Experimental Thermal and Fluid Science 84 (2017) 144–155

Contents lists available at ScienceDirect

Experimental Thermal and Fluid Science

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

Experimental studies on heat transfer to Newtonian fluids through spiral

coils
Rahul Harishchandra Patil
Department of Mechanical Engineering, Pillai Institute of Information Technology, Engineering, Media Studies & Research, New Panvel 410206, India

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

Article history: Spiral coils are investigated for flow of petroleum base oils SN70, SN150 and SN300 under isothermal
Received 25 October 2016 steady state and a combination of isothermal steady state & non-isothermal unsteady state conditions
Received in revised form 24 January 2017 for ‘‘Graetz problem”. The experiments were conducted with spiral coils with curvature ratios 0.01568,
Accepted 2 February 2017
0.01778, 0.019, 0.02466, 0.02525, 0.02666, 0.028055, 0.02952, 0.03011, under laminar flow condition.
Available online 5 February 2017
A new dimensionless number R0 number is introduced which is proved by the method of dimensionless
analysis. The Nusselt number is correlated with R0 number for Newtonian fluids based on the experimen-
Keywords:
tal data. A number of correlations are developed for Newtonian fluids under isothermal steady state (total
Heat transfer
Laminar flow
130 tests) and a combination of isothermal steady state & non-isothermal unsteady state conditions
Newtonian fluid (total 154 tests). It is found that R0 number characterizes the heat transfer phenomenon very well and
Spiral coil it also simplifies the correlations on curved coils. The developed correlations are compared with the cor-
relations developed by earlier investigators and are found to be in good agreement. These correlations
can be used for the design of spiral coil heat exchangers.
Ó 2017 Elsevier Inc. All rights reserved.

1. Introduction
Fluid flow through curvature pipe of path may increase the heat
transfer effectiveness. This concept of heat transfer is experi-
mented, investigated and correlated to Newtonians fluids under
laminar flow at isothermal steady state and combination of
isothermal steady state & non-isothermal unsteady state condi-
tions. Spiral coils are generally used in heat exchangers, electronic
cooling, chemical reactors, food industry, health industry etc. They
are compact and their heat transfer rate is high against the straight
tube or helical coiled tubes of same length. In a spiral tube the
transfer of heat and pressure drop is dependent upon the following
parameters and dimensionless numbers namely; the curvature
ratio, the number of turns, the Reynold’s number and Prandtl num-
ber. The influence of the above mentioned parameters and dimen-
sionless numbers on heat transfer to Newtonian fluids flowing
through the spiral coils has been described and investigated for
the first time. In spiral coils as compared to other curved coils
the variable centrifugal forces are present. These variable centrifu-
gal forces give rise to variable secondary flow occurring through
the whole length of the coil that ultimately increases the heat
transfer. The early investigations on curved tubes in general were
focused on experimental rather than the analytic or theoretical
E-mail address: rhpatil9@gmail.com
work. Dean [4,5] presented the first analytical proof of the exis-
tence of secondary flow in curved coils. The existence of critical
velocity for a fluid flowing in the circular cross-section straight
pipe was investigated by Reynolds in 1883. After Reynolds, who
studied the straight coil flow characteristics Grindley and Gibson
[1] were first to notice the difference in straight tube flow and
curved coil flow. Eustice [2,3] studied the flow in a transparent
‘U’ tube by injecting a number of coloured dies which flowed along
with the fluid through the ‘U’ tube. Eustice showed that a different
type of flow existed along with the primary flow when the liquid
flowed through the curved pipe. Noble et al. as cited in Kubair
and Kuloor [8] studied the spiral coils without considering the rep-
resentative parameters of the coil. Kubair and Kuloor [8] studied
the flow of fluids flowing through the two spiral coils for ‘Graetz
problem’. Their study depicts the influence of the ratio of length
to average diameter of the spiral coil on the heat transfer. The most
important representative geometrical parameter of the spiral coil
is its curvature ratio. Kubair and Kuloor [11] investigated the two
spiral coils for ‘Graetz problem’ taking into consideration the cur-
vature ratio of these two spiral coils. This is the only correlation
involving curvature ratio term available for ‘Graetz problem’ on
spiral coils. A comparison of helical coil, straight tube and spiral
coil for heat transfer is experimented by Kubair and Kuloor [9].
The performance of spiral coils was found to be better than either
straight tube or helical coil. The same authors [10] have developed

http://dx.doi.org/10.1016/j.expthermflusci.2017.02.002
0894-1777/Ó 2017 Elsevier Inc. All rights reserved.
 R.H. Patil / Experimental Thermal and Fluid Science 84 (2017) 144–155 145

