Anda di halaman 1dari 5

Experimental Thermal and Fluid Science 33 (2009) 203207

Contents lists available at ScienceDirect

Experimental Thermal and Fluid Science


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

Heat transfer coefcients of shell and coiled tube heat exchangers


M.R. Salimpour *
Department of Mechanical Engineering, Isfahan University of Technology, Isfahan 84156-83111, Iran

a r t i c l e

i n f o

Article history:
Received 6 February 2007
Received in revised form 28 July 2008
Accepted 31 July 2008

Keywords:
Empirical
Shell and coiled tube
Heat exchanger
Experimental
Correlation

a b s t r a c t
In the present study, the heat transfer coefcients of shell and helically coiled tube heat exchangers were
investigated experimentally. Three heat exchangers with different coil pitches were selected as test section for both parallel-ow and counter-ow congurations. All the required parameters like inlet and
outlet temperatures of tube-side and shell-side uids, ow rate of uids, etc. were measured using appropriate instruments. Totally, 75 test runs were performed from which the tube-side and shell-side heat
transfer coefcients were calculated. Empirical correlations were proposed for shell-side and tube-side.
The calculated heat transfer coefcients of tube-side were also compared to the existing correlations
for other boundary conditions and a reasonable agreement was observed.
2008 Elsevier Inc. All rights reserved.

1. Introduction
Heat exchangers are used in a wide variety of applications
including power plants, nuclear reactors, refrigeration and airconditioning systems, automotive industries, heat recovery systems, chemical processing, and food industries [13]. Besides
the performance of the heat exchanger being improved, the heat
transfer enhancement enables the size of the heat exchanger to
be considerably decreased. In general, the enhancement techniques can be divided into two groups: active and passive techniques. The active techniques require external forces like uid
vibration, electric eld, and surface vibration. The passive techniques require special surface geometries or uid additives like
various tube inserts. Both techniques have been widely used to
improve heat transfer performance of heat exchangers. Due to
their compact structure and high heat transfer coefcient, helically coiled tubes have been introduced as one of the passive heat
transfer enhancement techniques and are widely used in various
industrial applications.
Several studies have indicated that helically coiled tubes are
superior to straight tubes when employed in heat transfer applications [4,5]. The centrifugal force due to the curvature of the tube
results in the secondary ow development which enhances the
heat transfer rate. This phenomenon can be benecial especially
in laminar ow regime. Thermal performance and pressure drop
of a shell and helically coiled tube heat exchanger with and without helical crimped ns have been investigated by Naphon [6].
* Tel.: +98 311 391 5210; fax: +98 311 391 2628.
E-mail addresses: salimpour@cc.iut.ac.ir, m_salimpour@yahoo.com.
0894-1777/$ - see front matter 2008 Elsevier Inc. All rights reserved.
doi:10.1016/j.expthermusci.2008.07.015

Naphon and Wongwises [7] provided a literature review on heat


transfer and ow characteristics of single-phase and two-phase
ow in curved tubes including helically coiled tubes and spirally
coiled tubes.
The majority of the studies related to helically coiled tubes
and heat exchangers have dealt with two major boundary conditions, i.e. constant heat ux and constant wall temperature
[8,9]. However, these boundary conditions are not encountered
in most single-phase heat exchangers. Rennie [10] studied the
double-pipe helical heat exchangers numerically and experimentally neglecting the effect of coiled tube pitch. Though the
boundary condition of his work was different from the conventional boundary conditions of constant wall temperature and
constant heat ux, however, it is obvious that the geometry of
the double-pipe coiled tube heat exchanger is completely different from that of shell and coiled tube heat exchanger of the
present work.
Kumar et al. [11] studied a tube-in-tube helically coiled heat exchanger for turbulent ow regime numerically.
Numerical investigations were done to understand forced laminar uid ow in rectangular coiled pipes with circular cross-section by Conte and Peng [12]. Their focus was addressed on
exploring the ow pattern and temperature distribution through
the pipe.
One of the most frequent uses of helically coiled tubes is in shell
and coiled tube heat exchangers. Going through the existing literature, it was revealed that there are a few investigations on the
heat transfer coefcients of this kind of heat exchangers considering the geometrical effects like coil pitch. Also, this scarcity is more
prominent for shell-side heat transfer coefcients.

