Available online at www.sciencedirect.

com

Energy Conversion and Management 49 (2008) 8–15

www.elsevier.com/locate/enconman

Heat transfer behaviors in round tube with conical ring inserts

P. Promvonge *

Department of Mechanical Engineering, Faculty of Engineering, King Mongkut’s Institute of Technology Ladkrabang, Bangkok 10520, Thailand

Received 26 October 2006; received in revised form 26 October 2006; accepted 4 June 2007
Available online 20 July 2007

Abstract

To increase convection heat transfer in a uniform heat flux tube by a passive method, several conical rings used as turbulators are
mounted over the test tube. The effects of the conical ring turbulator inserts on the heat transfer rate and friction factor are experimen-
tally investigated in the present work. Conical rings with three different diameter ratios of the ring to tube diameter (d/D = 0.5, 0.6, 0.7)
are introduced in the tests, and for each ratio, the rings are placed with three different arrangements (converging conical ring, referred to
as CR array, diverging conical ring, DR array and converging–diverging conical ring, CDR array). In the experiment, cold air at ambient
condition for Reynolds numbers in a range of 6000–26,000 is passed through the uniform heat flux circular tube. It is found that the ring
to tube diameter ratio and the ring arrays provide a significant effect on the thermal performance of the test tube. The experimental
results demonstrate that the use of conical ring inserts leads to a higher heat transfer rate than that of the plain surface tube, and the
DR array yields a better heat transfer than the others. The results are also correlated in the form of Nusselt number as a function of
Reynolds number, Prandtl number and diameter ratio. An augmentation of up to 197%, 333%, and 237% in Nusselt number is obtained
in the turbulent flow for the CR, DR and CDR arrays, respectively, although the effect of using the conical ring causes a substantial
increase in friction factor.
Ó 2007 Elsevier Ltd. All rights reserved.

Keywords: Heat transfer; Friction factor; Heat exchanger; Conical ring turbulator; Turbulent/reverse flow

1. Introduction lator. It is well established [3–8] that the turbulent/reverse

flow created in the medium facilitates efficient transport
The technology of enhanced heat transfer has received of heat from the tube wall for both the laminar and turbu-
strong attention over the past 3–5 decades; heat transfer lent regimes. Several correlations have been suggested to
augmentation techniques can be applied mainly in the include the effects of increased turbulence due to spiraling
design of more compact heat exchangers found in various flow, increased circulation created with heating due to large
industries, especially refrigeration, automotive and chemi- centrifugal force and turbulator effects. The reverse flow,
cal processes. A passive method of heat transfer enhance- sometimes called ‘‘re-circulation flow’’, device or the turbu-
ment has been of interest for many researchers because lator is widely employed in heat transfer engineering appli-
additional external power is not needed. The passive tech- cations. The effect of reverse flow and boundary layer
nique includes the use of treated surfaces, rough surfaces, disruption (dissipation) is to enhance the heat transfer coef-
extended surfaces, displaced enhancement devices, coiled ficient and momentum transfers. The reverse flow with high
tubes, additives for liquids and gases and swirl flow devices turbulent flow can improve convection of the tube wall by
[1,2]. Convective heat transfer in a tube with turbulator increasing the effective axial Reynolds number, decreasing
inserts is dynamically different from the one without turbu- the cross section flow area and increasing the mean velocity
and temperature gradient. It can help to produce higher
*
Tel.: +662 3264197; fax: +662 3264198.
heat fluxes and momentum transfers not only due to the
E-mail address: [email protected] large effective driving potential force but also the higher

0196-8904/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.enconman.2007.06.009
 P. Promvonge / Energy Conversion and Management 49 (2008) 8–15 9

