appl. Therm. Eng. 31 (2011) 2936e heat transfer coefcients of heat exchangers under dehumidifying conditions,
2942. Korean J. Air-Cond. Refrig. 15 (2003) 73e85.
[8] C.C. Wang, J.S. Liaw, B.C. Yang, Airside performance of herringbone wavy n- [22] ESDU 98005, Design and Performance Evaluation of Heat Exchangers: the
and-tube heat exchangers e data with larger diameter tube, Int. J. Heat Mass Effectiveness and NTU Method, Engineering and Sciences Data Unit 98005
Trans. 54 (2011) 1024e1029. with Amendment A, London ESDU International plc., 1998, pp. 122e129.
[9] Y.C. Liu, S. Wongwises, W.J. Chang, C.C. Wang, Air-side performance of n- [23] J.P. Holman, Heat Transfer, tenth ed., McGraw-Hill Pub, 2010.
and-tube heat exchangers in dehumidifying conditions e data with larger [24] V. Gnielinski, New equations for heat and mass transfer in turbulent pipe
diameter, Int. J. Heat Mass Trans. 53 (2010) 1603e1608. ows, Int. Chem. Eng. 16 (1976) 359e368.
[10] S.M. Saboya, F.E.M. Saboya, Experiments on elliptic sections in one- and two- [25] J.C. Min, T. Tao, X.F. Peng, Efciency of ns used in a nned oval tube heat
row arrangements of plate n and tube heat exchangers, Exp. Therm. Fluid exchanger, J. Enhanc. Heat Trans. 10 (3) (2003) 323e334.
Sci. 24 (2001) 65e75. [26] T.E. Schmidt, Heat transfer calculations for extended surfaces, J. ASRE Refrig.
[11] J.C. Min, R.L. Webb, Numerical analyses of effects of tube shape on perfor- Eng. 4 (1949) 351e357.
mance of a nned tube heat exchanger, J. Enhanc. Heat Trans. 11 (1) (2004) [27] R.L. Webb, A. Iyengar, Oval nned tube condenser and design pressure limits,
61e73. J. Enhanc. Heat Trans. 8 (2001) 147e158.
[12] J.S. Leu, M.S. Liu, J.S. Liaw, C.C. Wang, A numerical investigation of louvered [28] C.C. Wang, R.L. Webb, K.Y. Chi, Data reduction for airside performance of n-
n-and-tube heat exchangers having circular and oval tube congurations, and-tube heat exchangers, Exp. Therm. Fluid Sci. 21 (2000) 218e226.
Int. J. Heat Mass Trans. 44 (2001) 4235e4243. [29] K. Torikoshi, G.N. Xi, Y. Nakazawa, H. Asano, Flow and heat transfer perfor-
[13] H. Han, Y.L. He, Y.S. Li, Y. Wang, M. Wu, A numerical study on compact mance of a plate n and tube heat exchanger (rst report: effect of n pitch),
enhanced n-and-tube heat exchangers with oval and circular tube congu- in: Proceedings of the 10th Int. Heat Transfer Conf., vol. 4, 1994, pp. 411e416.
rations, Int. J. Heat Mass Trans. 65 (2013) 686e695. [30] D.G. Rich, The effect of n spacing on the heat transfer and friction perfor-
[14] Y.B. Tao, Y.L. He, Z.G. Wu, W.Q. Tao, Three-dimensional numerical study and mance of multi-row plate n-and-tube heat exchangers, ASHRAE Trans. 79 (2)
eld synergy principle analysis of wavy n heat exchangers with elliptic (1973) 137e145.
tubes, Int. J. Heat Fluid Flow 28 (2007) 1531e1544. [31] C.C. Wang, W.L. Fu, C.T. Chang, Heat transfer and friction characteristics of
[15] R.S. Matos, T.A. Laursen, J.V.C. Vargas, A. Bejan, Three-dimensional optimiza- typical wavy n-and-tube heat exchangers, Exp. Therm. Fluid Sci. 14 (1997)
tion of staggered nned circular and elliptic tubes in forced convection, Int. J. 174e186.
Therm. Sci. 43 (2004) 477e487. [32] D.G. Rich, The effect of the number of tube rows on heat transfer performance
[16] N.H. Kim, K.J. Lee, J.C. Han, B.N. Choi, Thermal performance of sine wave n- of smooth plate n-and-tube heat exchangers, ASHRAE Trans. 81 (1) (1975)
and-oval tube heat exchangers, Int. J. Air-Cond. Refrig. 21 (1) (2013) 135008. 307e317.
[17] B.N. Choi, F. Yi, H.M. Sim, N.H. Kim, Air-side performance of n-and-tube heat [33] T.A. Rush, T.A. Newell, A.M. Jacobi, An experimental study of ow and heat
exchangers having sine wave ns and oval tubes, Korean J. Air-Cond. Refrig. transfer in sinusoidal wavy passages, Int. J. Heat Mass Trans. 42 (1999) 1541e
Eng. 25 (5) (2013) 279e288. 1553.

You might also like