204

M.R. Salimpour / Experimental Thermal and Fluid Science 33 (2009) 203207

Nomenclature
surface of coiled tube, m2
coil pitch, m
constant of Eq. (2)
Prandtl number, = lCP/k
curvature radius, m
unknown constants
coiled tube diameter, m
shell diameter, m
shell-side hydraulic diameter, m
Dean number, = Re(d/2Rc)0.5
averaged convective heat transfer coefcient, W/m2 K
Helical number, = De/(1 + c2)0.5
thermal conductivity, W/m2 K
heat exchanger length, m
exponent in Eq. (2)

N
Nu
Re
U
v

Greek letters
d
curvature ratio, = d/2Rc
c
dimensionless pitch, = b/2pRc
l
viscosity, kg/m s
q
density, kg/m3
Subscripts
i
inside condition
o
outside condition

exchanging system in which a hot water stream owing inside


the tube-side is cooled by a cold water stream owing in the
shell-side. The main parts of the cycle are coiled tube heat exchanger (1), centrifugal pump (2), storage tank (3), and
heater (4).
The heat exchangers include a copper coiled tube and an insulated shell. The dimensions of the heat exchangers are depicted
in Table 1. The water of storage tank is heated using an electric
heater. Reaching to a prescribed temperature, pump is started to
circulate the hot water in the cycle. A ball valve and a globe valve
are used to control the ow rate of coolant water and hot water,
respectively. To measure the ow rate of the cold stream a rotameter with the accuracy of 2.78  104 kg/s is installed upstream of
the heat exchanger while for the hot stream a measuring pot with
the accuracy of 3.3  103 kg/s is used. The inlet and outlet temperatures of hot and cold water were recorded manually using 4
RTD thermocouples inserted in the small holes made in the inlet
and outlet tubes of each heat exchanger and sealed to prevent
any leakage. Also, all the pipes and connections between the temperature measuring stations and heat exchanger were duly insulated. All the temperatures were measured three times with
accuracy of 0.1 C in the time steps of 10 min, and the average values were used for further analysis. Appropriate arrangements were
provided to measure the pressure loss of both tube-side and shellside streams. All the tube- and shell-side uids properties were assessed at the mean temperature of the uids (average of inlet and
outlet temperatures).

2. Geometry of shell and coiled tube heat exchanger


A typical shell and coiled tube heat exchanger is shown in Fig. 1.
In this gure, d is the diameter of the coiled tube, Rc is the curvature
radius of the coil, D is the inner diameter of shell, and b is the coil
pitch. The curvature ratio, d, is dened as the coil-to-tube diameter
ratio, d/2Rc, and the non-dimensional pitch, c, is dened as b/2pRc.
The other four important dimensionless parameters of coiled tube
namely, Reynolds number (Rei), Nusselt number (Nui), Dean number (De), and Helical number (He) are dened as follow:

Rei qvi di =l;

Nui hi di =k;
0:5

He De=1 c2 0:5

De Rei di =2Rc ;

where vi and hi are average velocity and convective heat transfer


coefcient of coiled tube, respectively.
Shell-side Reynolds number, Reo, and Nusselt number, Nuo, are
dened as

Reo qvo Dh =l;


where vo, ho and Dh

Nuo ho Dh =k
D2 2pRc d2o c1
D2pRc do c1

number of data points


Nusselt number
Reynolds number
overall heat transfer coefcient, W/m2 K
uid average velocity, m/s

are average velocity, convective

heat transfer coefcient, and the hydraulic diameter of shell-side,


respectively.
3. Experimental set-up
The schematic diagram of experimental set-up is shown in
Fig. 2. The set-up is a well instrumented single-phase heat

2Rc

Fig. 1. Schematic view of a typical shell and coiled tube heat exchanger.