Nomenclature

A heat transfer surface area, m2 V voltage, volt

Cp specific heat of air, J/kg K V_ volumetric flow rate, m3/s
d small end diameter of conical ring, m g heat transfer enhancement efficiency
D inner diameter of test tube, m m kinematic viscosity, m2/s
f friction factor q density, kg/m3
h heat transfer coefficient, W/m2K l dynamic viscosity, Ns/m2
I current, A
k thermal conductivity of air, W/m K Subscripts
L length of test tube, m a air
l conical ring length, m b bulk
m_ mass flow rate, kg/s conv convection
Nu Nusselt number i inlet
DP pressure drop, Pa m mean
Pr Prandtl number o outlet
Q heat transfer rate, W p plain tube
Re Reynolds number pp pumping power
t thickness of test tube, m t turbulator
T temperature, °C w wall
U mean axial velocity, m/s

pressure drop. The strength of the reverse flow and the for the Nusselt number and friction factor based on their
reattachment position are the main interests in many heat measured data was also reported.
transfer applications such as heat exchangers, combustion The goal of the present work is to investigate the use of
chambers, gas turbine blades and electronic devices. Many conical ring turbulators with different ring to tube diameter
studies have been conducted previously to analyze the heat ratios from both the heat transfer and friction factor points
transfer and friction factor of conical turbulator inserts. of view. This is because the previous work found in the
Most of the early work was concerned with the effect of open literature is focused on one diameter ratio and DR
the conical turbulator on turbulent flow conditions. Yakut array with different pitches only. Thus, the present work
et al. [3] experimentally investigated the effect of conical also investigates the effect of the conical ring arrangements
ring turbulators on the turbulent heat transfer, pressure (CR, DR and CDR arrays) on the enhancement of heat
drop and flow-induced vibrations. Their experiments were transfer by comparing the results of three different ring to
analyzed and presented in terms of the thermal perfor- tube diameter ratios (d/D = 0.5, 0.6, and 0.7) with those
mances of the heat transfer promoters with respect to their of the plain tubes. All of the experiments are conducted
heat transfer enhancement efficiencies for constant pump- at the same inlet conditions with the Reynolds number,
ing power. Yakut and Sahin [4] again reported the flow- based on the tube diameter, in a range of 6000–26,000. In
induced vibration characteristics of conical ring turbulators addition, the results of the heat transfer enhancement effi-
used for heat transfer enhancement in heat exchangers. ciency are also reported for performance evaluation of tur-
They pointed out that the Nusselt number increases with bulators in this work.
increasing Reynolds number, and the maximum heat trans-
fer is obtained for the smallest pitch arrangement. Durmus 2. Experimental description
[5] also studied the effect of cutting out conical turbulators,
placed in a heat exchange tube, on the heat transfer rates 2.1. Experimental set up
with four different types of turbulators and different conical
angles. Ayhan et al. [6] numerically and experimentally The experiments were conducted in an open loop exper-
studied the heat transfer in a tube by means of truncated imental facility as shown in Fig. 1. The loop consisted of a
hollow cone inserts. Eiamsa-ard and Promvonge [7] studied 7.5 kW blower, orifice meter to measure the flow rate and
the effect of V nozzle turbulators on the heat transfer and the heat transfer test section. The copper test tube had a
friction characteristics in a circular tube. They found that length of L = 1250 mm, with 48 mm inner diameter (D),
the nozzle turbulators have a significant effect on the 51 mm outer diameter (Do) and 1.5 mm copper tube thick-
enhancement of heat transfer. Promvonge and Eiamsa- ness (t) as depicted in Fig. 2. The tube was heated by con-
ard [8] investigated the effect of a conical nozzle and snail tinuously wound flexible electrical wire, which provided a
entrance on heat transfer and friction characteristics in a uniform heat flux boundary condition. The electrical out-
uniform heat flux tube, and a new proposed correlation put power was controlled by a variac transformer to obtain
10 P. Promvonge / Energy Conversion and Management 49 (2008) 8–15

Amp meter
Personal Pressure transducer
computer

Data logger Volt meter

Variac transformer

Pressure
Thermocouple tapping
probe

Blower

Inverter
Incline
manometer

Thermocouple set Thermocouple

Pressure tap probe
Pressure tap
Hot air

Thermocouple Cold air

probe Test section

Fig. 1. Schematic diagram of experimental apparatus.

