Applied Thermal Engineering 35 (2012) 196e201

Contents lists available at SciVerse ScienceDirect

Applied Thermal Engineering

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

The influence of artificial roughness shape on heat transfer enhancement:

Corrugated tubes, dimpled tubes and wire coils
A. García a, J.P. Solano a, *, P.G. Vicente b, A. Viedma a
a
Universidad Politécnica de Cartagena, Departamento de Ingeniería Térmica y de Fluidos, Campus Muralla del Mar, 30202 Cartagena, Spain
b
Universidad Miguel Hernández, Departamento de Ingeniería de Sistemas Industriales Avenida de la Universidad, 03202 Elche, Spain

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

Article history: This work analyzes the thermalehydraulic behaviour of three types of enhancement technique based on
Received 7 June 2011 artificial roughness: corrugated tubes, dimpled tubes and wire coils. The comparison has been performed
Accepted 18 October 2011 from the three best specimens selected among the wide range of geometries investigated by the authors
Available online 25 October 2011
in previous works. Heat transfer and pressure drop experimental data in laminar, transition and
turbulent regimes are used in this investigation.
Keywords:
Results show that the shape of the artificial roughness exerts a greater influence on the pressure drop
Heat transfer enhancement
characteristics than on the heat transfer augmentation. Likewise, this shape strongly affects the advance
Wire coil inserts
Corrugated tubes
of the transition to turbulence and its characteristics: smooth or sudden. The study concludes that for
Dimpled tubes Reynolds numbers lower than 200, the use of smooth tubes is recommended. For Reynolds numbers
Turbulence promotion between 200 and 2000, the employment of wire coils is more advantageous, while for Reynolds numbers
higher than 2000, the use of corrugated and dimpled tubes is favoured over the wire coils because of the
lower pressure drop encountered for similar heat transfer coefficient levels.
Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction In fully developed internal turbulent flow, the velocity and

temperature profiles across the tube are similar in shape and
Enhancement techniques based on artificial roughness are used relatively flat until very close to the wall. Artificial roughness
in numerous applications of heat exchangers. The choice of an techniques are particularly appropriate for heat transfer augmen-
enhancement technique depends on variables such as: the flow tation in this flow regime, as they contribute to disturbing the
regime (Reynolds number), the fluid properties (Prandtl number), thermal boundary layer.
the existence or not of fouling, the allowable pressure drop and the With regard to the transition from laminar to turbulent flow,
existence or absence of natural convection. experimental evidence proves that these techniques promote the
The use of an enhancement technique may be conditioned by advance of transition [3]. As a result of the flow perturbation in the
the specific application: for example, wire coils are not applicable in viscous sub-layer, turbulence spots at Reynolds numbers below
the food industry due to hygiene problems but corrugated and 2300 lead to early turbulence phenomena. When the transition
dimpled tubes are. In the petrochemical industry, the use of takes place, the heat transfer rate can be five times higher than the
mechanically deformed tubes is not allowed for safety reasons. one for the laminar flow in a smooth tube [4]. The use of the best
However, the use of wire coils does not present any problem. In enhancement technique will bring about an important increase of
boilers and heat recovery systems, wire coils are frequently used the heat transfer rate in the transition region, this presenting a high
because of their easy removal for cleaning operations. potential in applications with highly viscous fluids, e.g. in the
In the fully laminar regime, the use of artificial roughness petrochemical and food industries.
techniques does not significantly improve the heat transfer coeffi-
cients as they only promote mixing in the boundary layer near the
2. Background
wall. Instead, devices that mix the gross flow are suitably employed
for heat transfer enhancement in this flow regime [1,2].
The use of wire coils to enhance heat transfer in tubular heat
exchangers goes back to the works by Joule in the middle of the XIX
century. In fact, those works are considered the pioneers in the field
* Corresponding author. Tel.: þ34 968325938; fax: þ34 968325999. of heat transfer enhancement. The technology which allows the
E-mail address: [email protected] (J.P. Solano). manufacture of low-cost deformed tubes was developed in the last

1359-4311/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.applthermaleng.2011.10.030
 A. García et al. / Applied Thermal Engineering 35 (2012) 196e201 197