A
b
C
Pr
Rc
C1,. . .,C4
d
D
Dh
De
h
He
k
L
n

M.R. Salimpour / Experimental Thermal and Fluid Science 33 (2009) 203207

205

Fig. 2. Flow diagram of the experimental set-up.

related to the inner and outer heat transfer coefcients by the following equation [14]:

Table 1
Characteristic dimensions of shell and coiled tube heat exchangers
Heat exchanger

di, mm

do, mm

D, mm

b, mm

L, mm

1
2
3

9
12
12

12
16
16

120
120
120

17.0
21.4
26.7

0.113
0.157
0.157

230
225
280

4. Data collection and analysis


The range of operating parameters is given in Table 2. As is seen
from this table, a wide range of ow rates both in the tube-side and
shell-side is covered for counter-ow congurations. The tests
were performed for all three coiled tube heat exchangers which resulted in a total 75 test runs.
Heat transfer coefcients for the shell-side, ho, and for the coiled
tube-side, hi, were calculated using Wilson plots as described by
Rose [13]. Using Wilson plots, the heat transfer coefcients can be
calculated based on the overall temperature difference and the rate
of heat transfer. As there is no need for measuring the tube wall
temperature in this method, it was chosen to avoid the disturbance
of ow patterns and heat transfer while attempting to measure
wall temperatures. In this work, the ow in the coiled tube was
kept constant and the ow in the shell-side was varied for the ve
different ow rates. The overall heat transfer coefcient can be

1
Ao
Ao lndo =di 1

U o Ai hi
2pkL
ho

where di and do are inner and outer diameters of the tube, respectively; k is the thermal conductivity of the wall; and L is the length
of the heat exchanger. After calculating the overall heat transfer
coefcients, the only unknown variables in Eq. (1) are the heat
transfer coefcients. By keeping the mass ow rate in the inner tube
constant, it is then assumed that the inner heat transfer coefcient
is constant. The outer heat transfer coefcient is assumed to behave
in the following manner with the uid velocity in the shell, vo:

ho Cvno

Substituting Eq. (2) into Eq. (1), the values for the constant, C, and
the exponent, n, were determined through curve tting. The inner
and outer heat transfer coefcients could then be calculated. This
procedure was repeated for each inner ow rate, coil size, and conguration. This resulted in 15 Wilson plots, and 15 inner heat transfer coefcients. For each Wilson plot, ve outer heat transfer
coefcients were calculated, i.e. totally 75 outer heat transfer coefcients were calculated.
The uncertainty analysis was performed by the method proposed by Schultz and Cole [15] for all experiments, and it was
found that the expected experimental error was within 8% for
all the runs.

5. Results and discussion

Table 2
Range of operating parameters
Parameters

Range

Tube-side water ow rate


Shell-side water ow rate
Tube inlet temperature
Tube outlet temperature
Shell inlet temperature
Shell outlet temperature

0.0160.113 kg/s
0.0190.136 kg/s
33.453.2 C
23.544.9 C
10.919.2 C
14.637.3 C

Fig. 3 represents the tube-side Nusselt number versus Dean


number for shell and coiled tube heat exchangers with different
coil pitches. This gure also illustrates a comparison between the
results of this study and the empirical correlations proposed in
[8] for boundary conditions of constant heat ux and constant wall
temperature. In this gure, correlations for constant wall temperature and constant heat ux boundary conditions proposed by

206

M.R. Salimpour / Experimental Thermal and Fluid Science 33 (2009) 203207

the shell-side Nusselt numbers increase with Reynolds number.