a constant heat flux along the entire length of the test sec- surement unit in conjunction with Chromel-constantan
tion and by keeping the current less than 3 A. The outer thermocouples as can be seen in Fig. 2. Fifteen thermocou-
surface of the test tube was well insulated to minimize con- ples were taped on the local wall of the tube, and the ther-
vective heat loss to the surroundings, and necessary precau- mocouples were placed round the tube to measure the
tions were taken to prevent leakages from the system. The circumferential temperature variation, which was found
inner and outer temperatures of the bulk air were measured to be negligible. The mean wall temperature was deter-
at certain points with a multi-channel temperature mea- mined by means of calculations based on the reading of
the Chromel-constantan thermocouples. In the experi-
ments, the conical ring turbulator arrays used in the pres-
ent work are depicted in Fig. 2a and b. Each conical ring
was made of aluminum with 48 mm (1.0D) in length and
its small end diameters for three sizes were 24 mm (0.5D),
28.8 mm (0.6D) and 38.4 mm (0.7D), with 2 mm uniform
thickness throughout. In each test run, the conical rings
were placed in three type arrays: (1) diverging conical ring
(DR), (2) converging conical ring (CR) and (3) converging–
diverging conical ring (CDR), respectively.

2.2. Experimental procedure

In the apparatus setting above, the inlet bulk air at 25 °C

from a 7.5 kW blower was directed through an orifice
meter and passed to the heat transfer test section. The air
flow rate was measured by the orifice meter, built according
to an ASME standard [9]. Manometric fluid was used in U-
tube manometers with specific gravity (SG) of 0.826 to
ensure reasonably accurate measurement of the low pres-
Fig. 2. Test tube fitted with various conical ring arrays, (a) DR array, (b) sure drop encountered at low Reynolds numbers. The vol-
CR array and (c) CDR array. umetric air flow rates from the blower, situated before the
 P. Promvonge / Energy Conversion and Management 49 (2008) 8–15 11

inlet of the test tube, were adjusted by varying the motor where
speed through the inverter. During the experiments, the T b ¼ ðT o þ T i Þ=2 ð4Þ
bulk air was heated by an adjustable electrical heater
wrapped along the test section. Both the inlet and outlet and
temperatures of the bulk air from the tube were measured h ¼ mC_ p;a ðT o  T i Þ=Að Te w  T b Þ ð5Þ
by multi-channel Chromel-constantan thermocouples, cali- X
brated within ±0.2 °C deviation by thermostat before Te w ¼ T w =15 ð6Þ
being used. It was necessary to measure the temperature in which Tw is the local surface temperature at the outer
at 15 stations altogether on the outer surface of the heat wall of the inner tube. The average surface temperature
transfer test pipe to determine the average Nusselt number. Te w is calculated from 15 points of Tw lined between the in-
In each test run, it was necessary to record the data of tem- let and the exit of the test tube.
perature, volumetric flow rate and pressure drop of the The average Nusselt number, Nu is estimated as follows:
bulk air at steady state conditions in which the inlet air
temperature was maintained at 25 °C. The various charac- Nu ¼ hD=k ð7Þ
teristics of the flow, the Nusselt number, and the Reynolds
The Reynolds number is given by
number were based on the average of tube wall tempera-
ture and outlet air temperature. The local wall temperature, Re ¼ UD=v ð8Þ
inlet and outlet air temperature, the pressure drop across The friction factor, f can be written as
the test section and the air flow velocity were measured
for heat transfer of the heated tube. The average Nusselt DP
f ¼ ð9Þ
numbers were calculated and discussed where all fluid ðL=DÞðqU 2 =2Þ
properties were determined at the overall bulk mean tem- where U is the mean air velocity in the tube. All of the ther-
perature. In order to quantify the uncertainties, the reduced mophysical properties of the air are determined at the over-
data obtained experimentally were determined. The uncer- all bulk air temperature from Eq. (4).
tainty in the data calculation was based on Ref. [10]. The
maximum uncertainties of the non-dimensional parameters 4. Results and discussion
are ±5% for Reynolds number, ±10% for Nusselt number
and ±15% friction. The uncertainty in the axial velocity The results obtained in this study are presented and dis-
measurement was estimated to be less than ±7%, and the cussed in this section. The experimental studies of the heat
pressure has a corresponding estimated uncertainty of transfer and the friction factor in a uniform heat flux tube
±5%, whereas the uncertainty in temperature measurement fitted with conical ring turbulators of three different types
at the tube wall was about ±0.5%. The experimental results of arrays (CR/DR/CDR) and diameter ratios (d/D = 0.5,
were reproducible within these uncertainty ranges. 0.6 and 0.7) are proposed in terms of Nusselt number
(Nu) and friction factor (f).
3. Data reduction
4.1. Verification of the plain tube
In the present work, air is used as the tested fluid and
flowed through a uniform heat flux and insulated tube. Verification of the heat transfer and friction of the plain
The steady state of the heat transfer rate is assumed to tube is performed by comparing with previous correlations
be equal to the heat loss from the test section, which can from the open literature [11] under similar conditions as
be expressed as shown in Figs. 3 and 4, respectively. The present plain tube
Qa ¼ Qconv ð1Þ data is found to be in good agreement with the previous
correlations of Dittus-Boelter, Kay and Petukhov [11] for
where both the Nusselt number and the friction factor. The fol-
Qa ¼ mC
_ p;a ðT o  T i Þ ¼ VI ð2Þ lowing empirical Nusselt number and friction factor corre-
lations presented in Eqs. (10) and (11), respectively, are
The heat supplied by the electrical winding in the test tube derived for the plain tube of this work and found to repre-
is found to be 5–8% higher than the heat absorbed by the sent the experimental data within ±5–10% error limits
fluid for the thermal equilibrium test due to convection
and radiation heat losses from the test section to the sur- Nu ¼ 0:03277Re0:742 Pr0:4 ð10Þ
roundings. Thus, only the heat transfer rate absorbed by f ¼ 0:6165Re 0:317
ð11Þ
the fluid is taken for the internal convective heat transfer
coefficient calculation.
The convection heat transfer rate from the test section 4.2. Influence of conical ring arrangement
can be written as
The effects of the conical ring turbulator arrays (CR/
Qconv ¼ hAð Te w  T b Þ ð3Þ DR/CDR) and the diameter ratios (d/D = 0.5, 0.6, and
12 P. Promvonge / Energy Conversion and Management 49 (2008) 8–15