Nomenclature Gr Grashof number, gbd4q00 /y2l

Nu Nusselt number, ad/l
cp fluid specific heat (J kg1 K1) Pr Prandtl number, mcp/l
d tube inner diameter (m) Ra Rayleigh number, GrPr
e wire coil diameter (m) Re Reynolds number, rvd/m
h corrugation/dimple height (m)
l length between dimples (m) Greek symbols
lh length of the heat transfer test section (m) a heat transfer coefficient (W m2 K1)
lp length of the pressure drop test section (m) b thermal expansion coefficient (K1)
p helical/corrugation/dimple pitch (m) l thermal conductivity (W m1 K1)
DP pressure drop across the test section (Pa) m dynamic viscosity (kg m1 s1)
Q overall electrical power added (W) y kinematic viscosity (m2 s1)
Q[ heat losses in the test section (W) r fluid density (kg m3)
q00 heat flux (Q  Q[)/(pdlh) (W m2)
v mean fluid velocity (m s1) Subscripts
a augmented tube
Dimensionless groups s smooth tube
f Fanning friction factor, DPd/(2rv2lp) N asymptotic

third of the XX century. The patent of the corrugated tubes dates the laminar regime: in viscous fluids, the flow can be laminar in the
back to 1977 [5]. The patent of a method for manufacturing dimpled entrance, where the fluid is cold and its viscosity is high. Transition
tubes by cold external deformation dates back to 1989 [6]. takes place at an undefined point of the heat exchanger. Because of
In the field of enhanced heat transfer, there are very few this, the transition point (critical Reynolds number) is an important
experimental studies on laminar flow. Here, the entrance effects parameter to bear in mind in all enhancement techniques.
and the secondary flow induced by buoyancy forces greatly The use of a surface roughness in the transition region
complicate the analysis. On wire coils, the work of Uttarwar and (Re ¼ 250e3000) can be very effective to increase heat transfer
Raja Rao [7] has been widely mentioned in the open literature [3,8]. [4,18]. Oliver and Shoji [19] studied different insert devices in the
However, their heat transfer results were strongly influenced by the laminar and transition region. Although their work only covered
entry region. Recently, Akhavan-Behabadi et al. [9] studied the heat Reynolds numbers below 700, it was proven that in this region,
transfer augmentation in laminar flow in tubes with different wire wire coils increase heat transfer in a much more efficient way than
coil inserts. The experiments were carried out in a double-pipe other insert devices such as mesh inserts and twisted tapes. Rav-
configuration with constant wall temperature, and did not igururajan and Bergles [20] and Li et al. [21] demonstrated through
account for entry region effects nor mixed convection. visualization experiments that the presence of artificial roughness
Barba et al. [10] published an experimental paper on heat promotes turbulent flow at Reynolds numbers below 2000. The
transfer enhancement in a corrugated tube for laminar and tran- paper of Olsson and Sundén [22] was focused on the laminar-
sitional flow. They reported pressure drop increases of 2.5 times for transitional region (Re ¼ 500e6000). They studied rib-
Re ¼ 800 and Nusselt number augmentations of about 17 times at roughened, dimpled and offset strip fin small tubes for radiators,
Pr ¼ 200, compared to the smooth tube. employing air as test fluid (Prz0.7). Their measurements were
Experimental studies on surface roughness enhancement highly influenced by the entry region. Since the flow behaviour
techniques have been usually carried out for turbulent flow at low presented a smooth transition, both friction factor and Nusselt
Prandtl numbers (water and air). Ravigururajan and Bergles [11] number results were fitted to Reynolds number by a simple power
compiled a great amount of experimental data from seventeen series correlation. To extend the validity of heat transfer results, it
sources and seven different enhanced tubes. They developed was assumed that Nusselt number was proportional to Pr1/3. Meyer
general correlations for the friction factor and for the Nusselt and Olivier [23,24] obtained heat transfer coefficients, and diabatic
number for turbulent flow. They concluded that roughness shape and adiabatic friction factor data for four helical finned tubes for
exerts a higher influence on pressure drop than on heat transfer. fully developed and developing flow, covering the laminar, transi-
Zhang et al. [12] obtained very similar results in wire coils of tional and fully turbulent flow. They analyzed the influence of
circular and square section. On the other hand, Zimparov et al. [13] secondary flows on the advance of transition, and the impact of
studied corrugated tubes with the same dimensionless geometrical different inlet geometries.
parameters but different shape and observed differences up to 25% Compound enhancement techniques have been recently studied
in the friction factor. Chen et al. [14] performed an experimental by several authors. Thianpong et al. [25] investigated the ther-
investigation of different dimpled tube geometries in turbulent malehydraulic performance of combinations of three twisted tapes
flow, providing accurate correlations for heat transfer and pressure and two dimpled tubes. They proposed experimental correlations
drop analysis. They concluded that the size and weight of the heat of heat transfer and pressure drop as a function of the pitch ratio
exchanger could be reduced by a factor of almost 2 without and twist ratio. The effects of the entrance length and mixed
affecting any other system conditions. Wang et al. [15] have studied convection were not considered. Saha adopted new configurations
the thermalehydraulic characteristics in tubes with outward- in square ribbed channels with wire coil inserts, and studied
facing and raised dimples in staggered and aligned configura- separately the laminar [26] and turbulent [27] flow regimes.
tions, for Reynolds number in the range 15000e60000 and using The aim of the present work is to perform a well-reasoned
air as working fluid. comparison of the thermalehydraulic behaviour of three types of
Further to the well known interest in using corrugated tubes in enhancement technique based on artificial roughness: corrugated
turbulent flow, the most interesting region is undoubtedly the tubes, dimpled tubes and wire coils. The authors have revisited
transition region [4,16,17]. A heat exchanger can partially work in their own results on dimpled tubes [28,29], corrugated tubes
198 A. García et al. / Applied Thermal Engineering 35 (2012) 196e201