Nomenclature

0 lC
a0 ; b correlation constants Pr Prandtl number ¼ k p
a2; b2 correlation constants Q heat transfer rate (W)
c2; d2 correlation constants q heat flux (W/m)
a3; b3 correlation constants Re Reynold’s number ¼ qv Di =l
c3; d3 correlation constants Recrit:min: minimum critical Reynold’s number
u; e correlation constants Recrit:max: maximum critical Reynold’s number
Cp specific heat at constant pressure (J/kg K) R0 a new dimensionless number ¼ 4mC _ p =pkD
D average diameter of curvature for a spiral (m) t1 inlet temperature of fluid (°C)
Dco maximum diameter of coil (m) t2 outlet temperature of fluid (°C)
Dci minimum diameter of coil (m) Dtlm logarithmic mean temperature (°C)
Di inside diameter of spiral coil tube (m) tp temperature of the inside surface of the spiral coil tube
Do outside diameter of spiral coil tube (m) (°C)
p
De

 Dean number ¼ Re ðDi =DÞ t w1 to t w3 outer wall temperatures of the spiral coil tube (°C)
Di
variable curvature ratio v mean axial velocity (m/s)
D
V volumetric flow rate (m3/s)
Gz Graetz number ¼ mC _ p =kL
hi inside film heat transfer coefficient (W/m2 K)
k fluid thermal conductivity (W/m K) Greek symbols
L tube length (m) D prefix denoting a difference or change
m mass of the fluid (kg) h angle of spiral coiled tube as it advances from centre
m_ mass flow rate (kg/s) (rad)
n number of spiral coil turns l dynamic viscosity (Pa s)
Nu Nusselt number ¼ hDi =k q fluid density (kg/m3)
P pitch of coil (m)

correlations for friction factors under laminar, transition and tur-
bulent flow in spiral coiled tubes for different fluids. Many papers
on spiral coils are generally focused on numerical CFD study, sim-
ulation on various software and then experimentally verifying the
results obtained. Laminar flow and heat transfer in wholly devel-
oped flow in curved tubes was studied by Zapryanov et al. [12].
They have considered the forced convection problems in tubes that
are curved in nature, which are generally utilized in various heat
exchange equipments, cooling, heating systems, chemical reactors,
heat engines and such and other apparatus, devices and equip-
ments. Curvature ratio effect on the heat transfer and development
of flow in the spirally coiled horizontal tubes was investigated by
Naphon et al. [19]. The results that they predicted for the convec-
tive heat transfer and characteristics of flow were in nice agree-
ment with their experiment. Their investigations reveal that the
centrifugal force has an important effect on the enhancement of
transfer of heat and drop in pressure. Due to the centrifugal force
the Nusselt number and drop in pressure obtained from tubes
coiled spirally are 1.49–1.50 times higher than the straight tube.
A review of fluid flow and heat transfer for curved coils is done by
Naphon et al. [18]. An extensive review has been done on spiral
coil in literature by Shah [13] and Kakaҫ [21]. A number of papers
and books on curved coils are referred [6,16,17,20,23,27,29,33] in
the present study. Fluid flow and heat transfer characteristics of
spiral coiled tube in constant wall heat flux was investigated by
Geun-jong et al. [24]. A numerical analysis for investigation of
flow and heat transfer characteristics in tubes with spiral heat
exchanger was undertaken by them. Developing laminar flow in
spiral coil tubes is investigated by Altac and Altun [25]. They have
numerically investigated steady state combined developing flow
and heat transfer in spiral tube coils. The study on spiral coils
is limited in open literature and there is a need to study the dif-
ferent aspects of fluid flow and heat transfer in spiral coils (see
Table 1).

Table 1
Developed correlations by earlier researchers on spiral and helical coils.