200 250
DR, d/D=0.5
Plain tube
DR, d/D=0.6
180 Dittus and Boelter(1930)
CDR, d/D=0.5
Kays(1955)
CR, d/D=0.5
160
DR, d/D=0.7

140 200 CDR, d/D=0.6

CR, d/D=0.6
Nusselt number

CDR, d/D=0.7
120
CR, d/D=0.7
Plain tube
100

Nusselt number
150
80

60

40
100
20

0
4000 8000 12000 16000 20000 24000 28000
Reynolds number
50
Fig. 3. Verification of Nusselt number of plain tube.

0
0.7) on the heat transfer coefficient and friction factor char- 4000 8000 12000 16000 20000 24000 28000
Reynolds number
acteristics are depicted in Fig. 5. In the figure, it is observed
that the use of conical ring inserts leads to considerably Fig. 5. Variation of Nusselt number with Reynolds number for various
higher heat transfer rates than the plain tube for all turbulators.
arrangements. This indicates that the effects of reverse flow
and boundary layer disruption can help to enhance the
convection heat transfer and momentum processes. The CDR arrays at a similar d/D ratio. This is likely from the
reverse flow region can improve convection so that it helps higher re-circulation and higher contact surface area
to increase the effective axial Reynolds number, since the between the fluid and the heating wall surface when the
same throughput must be accommodated by a reduced fluid decelerates from the DR type. It is worth noting that
cross sectional area of flow, and it results in more severe
mean velocity and temperature gradients, which produce
higher fluxes of heat and momentum due to the larger effec-
tive driving potential for each. On the other hand, the
10
boundary layer disruption causes a better chaotic mixing DR, d/D=0.5
DR, d/D=0.6
between the core and the wall regions, thus enhancing the CDR, d/D=0.5
9
convective process. In addition, the use of the DR array CR, d/D=0.5
DR, d/D=0.7
provides better heat transfer than that of the CR or 8 CDR, d/D=0.6
CR, d/D=0.6
CDR, d/D=0.7
7 CR, d/D=0.7
Plain tube
0.08
Petukhov 6
Friction factor