[30,31] and wire coils [32], where a wide range of geometrical Table 1
parameters of each artificial roughness technique was investigated. Characteristic dimensions of the roughened tubes.

The specimens which yielded the best thermalehydraulic perfor- Enhancement d h (e) p l h/d p/d d2p/l
mance for each shape have been chosen in the present investiga- technique [mm] [mm] [mm] [mm] (e/d) [e] [e]
tion. The heat transfer and pressure drop experimental data [e]

obtained by the authors in laminar, transition and turbulent Dimpled, D05 16.0 1.83 14.50 9.02 0.114 0.906 1.957
Corrugated, C01 18.0 1.03 15.95 e 0.057 0.886 e
regimes have been analyzed on a comparative basis. The main
Wire coil, W01 18.0 1.34 21.12 e 0.074 1.173 e
advantage of this approach is that the range of Reynolds and
Prandtl numbers investigated is similar for the three techniques, as
well as the thermal boundary condition and development lengths.
pitch (p/d). Table 1 shows the geometrical and the dimensionless
This prevents an ambiguous interpretation when data from
parameters of the mechanically deformed tubes and the wire coil
different sources are analyzed. The criterion for the choice of the
analyzed in this work. The geometrical parameters analyzed in the
three specimens allows us to establish general conclusions on the
works from which the test specimens have been chosen, as deliv-
best eligibility of an artificial roughness shape with regard to the
ering the best thermalehydraulic performance, cover the next
flow regime.
ranges: for the corrugated tubes 0.024 < h/d < 0.057 and 0.608 < p/
d < 1.229; for the dimpled tubes 0.083 < h/d < 0.119 and
3. Artificial roughened tubes analyzed 1.650 < d2p/l < 2.639; for the wire coils 0.074 < e/d < 0.101 and
1.173 < p/d < 2.684. Further information on the measurement
Surface roughness is the most common and successful tech- technique, data reduction and uncertainties of the results can be
nique for enhancing tube-side heat transfer in single phase found in [28,29], [30,31] and [32].
turbulent flow. Large scale production of roughened tubes can be
manufactured through cold rolling. In corrugated and dimpled
tubes the roughness height and pitch are controlled by the roller 3.1. Pressure drop results
geometry, the feed rate and the pressure applied during the
process. A tube with a wire coil insert is another simple and cheap Fig. 2 shows the experimental results of isothermal pressure
method of creating a roughened tube. Fig. 1 shows a sketch of the drop for the wire coil, the corrugated tube and the dimpled tube
three enhancement techniques analyzed in this work: corrugated selected, obtained in the hydrodynamically developed region. The
tubes, dimpled tubes and wire coil inserts. experimental set-up was adjusted and verified through pressure
The dimensionless numbers characterizing the geometry of drop experiments with a smooth tube. Laminar results were
corrugated tubes are dimensionless roughness height (h/d) and compared to the analytical solution (fs ¼ 16/Re) while results in the
dimensionless pitch (p/d). For dimpled tubes, the dimensionless turbulent region were compared to Blasius equation
roughness height (h/d) and dimple density (d2/pl) are employed (fs ¼ 0.0791Re0.25). An excellent agreement with the mentioned
and for wire coil inserts, wire diameter (e/d) and dimensionless correlations is observed: 3% for 95% of the experimental data.
The three roughened tubes show the typical behaviour of arti-
ficial roughness techniques: advance of transition to turbulence
and high pressure drop increase in turbulent regime [33]. In
laminar regime, at Reynolds numbers below 350, the three devices
under study increase pressure drop around 30% as a result of an
increase of the skin friction drag: this is due to the reduction of the
cross-sectional area and to the increase of the wet perimeter
(decrease of the hydraulic diameter).
a At Reynolds numbers higher than 3000, the friction factor curve
corresponding to any of the three devices has the typical trend of
the turbulent flow in roughened tubes: fafRe0.2. This implies that
the flow is fully turbulent and the pressure drop is approximately