Investigator Curvature ratio Type of flow Regime Correlations developed and notes.
Noble et al. as cited in [8] 0.045 Laminar Graetz Problem Nu ¼ 0:215ðGz0:741 Þ
1200 < Gz < 5000.
Geometrical parameters of spiral coils are not considered
Kubair et al. [8] 0.022–0.031 Laminar Graetz Problem Nu ¼ ½0:27ðL=DÞ  2:3Gz0:6
11 6 Gz < 180
170 < Re < 6000
Nusselt Numbers are calculated at arithmetic mean temperature.
Length, as a geometrical parameter is taken into consideration
Kubair et al. [11] 0.022–0.031 Laminar Graetz Problem Nu ¼ ½ð1:98 þ 1:8ðDi =DÞGz0:7
9 6 Gz < 1000
80 < Re < 6000
20 < Pr < 100
Nusselt numbers are calculated at arithmetic mean temperature.
Effect of curvature ratio on heat transfer is studied
Rainieri et al. [35] 0.06 Laminar Graetz Problem Nu ¼ 1:168ðDe0:47 ÞPr 0:16
12 < De < 280
125 < Pr < 300
Corrugated helical coils were investigated, highly viscous fluids were used
146 R.H. Patil / Experimental Thermal and Fluid Science 84 (2017) 144–155

2. Materials and methods
2.1. Experimental set-up & procedure
Fig. 1 shows the schematic diagram of the spiral coil with the
secondary flow pattern. Fig. 2 is a schematic diagram of the set-
up. Fig. 3 is the actual view of the experimental set-up and the spi-
ral coil-3. The experimental set-up consists of a helical coil heat
exchanger, a steam boiler, water tanks, oil tank, gear pump, water
pump, and stainless steel pipeline fittings and pipeline. The steam
pipeline is of size 16 mm and oil circuit pipeline is of size 12 mm.
The instruments used in the experiment are three rotameters of
range, 0.5–5 LPH, 5–50 LPH, and 50–500 LPH, used as flow indicat-
ing devices. The flow rate is measured by collecting a known
amount of fluid in a measuring cylinder fitted on the pipeline for
known time. Total six RTD’s are used in the experiment. Three
RTD’s, are used to measure the outside wall surface temperature
of the spiral coiled tubes. The maximum deviation in outside wall
temperature of the spiral coils along the length was found to be
1 °C. Kubair and Kuloor [8,11] had used three thermocouples in
their experiments, whereas Janssen and Hoogendoorn [38] and
Pawar et al. [22] had used four thermocouples in their experiment
on helical coils to measure the wall temperature. This justifies the
use of three RTD’s in the present experiment to measure the wall
temperature. The RTD’s used to measure the wall surface temper-
ature are placed at the inlet, middle and outlet position on the coil.
The temperature in the steam chamber is sensed by an RTD and
two RTD’s are used to measure the inlet and outlet temperature
of the test fluid. The RTD’s used are calibrated; their accuracy is
±0.1 °C also the temperature indicators used have an accuracy of
±0.1 °C The range of experimental conditions used in the study is
given in Table 3. A number of experiments were performed by

Fig. 1. Schematic diagram of the spiral coil with secondary flow pattern.

Fig. 2. Schematic diagram of the set-up.

 R.H. Patil / Experimental Thermal and Fluid Science 84 (2017) 144–155
Fig. 3. An actual view of the experimental set-up and spiral coil-3 used in the experiment.
employing different oils namely Coconut oil, Soyabean oil, Biodie-
sel and Diesel oil. Every oil had its own advantages and disadvan-
tages. Coconut and Soyabean oils were cheap but their visibility in
rotameter glass tube became poor after repeated use. Diesel and
Biodiesel oil were also economical; their visibility through rotame-
ter was also good but since their flash point was low possibility of
fire hazard was always there hence, it was not safe to use these
oils. Finally SN70, SN150 and SN300 petroleum base oils with high
viscosity and low heat transfer coefficient were chosen for the
experiment, these oils were also used by different researchers ear-
lier in their work. A number of experiments were performed in
order to check the reproducibility of the readings and it was
noticed that the reproducibility was good. A summary of critical
Reynold’s number for curved coils is given in Table 4. The critical
Reynold’s number is used to find the transition from laminar flow
to turbulent flow. In present investigation the correlation proposed
by Kubair and Kuloor [8] is used to judge the type of flow; laminar
or turbulent. The test fluids, petroleum base oils SN70, SN150 and
SN300 are pumped by a gear pump (supremo make) 2800 RPM,
capacity of 360 LPH. The pump is driven by a motor (Lawkim, God-
rej make) of 1.5 kW, 1440 RPM using a ‘v’ belt and pulley arrange-
ment. The oil flows through a stainless steel pipeline and enters the
spiral coil which is placed inside the steam chamber. An isothermal
calming section whose length is determined by the equation given
by Austin and Seader [7] is provided. The length of the upstream
and downstream isothermal calming section provided was 0.7 m.
The spiral coil is first placed and fitted in the steam chamber.
147
The outer wall of the spiral coil is maintained at a constant temper-
ature by injecting steam with the help of boiler inside the steam
chamber for isothermal steady state condition reading. Isothermal
steady state and non-isothermal unsteady state conditions are the
boundary conditions to which the outer wall of the spiral coil is
subjected. The pressure inside the steam chamber is maintained
constant to one bar. The temperature of steam inside the steam
chamber varies from 99 to 101 °C. The fluid which enters the spiral
coil first passes through the isothermal calming section and then
through the heating section of the spiral coiled tubes. Thus the
velocity profile gets fully developed and the temperature profile
starts developing at the entrance of the spiral coil. This is the
‘Graetz Problem’. After passing through the spiral coil the fluid
passes to the heat exchanger where it is cooled to the temperature
of the oil in the oil storage tank. The water which is supplied to the
heat exchanger for cooling the oil is circulated by the water pump,
(Kirloskar make) 0.37 kW. After passing through the heat exchan-
ger the hot water is cooled in the cooling tower to the temperature
of the water in the water tank i.e., 30–35 °C. A separate galvanized
water tank of 0.2 m3 capacity is provided to supply the feed water
to the boiler. Table 2 gives the list of nine spiral coils along with
their dimensions used in the experiment. The readings are taken
for different flow rates by placing one coil at a time in the steam
chamber. About 13–15 readings per coil are recorded. For non-
isothermal unsteady state condition the wall temperature of the
spiral coil continuously goes on changing with flow rate remaining
constant. First reading is taken at 60 °C and other readings are
148 R.H. Patil / Experimental Thermal and Fluid Science 84 (2017) 144–155