0.07 Plain tube

5
0.06

4
Friction factor

0.05

0.04 3

0.03 2

0.02
1
0.01
0
4000 8000 12000 16000 20000 24000 28000
0
4000 8000 12000 16000 20000 24000 28000 Reynolds number
Reynolds number
Fig. 6. Variation of friction factor with Reynolds number for various
Fig. 4. Verification of friction factor of plain tube. turbulators.
 P. Promvonge / Energy Conversion and Management 49 (2008) 8–15 13

1:31
the heat transfer rate of the DR array is found to be higher Nu ¼ 0:09155Re0:655 Pr0:4 ðd=DÞ ð12Þ
than that of the CR and CDR arrays by around 36–45% 0:258 4:4
f ¼ 1:12Re ðd=DÞ ð13Þ
and 19–29%, respectively. Under the present experimental
conditions, the average Nusselt number for the DR array For the diverging conical ring array (DR):
at all ring to tube diameter ratios is found to be about
175–333% more than that of the plain tube.
The influences of the conical ring arrays (CR/DR/ a 250
CDR type) on the friction factor are presented in DR
Fig. 6. As can be seen in the figure, a considerable
200
increase in friction factor occurs when the conical ring
is mounted in the entire length of the test section in com-

Predicted Nu
parison with the plain tube. This mainly comes from the 150
+10%
dissipation of the dynamical pressure of the fluid due to -10%
high viscous losses near the pipe wall and to the forces
exerted by the ring blockage. Moreover, the friction 100
increase is probably due to the secondary flows occurring
as a result of the interaction of pressure forces with iner-
50
tial forces in the boundary layer. For the DR array, it is
observed that the friction factor, having a peak value at
lower diameter ratio, shows a rapid reduction as the Rey- 0
nolds number increases. As expected, the DR arrange- 0 50 100 150 200 250
ment yields a higher pressure loss across the test tube Experimental Nu
than those of the CR and CDR arrays at similar diameter
ratios, especially for low Reynolds number. b 250
CR

4.3. Effect of ring to tube diameter ratio 200

Fig. 5 again presents the effect of the ring to tube diam-

Predicted Nu

+10%
eter ratio on heat transfer rate in the test tube. In the figure, 150
-10%
it is visible that the heat transfer from using the conical ring
tends to increase with reduction of the diameter ratio, apart
100
from rising Reynolds number for all the arrays. This may
be attributed to the higher turbulent intensity imparted to
the flow and a better and rapid mixing of the fluid between 50
the core and the tube surface regions from the turbulent
fluctuation flow and the appearance of reverse flow
between the ring elements, leading to higher temperature 0
0 50 100 150 200 250
gradients. At the lowest diameter ratio of d/D = 0.5, the Experimental Nu
rise of Nusselt number is found to be in a range of 197–
333% more than that of the plain tube while at the diameter c 250
ratios of d/D = 0.6 and 0.7, it is found to be approximately CDR
138–234% and 91–175%, respectively.
200
The effect of the diameter ratio on friction factor in a
uniform heat flux tube is also depicted in Fig. 6. In the fig-
+10%
Predicted Nu

ure, it can be seen that the friction factor values obtained 150
from using three different diameter ratios are in a similar -10%
trend and decrease with the rise of the diameter ratio val-
ues. As can be observed, there is a considerable decrease 100
in the friction factor at around 50% and 75% for using
d/D = 0.6 and 0.7, respectively, instead of d/D = 0.5. The
50
friction factor tends to decrease with increasing Reynolds
number values.
With the help of the experimental data in the test tube, 0
the following empirical Nusselt number and friction factor 0 50 100 150 200 250
relationships are derived for all the conical ring Experimental Nu
arrangements. Fig. 7. Nusselt numbers obtained by the present correlation and exper-
For the converging conical ring array (CR): imental data for (a) DR, (b) CR and (c) CDR arrays.
14 P. Promvonge / Energy Conversion and Management 49 (2008) 8–15