Also, it appears that the increase of coiled tube pitch leads to higher values of shell-side Nusselt number. This may be explained as in
smaller coil pitches, the coolant water is conned in the space between the successive coil rounds and a semi-dead zone is formed.
As in this region, the ow of shell-side uid is decelerated; heat
transfer coefcients will be descended. Also, it can be easily seen
that the difference among the coils with different geometries are
more severe in high Reynolds number region; i.e. when the shellside Reynolds number is increased choosing the appropriate coiled
tube geometry becomes more critical.

80

70

60

Nui

50

40

30

20

10

6. Proposing correlations to predict tube- and shell-sides


Nusselt numbers

coil 1
coil 2
coil 3
ref. 8, case I
ref. 8, case II

1000

2000

3000

4000

5000

De
Fig. 3. Variation of coiled tube Nusselt number with Dean number.

Manlapaz and Churchill [8] are indicated by case I, and case II and
illustrated by solid lines and dash lines, respectively.
From this gure, it is seen that increasing the coiled tube pitch
will decrease the inner Nusselt number. This matter can be justied as the higher values of coil pitches correspond to the lower
values of Helical number or loose-coiling conditions. Therefore,
the heat transfer coefcients of these coiled tubes are closer to
those of the straight pipes expectedly. A comparison of the present
data with the previous correlations revealed that when coil pitches
are negligible, in low Dean numbers (De < 3000), the correlation of
constant temperature boundary condition predicts the present
data quite well; however, in high Dean numbers (De > 3000), this
correlation overestimates heat transfer coefcients. Also, the correlation pertained to constant heat ux boundary condition overestimates all the data of the present study.
Fig. 4 represents the variations of shell-side Nusselt number
with shell-side Reynolds number which was calculated based on
hydraulic diameter of the shell. From this gure, it is observed that

As discussed in the previous section, the existing correlations


for estimation of heat transfer coefcients of helical coils with constant temperature boundary condition do not conform to the results of the present study, thus a new correlation was developed
to predict the inner Nusselt number. Based on the experimental results, a correlation between the tube-side Nusselt number and
Dean number, Prandtl number, and dimensionless coil pitch is obtained using least square analysis. For this purpose, the following
functional relationship is assumed:

Nui C 1 DeC 2 Pr C 3 cC 4

Taking logarithm of Eq. (3), and introducing the data of Nui and the
corresponding De, Pr and c into it, an error function can be obtained
as

EC 1 ; C 2 ; C 3 ; C 4
N n
h
io 2
X
lnNui j  ln C 1 C 2 ln Dej C 3 ln Pr j C 4 ln cj

j1

Minimizing the above error function using least square regression,


the following correlation was found with average error of 0.91%:

Nui 0:152De0:431 Pr1:06 c0:277

Fig. 5 shows the comparison between the predicted Nusselt numbers by the proposed correlation, Eq. (5), and the experimental Nusselt numbers.

70
70

60

predicted Nusselt number

60

Nuo

50

40

30

shell 1
shell 2
shell 3

20

10

50

40

30

20

10

0
100

200

300

400

500

Reo
Fig. 4. Variation of shell-side Nusselt number with Reynolds number.

10

20

30

40

50

60

70

experimental Nusselt number


Fig. 5. Comparison of tube-side experimental Nusselt number with predicted
Nusselt number.

M.R. Salimpour / Experimental Thermal and Fluid Science 33 (2009) 203207

exchangers. Heat exchangers with three different coil pitches were


tested for counter-ow conguration. It was revealed that the
empirical correlation for constant temperature boundary condition
is quite in agreement with the present data in low Dean number
region. From the results of the present study, it was found out that
the shell-side heat transfer coefcients of the coils with larger
pitches are higher than those for smaller pitches. Finally, based
on the results of this study, two correlations were developed to
predict the inner and outer heat transfer coefcients of the coiled
tube heat exchangers.