Table 2
The physical dimensions of the spiral coils used in the experiment.

Spiral Maximum diameter Minimum diameter Average diameter Pitch of Number of Outer diameter of Inner diameter of Curvature


coil no. of coil Dco of coil Dci of coil D coil P P turns n coil tube Do coil tube Di ratio DDi

1 0.71 0.13 0.42 0.026 12 0.00985 0.008 0.019

2 0.55 0.11 0.33 0.0185 12 0.00791 0.0059 0.01778
3 0.41 0.1 0.26 0.0125 12 0.00591 0.004 0.01568
4 0.52 0.12 0.32 0.0505 4.375 0.00985 0.008 0.02466
5 0.34 0.13 0.23 0.031 4 0.00791 0.0059 0.02525
6 0.21 0.1 0.16 0.0125 4 0.00591 0.004 0.02666
7 0.47 0.098 0.29 0.015 12 0.00985 0.008 0.028055
8 0.32 0.08 0.2 0.0105 9.6 0.00791 0.0059 0.02952
9 0.18 0.09 0.13 0.0075 5.35 0.00591 0.004 0.03011

taken at 70, 80 and 90 °C. For isothermal steady state, after first The technique of dimensional analysis is used to find the partic-
reading second flow rate is selected and second reading is taken ular form of the correlation required. According to Buckingham’s Pi
and so on all the readings are taken. theorem the desired form of correlation is as follows:

2.2. Determination of the coefficient of heat transfer Nu ¼ a0 Gzb0

0
The constants a0 &b in the above equation are determined by
The thermo physical properties of the fluid are assumed to be using the least square power law fitting technique, depending on
constant along the length of the spiral coil and are found at the the experimental data obtained. A log-log plot of Gz against Nu is
fluid bulk temperature. The properties are taken from Andrew shown in Fig. 4. The trend line on the plot is the curve fitting line
[14]. The heat which is absorbed by the liquid while fluid flowing through the experimental readings, using the least square power
through the spiral coiled tube would be same as the heat passing law fitting technique. The Pearson’s correlation coefficient is
in the pipe from all the directions at right angle to the spiral axis: 0.97351. The following correlation is obtained from the plot:
_ p ðt2  t1 Þ ¼ hi Ai Dt lm
Q ¼ mC ð1Þ
Nu ¼ 2:96 Gz0:545 ð6Þ
where
Dt lm , is the LMTD temperature and is calculated from the for
equation:  
2 3 Di
3 6 Gz 6 786; 0:01568 6 6 0:03011; 12 < Re < 6013:
ð D t  D t Þ D
¼4 h i 5;
2 1
Dtlm ð2Þ
ln DDtt21 There is no need of a separate Prandtl number term in the above
correlation when Graetz number is used in the analysis, since
where Prandtl number is already present in the Graetz number and it is
enough to characterize the heat transfer phenomenon in spiral
Dt2 ¼ tp  t 1 coils. This is also confirmed by the various correlations available
Dt1 ¼ tp  t 2 in literature on helical coils. The curvature ratio term is not consid-