b
10
Friction factor, f

f=16/Re

10 Wire coil: W01

Dimpled tube: D05
Corrugated tube: C01
Smooth tube f=0.079 Re

c 10
10 10 10 10 10
Reynolds number, Re
Fig. 1. Types of surface roughness that this paper contemplates. (a) Wire coils: helical
pitch p, wire diameter e; (b) Corrugated tubes: corrugation pitch p, corrugation height Fig. 2. Fanning friction factor vs. Reynolds number. Experimental results for the wire
h; (c) Dimpled tubes: corrugation pitch p, length between dimples l, dimple height h. coil, the corrugated tube and the dimpled tube.
 A. García et al. / Applied Thermal Engineering 35 (2012) 196e201 199

proportional to the square of the flow velocity. The wire coil 10

produces higher pressure drop than the deformed tubes: at TURBULENT FLOW
Re ¼ 10000, the friction factor increase is fa/fs ¼ 5 as compared to fa/ Gnielinski [36]

fs ¼ 3.7 obtained in corrugated and dimpled tubes. Pr=16.8

Pr=4.4

It can be stated that the three roughened tubes studied in this Pr=33.5
paper present similar behaviour both in pure laminar and turbulent Pr=74 Pr=2.8

Nusselt number, Nu
regimes. However, roughness shape plays an important role in how
transition occurs. During the performance of the pressure drop 10
experiments, it was observed that the transition to turbulence in LAMINAR FLOW
the dimpled tube took place with strong flow instabilities. These −−− Petukhov & Polyakov [35]
instabilities were not so strong in the experiments carried out in
the corrugated tube. In wire coils, transition occurred smoothly and
without any kind of fluctuation. Fig. 2 illustrates the different
behaviours: the friction factor curve of the dimpled tube presents
a high jump within a limited Reynolds number range, which goes 10
between 1200 and 1600. For the wire coil, the critical Reynolds
number cannot be clearly identified as the jump is very small and it
10 10 10 10
takes place within a wider Reynolds range that extends from 350 to
Reynolds number, Re
2000. However, flow visualization tests in this wire coil performed
by the authors [34] demonstrate that at Re ¼ 700 the flow has Fig. 3. Nusselt number vs. Reynolds number in laminar, transition and turbulent flow.
turbulent characteristics and that at Reynolds numbers between Experimental smooth tube results compared with Petukhov and Polyakov [35] and
350 and 700, the laminar flow is strongly disturbed. Gnielinski [36] equations.
Pressure drop results of the three roughened tubes are quali-
behaviour to that of the smooth tube. However, the onset of the
tatively different and it can be affirmed that this is due to the
buoyancy-driven recirculations is greatly affected by the roughness
roughness shape. Wire coils produce two effects in the flow
shape. In the wire coil, the appearance of a rotational component of
structure: rotation of the core flow and flow separation down-
the velocity was observed [34], which delays the appearance of the
stream of the wire. It is reasonable to affirm that corrugated tubes
mixed convection: it only occurs at Reynolds numbers below 200
produce a slight rotation of the core flow and no flow separation
and at higher Rayleigh numbers than for the smooth tube. In
downstream of the corrugations. On the other hand, the three-
corrugated tubes, this rotational component is weaker and mixed-
dimensional artificial roughness of dimpled tubes is similar to
convection flow does not occur at Reynolds numbers above 700.
natural roughness, which does not generate either rotating flow or
Finally, in the dimpled tube, viscous flow occurs in a similar way to
large-scale separations. It can be stated that the rotation of the core
the smooth tube, and there is not a significant delay in the devel-
flow affects mainly the advance of transition from laminar to
opment of mixed-convection flow.
turbulent flow and its characteristics: smooth or sudden. Moreover,
Figs. 4e6 clearly show great differences in how each enhanced
coil inserts of round wire shape produce flow separations that yield
tube promotes the transition (Region II) from laminar to turbulent
high friction factor coefficients in turbulent flow, suggesting that
regime. Transition from the fully laminar to the turbulent flow
bluff body drag exceeds the skin friction drag on the wall.
takes place smoothly in the wire coil in the Reynolds number range
from 200 to 700. Visualization tests showed that at Re ¼ 700 the
3.2. Heat transfer results
flow is turbulent. For Reynolds numbers between 200 and 700, the
flow remains laminar, but separation occurs downstream of the
This section aims to compare the increase in the heat transfer
wire. The fluid near the wall is mixed and the heat transfer coef-
coefficient produced by the different enhancement techniques in
ficient increases significantly. Conversely, in the dimpled tube, fully
laminar, transition and turbulent regimes. Experimental results are
developed laminar mixed-convection flow is found up to Reynolds
presented in terms of Nusselt number vs. Reynolds number. Firstly,
number 1200. The rotational component on the core flow produced
experimental results of Nusselt number for a smooth horizontal
tube are presented in Fig. 3.
Heat transfer in the laminar regime can occur either in forced
3
convection (continuous line) or in mixed convection (dashed line). 10
I II III
Experimental results in the laminar regime were obtained in mixed Pr=4.1
convection and at Rayleigh numbers from 2107 to 7.5107. These Pr=59 Pr=37
Pr=92
results agree with the correlation of Petukhov and Polyakov [35].
Experiments for the turbulent flow were carried out at five
different Prandtl numbers from 2.8 to 74. The results agree to 2
10
a great extent with the Gnielinski equation [36].
The different heat transfer regions for each of the three
enhanced tubes under study are illustrated in Fig. 4 (dimpled tube),
Fig. 5 (wire coil), and Fig. 6 (corrugated tube). Results are con-
D05: Pr=2.9, 4.1
fronted with the correlations for the smooth tube presented in D05: Pr=92, 59, 37
Fig. 2. 1 Smooth tube, Pr=2.9, 4.1
10 Smooth tube, Pr=92, 59, 37
In laminar flow, heat transfer in horizontal tubes can occur Smooth tube, Ra=constant
either in forced convection (entry region) or in mixed convection
2 3 4 5
(fully developed region), where the flow is affected by the existence 10 10 10 10