70

60

predicted Nusselt number

207

50

40

30

References

20

10

10

20

30

40

50

60

70

experimental Nusselt number


Fig. 6. Comparison of shell-side experimental Nusselt number with predicted
Nusselt number.

Investigating the literature, no correlation was found to evaluate the shell-side heat transfer coefcients of helically coiled tube
heat exchangers. Therefore, in the present study, one correlation
was developed to predict the shell-side Nusselt number as well.
To predict the shell-side Nusselt numbers of coiled tube heat
exchangers, a similar treatment to the inner Nusselt numbers
was done for the outer Nusselt numbers which led to the following
correlation with average error of 1.19%:

Nuo 19:64Re0:513
Pr0:129 c0:938
o

The comparison between the predicted Nusselt numbers by the


proposed correlation, Eq. (6), and the experimental Nusselt numbers, is presented in Fig. 6.
From Figs. 5 and 6, it is evident that the proposed correlations
are in good agreement with the present data.
7. Conclusions
An experimental investigation was carried out to study heat
transfer coefcients of the shell and helically coiled tube heat

[1] W.H. Fleming, J.A. Khan, C.A. Rhodes, Effective heat transfer in a metalhydridebased hydrogen separation process, Int. J. Hydrogen Energ. 26 (2001)
711724.
[2] P.K. Sahoo, M.I.A. Ansari, A.K. Datta, A computer based iterative solution for
accurate estimation of heat transfer coefcients in a helical tube heat
exchanger, J. Food Eng. 58 (2003) 211214.
[3] J. Yi, Z.H. Liu, J. Wang, Heat transfer characteristics of the evaporator section
using small helical coiled pipe in a looped heat pipe, Appl. Therm. Eng. 23
(2003) 8999.
[4] L.A.M. Janssen, C.J. Hoogendoorn, Laminar convective heat transfer in helical
coiled tubes, Int. J. Heat Mass Tran. 21 (1978) 11971206.
[5] D.G. Prabhanjan, G.S.V. Raghavan, T.J. Rennie, Comparison of heat transfer rates
between a straight tube heat exchanger and a helically coiled heat exchanger,
Int. Commun. Heat Mass Tran. 29 (2) (2002) 185191.
[6] P. Naphon, Thermal performance and pressure drop of the helical-coil heat
exchangers with and without helically crimped ns, Int. Commun. Heat Mass
Tran. 34 (3) (2007) 321330.
[7] P. Naphon, S. Wongwises, A review of ow and heat transfer characteristics in
curved tubes, Renew. Sust. Energ. Rev. 10 (2006) 463490.
[8] R. Manlapaz, S.W. Churchill, Fully developed laminar convection from a helical
coil, Chem. Eng. Commun. 9 (1981) 185200.
[9] C.E. Kalb, J.D. Seader, Fully developed viscous-ow heat transfer in curved
circular tubes with uniform wall temperature, AICHE J. 20 (1974)
340346.
[10] T.J. Rennie, Numerical and experimental studies of a double-pipe helical
heat exchanger, Ph.D. Thesis, McGill University, Montreal, Canada,
2004.
[11] V. Kumar, B. Faizee, M. Mridha, K.D.P. Nigam, Numerical studies of a tube-intube helically coiled heat exchanger, Chem. Eng. Process. (2008), doi:10.1016/
j.cep.2008.01.001.
[12] I. Conte, X.F. Peng, Numerical investigations of laminar ow in coiled pipes,
Appl. Therm. Eng. 28 (2008) 423432.
[13] J.W. Rose, Heat-transfer coefcients, Wilson plot and accuracy of thermal
measurements, Exp. Therm. Fluid Sci. 28 (2004) 7786.
[14] F.M. White, Heat Transfer, Addison-Wesley Publishing Company Inc., New
York, 1984.
[15] R.R. Schultz, R. Cole, Uncertainty analysis in boiling nucleation, AICHE Symp.
Ser. 189 (75) (1979) 3238.