ered in the above correlation. Fig. 5 is a log-log plot of DDi against
where tp is the inside surface temperature of the spiral coiled tube
and is calculated by equation given by Kern [15]. The equation is Nu=Gz0:545 . The Pearson’s correlation coefficient is 0.917. The fol-
given below: lowing correlation is obtained from the plot:
   0:48
2:3q Do Di
tp ¼ tw  log ð3Þ Nu ¼ 0:495 Gz0:545 ð7Þ
2p k Di D
where q (Watts/meter) = Q =L
After determination of Dt lm , hi is calculated by the formula as for
stated below:  
Di
_ p ½t 2  t 1 
mC 3 6 Gz 6 786; 0:01568 6 6 0:03011; 12 < Re < 6013:
hi ¼ ð4Þ D
ðAi  Dtlm Þ
The hydrodynamics of flow and the heat transfer characteriza-
Once inside heat transfer coefficient is known all the unknown tion for curved coils is done in literature by correlating Dean num-
dimensionless numbers are determined since all other data is ber with friction factors and Nusselt numbers respectively. An
already known. attempt is done in the present paper to find the most general form
of correlation involving Dean number for spiral coil geometry. A
3. Results and discussion log-log plot of De against Nu=Pr0:4 is shown in Fig. 6. The trend line
on the plot is the curve fitting line through the experimental read-
Experiments are carried in laminar flow with petroleum base ings, using the least square power law fitting technique. The fol-
oils SN70, SN150 & SN300 on nine spiral coils under isothermal lowing correlation is obtained from the plot, Nu ¼ 0:219De0:639 .
steady state and a combination of isothermal steady state & non-

Fig. 7. is a log-log plot of DDi against Nu=ðPr 0:4 De0:639 Þ and Fig. 8.
isothermal unsteady state conditions. The thermo physical proper-

0:378
ties of the fluids SN70, SN150 & SN300 @ 40 °C are given in Table 5. is a log-log plot of n against Nu=ðPr 0:4 De0:639 DDi Þ. The Pearson’s
The dimensionless Graetz number is defined as:
correlation coefficient is 0.9699. The following correlation is
_
Gz ¼ mCp=kL ¼ ReðPrÞðDi =LÞðp=4Þ ð5Þ obtained from the plot:
 R.H. Patil / Experimental Thermal and Fluid Science 84 (2017) 144–155 149

Table 3
The experimental conditions.

Coil No. Fluid used Rate of flow (kg/s) Experimental condition

1 Petroleum base oil SN150, SN300 0.0017–0.1 Isothermal
Petroleum base oil SN70 0.0139–0.0739 Non-isothermal
2 Petroleum base oil SN150 0.001902–0.065116 Isothermal
3 Petroleum base oil SN150 0.001879–0.049267 Isothermal
4 Petroleum base oil SN70, SN150 0.001846–0.12445 Isothermal
Petroleum base oil SN70 0.007423–0.082926 Non-isothermal
5 Petroleum base oil SN70, SN150 0.002056–0.109677 Isothermal
Petroleum base oil SN70 0.02222–0.109677 Non-isothermal
6 Petroleum base oil SN70, SN150 0.002444–0.1 Isothermal
7 Petroleum base oil SN150 0.001766–0.079091 Isothermal
8 Petroleum base oil SN150 0.001941–0.082857 Isothermal
9 Petroleum base oil SN150, SN70 0.001935–0.11333 Isothermal

Table 4
Critical Reynold’s number for spiral coils according to the correlations of various investigators.

Coil No. Dco/Di Dci/Di Di/D Ito [28] Kubair and Srinivasan et al. [30], Srinivasan et al. [30], Schmidt [32] Kubair and
Varrier [31] Recrit:min: @ (Dco/2) Recrit:max: @ (Dci/2) Kuloor [8]
1 88.75 16.25 0.019 4448.25 3581.21 4774.95 8351.35 7864.52 5762.08
2 93.22 18.64 0.0177 4352.68 3505.96 4710.03 7936.84 7747.59 5686.10
3 102.5 65 0.0156 4181.78 3367.74 4589.08 5225.68 7532.83 5544.95
4 65 15 0.0246 4862.85 3892.84 5225.68 8606.61 8348.72 6070.54
5 57.63 22.03 0.0252 4903.76 3922.40 5419.53 7469.00 8394.67 6099.31
6 52.5 40 0.0266 4999.99 3991.20 5577.93 6084.47 8501.57 6165.96
7 58.75 36.25 0.0280 5092.88 4056.64 5387.73 6285.49 8603.25 6228.95
8 54.24 13.56 0.0295 5189.14 4123.50 5521.69 8943.38 8707.13 6292.91
9 45 22.5 0.0301 5227.23 4149.69 5856.59 7412.63 8747.83 6317.87

Table 5
Thermo-physical properties of petroleum base oils @ 40 °C.