of two buoyancy-driven recirculations. The results shown in

Figs. 4e6 for the pure laminar region (Region I) are for mixed- Fig. 4. Nusselt number vs. Reynolds number. Experimental results for the dimpled
convection flow. Here, the three roughened tubes have a similar tube in the: laminar region (I), transition region (II) and turbulent region (III).
200 A. García et al. / Applied Thermal Engineering 35 (2012) 196e201

3 3
10 10
I II III

Pr=11 Pr=4.3 Pr=2.8 ed

Pr=76 rc
Fo
Pr=23
be,
Pr=168 Tu
th
Nusselt number, Nu

Nusselt number, Nu
oo
2 Sm
10
2
10

Smooth Tube, Mixed (Ra=10 )

1
10 ed
Forc
Ra=7.5 10 W01: Pr=168, 23, 4.3 ube, Wire coil W01
oth T
Ra=2.0 10 Smo Corrugated tube C01
W01: Pr=76, 11, 2.8
Smooth tube, Pr=168, 23, 4.3 Dimpled tube D05
1
10 Smooth tube, Pr=76, 11, 2.8
Smooth tube, Ra=constant
0
2 3 4 5
10 1 2 3 4 5
10 10 10 10 10 10 10 10 10
Reynolds number, Re Reynolds number, Re

Fig. 5. Nusselt number vs. Reynolds number. Experimental results for the wire coil in Fig. 7. Nusselt number vs. Reynolds number. Experimental correlations for the wire
the: laminar region (I), transition region (II) and turbulent region (III). coil, the corrugated tube and the dimpled tube.