1:32
Nu ¼ 0:863Re0:459 Pr0:4 ðd=DÞ ð14Þ tively. For all, the experimental data are compared at the
0:42 4:31 same pumping power. It is seen in the figure that the
f ¼ 12:52Re ðd=DÞ ð15Þ
enhancement efficiency (g) generally tends to decrease with
For the converging–diverging conical ring array (CDR):
1:34
Nu ¼ 0:1986Re0:586 Pr0:4 ðd=DÞ ð16Þ
a 3.5
4:58 DR
f ¼ 1:038Re0:23 ðd=DÞ ð17Þ
3
Figs. 7 and 8 show comparisons between the present exper-
imental data and the predictions obtained from the present 2.5
correlations, Eqs. (12) through (17). In the figures, the re- +15%

sults of the present work agree reasonably well with the fit-

Predicted f
2 -15%
ted correlations within ± 10–15% in comparison with
experimental data for the friction factor, and within 1.5
±10% for the Nusselt number.
1
4.4. Performance evaluation
0.5
As preliminary design guidance for selection of a tech-
nique, the heat transfer enhancement efficiency can be eval- 0
uated based on the power consumption per unit mass of 0
Experimental f
fluid. The heat transfer enhancement efficiency criteria, g
is defined as the ratio of the heat transfer coefficient for b 3.5
the tube fitted with the turbulator (ht) to that for the plain CR
tube (hp) at a similar pumping power. Following Ref. [3], 3
the criterion can be defined as:
2.5
ðfRe3 Þp ¼ ðfRe3 Þt ð18Þ
+10%
Predicted f

2
The heat transfer enhancement efficiency (g) at a constant -10%
pumping power can be written as follows: 1.5

ht 
g¼  ð19Þ 1
h p pp

Using Eqs. (10), (13), (15), (17) and (18), the Reynolds 0.5
number for the plain tube (Rep) can be written as a func-
tion of the Reynolds number for the conical ring turbulator 0
0 0.5 1 1.5 2 2.5 3 3.5
(Ret):
Experimental f
1:64
Rep ¼ 1:25Re1:02
t ðd=DÞ for CR array ð20Þ
1:6
c 3.5
Rep ¼ 3:07Re0:961
t ðd=DÞ for DR array ð21Þ CDR
1:71 3
Rep ¼ 1:214Re1:03
t ðd=DÞ for CDR array ð22Þ
+10%
Employing Eqs. (10), (12), (14), (16), (19), (20), (21) and 2.5
(22), the enhancement efficiency for conical ring turbula-
Predicted f

2 -10%
tors can be written as

ht 
g ¼  ¼ 2:37Re0:103
t ðd=DÞ0:091 for CR array ð23Þ 1.5
hp pp

ht  1
g ¼  ¼ 11:46Ret0:255 ðd=DÞ0:126 for DR array ð24Þ
hp pp
 0.5
ht 
g ¼  ¼ 5:25Re0:1799
0:076
t ðd=DÞ for CDR array ð25Þ
hp pp 0
0 0.5 1 1.5 2 2.5 3 3.5
Fig. 9 shows the variation of heat transfer enhancement Experimental f
efficiency (g) with Reynolds number for all the conical ring Fig. 8. Friction factors obtained by the present correlation and experi-
arrays (CR/DR/CDR) at d/D = 0.5, 0.6, and 0.7, respec- mental data for (a) DR, (b) CR and (c) CDR arrays.
 P. Promvonge / Energy Conversion and Management 49 (2008) 8–15 15

2
DR, d/D=0.5
d/D = 0.7–0.5 are found to be around 197–333%,
DR, d/D=0.6 138–234% and 91–175%, more than those of the plain tube,
DR, d/D=0.7
1.8 CDR, d/D=0.5
respectively. However, the increase in friction factor is
CDR, d/D=0.6 much higher than that in Nusselt number at the same Rey-
Enhancement Efficiency