Base oil m (Pa s) k (W/m K) Cp (J/kg K) q (kg/m3) Flash point °C Pour point °C
SN70 0.011–0.012 0.13 2200 849 190 12
SN150 0.030–0.032 0.1341 2177 850 210 15
SN300 0.058–0.060 0.13 2100 870 200 12

Fig. 4. A log-log plot of Gz vs. Nu used to develop correlation (6) under isothermal steady state condition.
150 R.H. Patil / Experimental Thermal and Fluid Science 84 (2017) 144–155

Fig. 5. A log-log plot of (Di/D) vs. Nu/Gz0545 used to develop correlation (7) under isothermal steady state condition.

Fig. 6. A log-log plot of De vs. Nu/Pr0.4 used to develop correlation (8) under isothermal steady state condition.

Fig. 7. A log-log plot of (Di/D) vs. Nu/Pr04 De0639 used to develop correlation (8) under isothermal steady state condition.
 R.H. Patil / Experimental Thermal and Fluid Science 84 (2017) 144–155 151

Fig. 8. A log-log plot of n vs. Nu/(Di/D)0378 Pr04 De0639 used to develop correlation (8) under isothermal steady state condition.

 0:378 e
Nu ¼ u½R0  ¼ u½qVD2i C p =kDÞ ¼ u½RePrðDi =DÞe
e
Di ð12Þ
Nu ¼ 1:014 n0:07 Pr 0:4 De0:639 ð8Þ
D
where u and e are constants that can be found by performing
for experiments. Fig. 9 is a log-log plot of R0 against Nu. The trend line
is obtained by least square power law fitting method. The Pearson’s
 
Di correlation coefficient is 0.98082. The following correlation is
3:35 6 De 6 1043; 0:01568 6 6 0:03011; 12 < Re < 6012 obtained from the plot:
D
Eqs. (6)–(8) are developed for isothermal steady state condition Nu ¼ 0:222R00:646 ð13Þ
in order to compare them with a newly introduced dimensionless
number the R0 number which is proved on the basis of Bucking- for
ham’s Pi theorem. It is observed that the convective heat transfer
coefficient ‘h’ depends on the following variables in the present  
Di
experiment; the mass flow rate of the fluid m, _ the specific heat 117 6 R0 6 18382; 0:01568 6 6 0:03011; 12 < Re < 6013:87
D
capacity of the fluid C p , the thermal conductivity of the fluid k,
the internal diameter of the spiral coil tube Di and the average
Eq. (13) is compared with Eqs. (6)–(8) and it is found that Eq.
diameter of the spiral coil D. Hence:
(13) deviated least from the experimental data and also from the
_ C p ; k; Di ; DÞ
h ¼ f ðm; ð9Þ correlations of Kubair and Kuloor [8,11] as shown in Fig. 13. The
form of Eq. (13) is simple and it contains less number of variables
The number of variables are six the prime variables selected are
than Eqs. (6)–(8). There is no need of a separate ðDi =DÞ term in the
two namely h and D. The repeating variables selected are four
correlation and R0 number is enough to accurately predict the heat
_ C p ; k&Di Þ there are thus two dimensional groups:
namely ðm;
transfer phenomenon occurring in spiral coils. The oils used in the
p1 ¼ hðm_ a ; C bp ; kc ; and Ddi Þ; p2 ¼ Dðm_ a1 ; C b1 c1 d1
p ; k ; and Di Þ
present experiment are viscous in nature hence a correction is nec-
essary to incorporate, the variation in temperature dependent
wherea; b; c; d; a1; b1; c1&d1, are the exponents of the repeating property relations existing in the experiment. This correction is
variables. After analysis according to Buckingham’s Pi theorem the done by multiplying the right hand side of Eq. (13) by (mb/mw)1/6.
two non-dimensional numbers obtained are Nusselt number The exponent (1/6) is taken from Kuzneltsova as cited in Kakaҫ
Nu ¼ ðhDi Þ=k and, [21]. He has used oils with nearly the same viscosities as in the pre-
! sent experiment. It is to be noted that for viscous oils and for spi-
_ p
4mC qVD2i C p
R0 number ¼ ¼ ð10Þ rals, the viscosity correction is higher than the viscosity correction
pkD kD in helical coils according to Kubair & Kuloor [26]. In Eq. (13) the
effect of characteristic dimensions of the spiral coils excluding
This dimensional number is not available in literature. The role
the curvature ratio is not considered hence, a most general corre-
played by R0 number in convective heat transfer in curved tubes is
lation including the influence of curvature ratio, Prandtl number
similar to the role played by Graetz number in straight tube
and number of turns on heat transfer to a fluid flowing through spi-
convective heat transfer. The two non-dimensional numbers
ral coils under isothermal steady state condition is considered. The
Nu and R0 numbers can be functionally written as, form of correlation as per Buckingham’s Pi theorem is given below:
e
Nu ¼ u½R0  ¼ u½qVD2i C p =kDÞ
e
ð11Þ
Nu ¼ a2nb2 R0c2 ð14Þ