in corrugated tubes hinders the establishment of the two

buoyancy-driven recirculations: at Reynolds numbers above 700 possible to obtain reliable correlations. Wire coils are the best
heat transfer takes place under forced convection. choice for heat exchangers working in this region since they
In the turbulent regime at Reynolds numbers above 2000 produce the best heat transfer enhancement and they have
(Region III), the assertions by Ravigururajan and Rabas [37] are a predictable behaviour.
validated: wire coil inserts have approximately the same heat At Reynolds numbers above 2000, the deformed tubes produce
transfer coefficient as integral surface roughness. Bergles [38] slightly higher heat transfer coefficients than the wire coil: at
affirms that with these techniques, maximum Nusselt number Re ¼ 10000 and Pr ¼ 200, Nua/Nus ¼ 2.9 for mechanically deformed
augmentations of 250% can be expected at low Prandtl numbers. tubes and Nua/Nus ¼ 2.4 for the wire coil. The correlations
employed offer the next influence of Prandtl number on Nua/Nus:
for dimpled tubes, Nua/NusfPr0.01; for corrugated tubes, Nua/
4. Discussion of results and conclusions NusfPr0.05; for wire coils, Nua/NusfPr0.02. These relations yield
negligible differences between the heat transfer augmentations
Fig. 7 presents the Nusselt number correlations proposed by the reported above and the values averaged over the range of Prandtl
authors [28e32] for the corrugated tube, the dimpled tube and the number investigated. Since wire coil inserts produce the highest
wire coil at Prandtl number 200 in the Reynolds number range friction factor coefficients in the turbulent regime, it is obvious that
from 20 to 20000. they perform worse than corrugated and dimpled tubes. In any
At Reynolds numbers below 200, the use of roughened tubes case, wire coils would find use in many applications since they are
will not produce higher heat transfer coefficients than those easy to install on existing smooth-tube heat exchangers.
produced by smooth tubes. Moreover, wire coils and to a lesser The conclusions of this comparative analysis can be summarized
extent corrugated tubes, can even reduce the heat transfer rate in the next points:
when they delay the establishment of mixed-convection flow.
Therefore the use of these enhancement techniques is not recom-  The roughness shape determines the existence or absence of
mended within this range of Reynolds numbers. a rotational velocity component in the flow and its magnitude.
For Reynolds numbers between 200 and 2000, the authors The core flow rotation affects mainly the advance of transition
recommend the use of wire coil inserts. In this region, Nusselt from laminar to turbulent flow and its characteristics: smooth
number and friction factor curves are continuous and therefore it is or sudden.
 In coil inserts, transition from the fully laminar to the turbulent
3
10
III
flow takes place smoothly. In the Reynolds number range from
I II
Pr=6 Pr=3 200 to 700 the flow remains laminar, but separation occurs
Pr=35
Pr=60 downstream of the wire. This separation promotes heat
Pr=100
transfer enhancement and eventually yields bluff body drag
and high friction factors in turbulent flow.
Nusselt number, Nu

2
 In the pure laminar region, heat transfer in roughened tubes is
10
very similar to that observed in smooth tubes. The rotational
velocity component induced in wire coils and corrugated tubes
hinders the appearance of mixed convection: for wire coils, the
buoyancy-driven recirculations only occur at Reynolds
C01: Pr=3, 6
Ra=15 10
Ra=7.5 10 C01: Pr=35, 60, 100
numbers below 200 and at high Rayleigh numbers. In corru-
1
Ra=3.6 10 Smooth tube, Pr=3, 6 gated tubes, mixed convection is not produced at Reynolds
10 Smooth tube, Pr=35, 60,100
Smooth tube, Ra=constant numbers above 700.
 At Reynolds numbers above 2000 the three roughened tubes
2 3 4 5
10 10 10 10 produce similar heat transfer coefficients, but wire coils have
Reynolds number, Re
higher friction factor coefficients.
Fig. 6. Nusselt number vs. Reynolds number. Experimental results for the corrugated  The roughness shape is the key factor in the selection of
tube in the: laminar region (I), transition region (II) and turbulent region (III). roughened tubes: at Reynolds numbers below 200, the use of
 A. García et al. / Applied Thermal Engineering 35 (2012) 196e201 201