CDR, d/D=0.7
nolds number. For a fixed Reynolds number, both the Nus-
1.6 CR, d/D=0.5
CR, d/D=0.6 selt number and friction factor increase with the decrease in
CR, d/D=0.7
diameter ratio. Correlations for the Nuseselt number and
1.4
friction factor are provided and are found to be within
10% with the experimental data for all the conical ring
1.2 arrays and for 6000 < Re < 26,000. In addition, the
enhancement efficiency increases with decreasing Reynolds
1 number and diameter ratio. The maximum values of
enhancement efficiency for using the DR and CR at
0.8 d/D = 0.5, 0.6 and 0.7 are found to be 1.8 and 1.15; 1.76
0 4000 8000 12000 16000 20000 24000 and 1.13 and 1.73 and 1.12; respectively, while the maxi-
Reynolds number
mum one for the CDR is compromised between the maxi-
Fig. 9. Variation of enhancement efficiency with Reynolds number for mum ones for the DR and CR arrays. In all the cases of
various turbulators. conical ring inserts, the heat transfer rates increase at the
expense of high friction losses.
Table 1
Summary of enhancement efficiency of various conical ring turbulators
References
d/D = 0.5 d/D = 0.6 d/D = 0.7 [1] Ray S, Date AW. Friction and heat transfer characteristics of flow
Re = 2000–26,000 Re = 2000–26,000 Re = 2000–26,000 through square duct with twisted tape insert. Int J Heat Mass
Transfer 2003;46(5):889–902.
DR array 1.8–0.93 1.76–0.91 1.73–0.89
[2] Durmus Aydin, Durmus Ayla, Esen M. Investigation of heat transfer
CR array 1.15–0.885 1.13–0.87 1.12–0.857
and pressure drop in a concentric heat exchanger with snail entrance.
CDR array 1.41–0.89 1.39–0.876 1.37–0.866
Appl Therm Eng 2002;22(3):321–2.
[3] Yakut K, Sahin B, Canbazoglu S. Performance and flow-induced
vibration characteristics for conical-ring turbulators. Appl Energy
the rise of Reynolds number for all the turbulator arrays. 2004;79(1):65–76.
The enhancement efficiency increases with the reduction [4] Yakut K, Sahin B. Flow-induced vibration analysis of conical rings
used of heat transfer enhancement in heat exchanger. Appl Energy
of the diameter ratio, especially at lower Reynolds number.
2004;78(3):273–88.
This can be expressed that the conical ring turbulators are [5] Durmus A. Heat transfer and exergy loss in cut out conical
not feasible in terms of energy saving at higher Reynolds turbulators. Energy Convers Manage 2004;45(5):785–96.
number values. It is worth noting that the enhancement [6] Ayhan T, Azak Y, Demirtas C, Ayhan B. Numerical and exper-
efficiency (g) of the DR array is found to be the highest imental investigation of enhancement of turbulent flow heat transfer
in tubes by means of truncated hollow cone inserts. Heat transfer
compared to those of the CR and CDR arrays for all sim-
enhancement of heat exchangers. Kluwer Academic Publishers; 1999.
ilar diameter ratios. The decrease of the average enhance- p. 347–56.
ment efficiency of the CR and CDR arrays below the DR [7] Eiamsa-ard S, Promvonge P. Experimental investigation of heat
array is about 20% and 33%, 16% and 32% and 19% and transfer and friction characteristics in a circular tube fitted with V-
31%, for d/D = 0.5, 0.6 and 0.7, respectively. The details nozzle turbulators. Int Commun Heat Mass Transfer
2006;33(5):591–600.
of the enhancement efficiency for using various conical ring
[8] Promvonge P, Eiamsa-ard S. Heat transfer enhancement in a tube
arrays are summarized in Table 1. with combined conical-nozzle inserts and swirl generator. Energy
Convers Manage 2006;47(18–19):2867–82.
5. Conclusions [9] ASME Standard. Measurement of fluid flow in pipes using orifice,
nozzle and venturi. ASME MFC-3M-1984. United Engineering
Center 345 East 47th Street, New York: 1984:1–56.
Experimental investigations have been conducted to
[10] ANSI/ASME, Measurement uncertainty, PTC 19, 1-1985, Part I,
study the heat transfer in a circular tube with conical ring 1986.
inserts. The mean heat transfer rates obtained from using [11] Incropera F, Dewitt PD. Introduction to heat transfer. third ed. John
the conical rings of DR, CDR and CR arrays with Wiley and Sons Inc.; 1996.