e
l
Synthetically, multiplying the right hand side of Eq. (8) by l The constants a2, b2 and c2 are found by performing experi-
results in the following equation: ments. Fig. 10 is a log-log plot of n against Nu=R00:646 . The Pearson’s
152 R.H. Patil / Experimental Thermal and Fluid Science 84 (2017) 144–155

Fig. 9. A log-log plot of R0 vs. Nu used to develop correlation (13) under isothermal steady state condition.

Fig. 10. A log-log plot of n vs. Nu/R0 0.646 used to develop correlation (15) under isothermal steady state condition.

correlation coefficient is 0.98012. The following correlation is Nu ¼ 0:225R00:644 ð16Þ

obtained from the plot:
for;
Nu ¼ 0:281R00:646 n0:16 ð15Þ
 
Di
for 117 6 R0 6 18382; 0:01568 6 6 0:03011; 12 < Re < 6013:87
D
 
Di
117 6 R0 6 18382; 0:01568 6 6 0:03011; 12 < Re < 6013:87 To account for the variable property relation the right hand side
D
of Eq. (16) should be multiplied by (mb/mw)1/6. When the effect of
number of turns of spiral coils is considered the form of correlation
To account for the variable property relation the right hand side
obtained is:
of Eq. (15) should be multiplied by (mb/mw)1/6. In the similar way Eq.
(16) is obtained by considering the combination of isothermal Nu ¼ a3nb3 R0c3 ð17Þ
steady state and non-isothermal unsteady state data points as
The constants a3, b3 and c3 are found by performing experi-
shown in Fig. 11 which is graph of R0 against Nu. The least square
power law fitting method is used to plot the trend line. The Pear- ments. Fig. 12 is a log-log plot of n against Nu=R00:644 . The Pearson’s
son’s correlation coefficient is 0.98118. The following correlation correlation coefficient is 0.98016. The following correlation is
is obtained from the plot: obtained from the plot:
 R.H. Patil / Experimental Thermal and Fluid Science 84 (2017) 144–155 153

Fig. 11. A log-log plot of R0 vs. Nu used to develop correlation (16) under isothermal steady state condition and non-isothermal unsteady state condition.

Fig. 12. A log-log plot of n vs. Nu/R0 0.644 used to develop correlation (18) under isothermal steady state and non-isothermal unsteady state condition.