roughened tube is not recommended: smooth tubes will [16] E.B. Esen, N.T. Obot, T.J. Rabas, Enhancement: part I. Heat transfer and pres-
sure drop results for air flow through passages with spirally-shaped rough-
produce the same results; at Reynolds numbers between 200
ness, Journal of Enhanced Heat Transfer 1 (1994) 145e156.
and 2000, the use of wire coils is recommended; at Reynolds [17] E.B. Esen, N.T. Obot, T.J. Rabas, Enhancement: part II. The role of transition to
numbers above 2000, the use of corrugated and dimpled tubes turbulent flow, Journal of Enhanced Heat Transfer 1 (1994) 157e167.
is recommended. In any case, wire coils would find use in many [18] R. Tauscher, F. Mayinger, Enhancement of heat transfer in a plate heat
exchanger by turbulence promoters, in: Proceedings of the International
applications since they are easy to install on existing smooth- Conference on Compact Heat Exchangers for the Process Industries, Begell
tube heat exchangers. House, New York, pp. 243e260, 1997.
[19] D.R. Oliver, Y. Shoji, Heat transfer enhancement in round tubes using three
different tube inserts: non-Newtonian liquids, Transactions of the Institution
The conclusions reported in this work aim to ease the eligibility of Chemical Engineers 70 (1992) 558e564.
of an artificial roughness technique for a given application, [20] T.S. Ravigururajan, A.E. Bergles, Visualization of flow phenomena near
provided that the flow conditions are known. For example, this enhanced surfaces, Journal of Heat Transfer 116 (1994) 54e57.
[21] H.M. Li, K.S. Ye, Y.K. Tan, S.J. Deng, Investigation on tube-side flow visuali-
knowledge is being at present employed for the tube-side zation, friction factors and heat transfer characteristics of helical-ridging
enhancement of flat plate solar collectors with coil inserts [39], tubes. in: U. Grigull, et al. (Eds.), Proceedings of the Seventh International
that typically operate with transitional flow Reynolds numbers. Heat Transfer Conference, vol. 3. Hemisphere Publishing Corp., Washington,
D.C, 1982, pp. 75e80.
[22] C.O. Olsson, B. Sundén, Heat transfer and pressure drop characteristics of ten
radiator tubes, International Journal of Heat and Mass Transfer 39 (1996)
Acknowledgements 3211e3220.
[23] J.P. Meyer, J.A. Olivier, Transitional flow inside enhanced tubes for fully
developed and developing flow with different types of inlet disturbances: part
This research has been partially financed by the DPI2007- I e adiabatic pressure drops, International Journal of Heat and Mass Transfer
66551-C02 grant of the Spanish Ministery of Science and the 54 (2011) 1587e1597.
company “HRS Spiratube”. [24] J.P. Meyer, J.A. Olivier, Transitional flow inside enhanced tubes for fully
developed and developing flow with different types of inlet disturbances: part
IIeheat transfer, International Journal of Heat and Mass Transfer 54 (2011)
1598e1607.
References [25] C. Thianpong, P. Eiamsa-ard, K. Wongcharee, S. Eiamsa-ard, Compound heat
transfer enhancement of a dimpled tube with a twisted tape swirl generator,
[1] J.P. Solano, A. García, P.G. Vicente, A. Viedma, Flow field and heat transfer International Communications in Heat and Mass Transfer 36 (2009) 698e704.
investigation in tubes of heat exchangers with motionless scrapers, Applied [26] S.K. Saha, Thermal and friction characteristics of laminar flow through rect-
Thermal Engineering 31 (2011) 2013e2024. angular and square ducts with transverse ribs and wire coil inserts, Experi-
[2] P. Joshi, K.D.P. Nigam, E. Bruce Nauman, The kenics static mixer: new data and mental Thermal and Fluid Science 34 (2010) 63e72.
proposed correlations, The Chemical Engineering Journal 59 (1995) 265e271. [27] S.K. Saha, Thermal and friction characteristics of turbulent flow through
[3] R.L. Webb, N.H. Kim, Principles of Enhanced Heat Transfer, second ed. Taylor rectangular and square ducts with transverse ribs and wire-coil inserts,
and Francis, New York, 2005. Experimental Thermal and Fluid Science 34 (2010) 575e589.
[4] N.T. Obot, E.B. Esen, T.J. Rabas, The role of transition in determining friction [28] P.G. Vicente, A. García, A. Viedma, Experimental study of mixed convection
and heat transfer in smooth and rough passages, International Journal of Heat and pressure drop in helically dimpled tubes for laminar and transition flow,
and Mass Transfer 33 (1990) 2133e2143. International Journal of Heat and Mass Transfer 45 (2002) 5091e5105.
[5] J.G. Withers, Tube-side heat transfer and pressure drop for tubes having [29] P.G. Vicente, A. García, A. Viedma, Heat transfer and pressure drop for low
helical internal ridging with turbulent/transitional flow of single-phase fluid. Reynolds turbulent flow in helically dimpled tubes, International Journal of
Part 1. Single-helix ridging, Heat Transfer Engineering Journal 2 (1980) 48e58. Heat and Mass Transfer 45 (2002) 543e553.
[6] H. Kuwakara, K. Takahashi, T. Yanagida, T. Nakayama, S. Sugimoto, K. Oizumi, [30] P.G. Vicente, A. García, A. Viedma, Mixed convection heat transfer and
Method of producing a heat transfer tube for single phase-flow, U.S. Patent isothermal pressure drop in corrugated tubes for laminar and transition flow,
4794775 (1989). International Communications in Heat and Mass Transfer 31 (2004) 651e662.
[7] S.B. Uttarwar, M. Raja Rao, Augmentation of laminar flow heat transfer in [31] P.G. Vicente, A. García, A. Viedma, Experimental investigation on heat transfer
tubes by means of wire coil inserts, Journal of Heat Transfer 107 (1985) and frictional characteristics of spirally corrugated tubes in turbulent flow at
930e935. different Prandtl numbers, International Journal of Heat and Mass Transfer
[8] L. Wang, B. Sundén, Performance comparison of some tube inserts, Interna- 47–4 (2004) 671e681.
tional Communications in Heat and Mass Transfer 29 (2002) 45e56. [32] A. García, P.G. Vicente, A. Viedma, Experimental study of heat transfer
[9] M.A. Akhavan-Behabadi, R. Kumar, M.R. Salimpour, R. Azimi, Pressure drop enhancement with wire coil inserts in laminar-transition-turbulent regimes
and heat transfer augmentation due to coiled wire inserts during laminar flow at different Prandtl numbers, International Journal of Heat and Mass Transfer
of oil inside a horizontal tube, International Journal of Thermal Sciences 49 48 (2005) 4640e4651.
(2010) 373e379. [33] R.L. Webb, R.G. Eckert, R.J. Goldstein, Heat transfer and friction in tubes with
[10] A. Barba, S. Rainieri, M. Spiga, Heat transfer enhancement in a corrugated tube, repeated-rib roughness, International Journal of Heat and Mass Transfer 14
International Communications in Heat and Mass Transfer 29 (2002) 313e322. (1971) 601e617.
[11] T.S. Ravigururajan, A.E. Bergles, Development and verification of general [34] A. García, J.P. Solano, P.G. Vicente, A. Viedma, Flow pattern assessment in
correlations for pressure drop and heat transfer in single-phase turbulent flow tubes with wire coil inserts in laminar and transition regimes, International
in enhanced tubes, Experimental Thermal and Fluid Science 13 (1996) 55e70. Journal of Heat and Fluid Flow 28 (2007) 516e525.
[12] Y.F. Zhang, F.Y. Li, Z.M. Liang, Heat transfer in spiral-coil-inserted tubes and its [35] B.S. Petukhov, A.F. Polyakov, Heat Transfer in Turbulent Mixed Convection.
applications, in: advances in heat transfer augmentation and mixed convec- Hemisphere, New York, 1988.
tion, ASME Heat Transfer Division 169 (1991) 31e36. [36] V. Gnielinski, New equations for heat and mass transfer in turbulent pipe and
[13] V.D. Zimparov, N.L. Vulchanov, B. Delov, Heat transfer and friction charac- channel flow, International Chemical Engineering 16 (1976) 359e368.
teristics of spirally corrugated tubes for power plant condensers-1. Experi- [37] T.S. Ravigururajan, T.J. Rabas, Turbulent flow in integrally enhanced tubes,
mental investigation and performance evaluation, International Journal of part 1: comprehensive review and database development, Heat Transfer
Heat and Mass Transfer 34-9 (1991) 2187e2197. Engineering 17 (1996) 19e29.
[14] J. Chen, H. Müller-Steinhagen, G.G. Duffy, Heat transfer enhancement in [38] A.E. Bergles, ExHFT for fourth generation heat transfer technology, Experi-
dimpled tubes, Applied Thermal Engineering 21 (2001) 535e547. mental Thermal and Fluid Science 26 (2002) 335e344.
[15] Y. Wang, Y. He, Y. Lei, R. Li, Heat transfer and friction characteristics for [39] R. Herrero Martín, J. Pérez-García, A. García, F.J. García-Soto, E. López-Galiana,
turbulent flow of dimpled tubes, Chemical Engineering & Technology 32-6 Simulation of an enhanced flat-plate solar liquid collector with wire-coil
(2009) 956e963. insert devices, Solar Energy 85-3 (2011) 455e469.