Nu ¼ 0:286n0:17 R00:644
for
 
Di
117 6 R0 6 18382; 0:01568 6 6 0:03011; 12 < Re < 6013:87
D
It is observed that the exponent of R0 number in Eq. (16) is lower
than the exponent of R0 number in Eq. (13), this is because of low
values of Nusselt numbers obtained in non-isothermal heat trans-
fer condition. To account for the variable property relation the
right hand side of Eq. (18) should be multiplied by (mb/mw)1/6. Heat
exchangers are expected to work on isothermal operating condi-
tions but because of disturbances in flow due to sudden stopping
or starting the equipment or due to fluctuations in pump speed
etc. the heat exchanger no longer operates on isothermal condi-
tion. A combination of isothermal steady state and non-
isothermal unsteady state condition exists and a correlation to deal
with such a situation is necessary hence the two correlations Eqs.
(16), (18) are developed. The differences in Nusselt numbers in
isothermal and non-isothermal conditions are less because of
ð18Þ intense mixing of fluid particles in the spiral coils due to variable
secondary forces present. Fig. 13 is a plot of Re against Nu. The plot
compares the value of Nusselt numbers obtained by Eqs. (6)–(8),
(13), (15), (16), (18) against the Nusselt number values as pre-
dicted by the two correlations proposed by Kubair and Kuloor
[8,11]. The plot reveals that there is a good agreement between
the values of Nusselt numbers as predicted by developed correla-
tions in the present experiment and the values of Nusselt numbers
predicted by the correlations of Kubair and Kuloor. The values of
Nusselt numbers as obtained from the correlation of Kubair and
Kuloor [11] are 19.59%, 27.75%, 16.17%, 18.21%, 20.84%, 18.56%,
and 22.16%. These values are all higher than the corresponding val-
ues as developed by Eqs. (6)–(8), (13), (15), (16), (18) for coil-9.
Similar trend is observed in other coils (not shown here). Predicted
Nusselt number values by Eqs. (6)-(8), (13), (15), (16), (18) with the
experimental values of Nusselt number shows an acceptable range
of deviation of ±12%, for coil-9. Similarly, the values of Nusselt
numbers as obtained from the correlation of Kubair and Kuloor
[8] are 13.87%, 2.32%, 13.78, 16.64%, 12.89%, 16.13% and 11%. These
values are all lower than the corresponding values as developed by
Eqs. (6)–(8), (13), (15), (16), (18) for coil-9. Similar trend is
154 R.H. Patil / Experimental Thermal and Fluid Science 84 (2017) 144–155
Fig. 13. A plot of Re vs. Nu for comparison of developed correlations Eqs. (6)–(8), (13), (15), (16), (18) based on present experimental readings (for Re = 12–6013, (Di/D) =
0.01568–0.03011, R0 = 117–18382 with the works of earlier investigators for coil-9.
observed in other coils (not shown here). Eqs. (6)–(8), (13), (15),
(16), (18) are also compared with the equation of Rainieri et al.
[35] given in Table 2, the equation is for helical coils with corruga-
tions. The Nusselt numbers predicted by this equation are less than
all other equations developed on spiral coils. This indicates that
spiral coils show better performance than helical coils for the given
conditions, this is in accordance to the findings of Kubair and
Kuloor [9]. It is seen from Fig. 13 that the deviations are least for
Eq. (13) which implies that only R0 number is enough to character-
ize the heat transfer phenomenon in spiral coils or curved coils.
There is no need of a separate ðDi =DÞ or Prandtl number or geomet-
rical parameters such as pitch, length or number of turns of the spi-
ral coils to be considered in the correlations on spiral coils. The
main problem encountered while dealing with ‘Graetz Problem’
is the uneven distribution of temperature along the circumference
of the spiral coil wall mainly because of the local heat transfer coef-
ficient which do not decrease continuously with distance along the
length of the spirally coiled tube, but it undergoes cyclic oscilla-
tions along the axial distance. The oscillations get damped in fully
developed thermal region. This phenomenon was observed by
investigators such as, Seban and McLaughlin [36], Dravid et al.
[37], Janssen and Hoogendoorn [38], Tarbell and Samuels [39]
and Akiyama and Cheng [40]. Seban and McLaughlin [36] were
unable to give an explanation for the oscillations. According to Dra-
vid et al. [37] the reason for the occurrence of thermal oscillations
was due to the exposure of the tube wall to alternately hot and
cold fluids because of secondary flow. Dravid et al. [37] have rec-
ommended their correlation on helical coils for design purpose
by neglecting the effects of thermal oscillations. Janssen and
Hoogendoorn [38] and Kubair and Kuloor [8,11] have also
neglected the effect of thermal oscillations in their correlations.
In general thermal oscillations increase because of higher pitch val-
ues of the spiral coiled tube, lower values of curvature ratios, tube
material of low thermal conductivity and thin walled tubes
although no quantitative data is available in this respect. In the
present experiment while designing the spiral coiled tubes, low
values of pitch, higher curvature ratios, high thermal conductivity
material and tube of medium thickness was selected so that the
effect of thermal oscillations was minimized in the final correla-
tions. The effect of thermal oscillations in the present experiment
was negligible, since the calculations indicated that the variation
of temperature around the circumference of the spiral coiled tubes
was (0.1 °C). Another possible explanation to neglect the problem
of thermal oscillations in the present experiment is the existence
of variable secondary flow occurring in spiral coiled tubes. The
presence of variable centrifugal force on the fluid in the same spiral
coil assists in variable secondary flow due to which the fluid parti-
cles tend to get mixed properly thereby increasing the possibility
of heat and momentum transfer within the fluid to a great extent.
Also, the contribution of this variable secondary flow in enhancing
the rate of momentum transfer is greater in laminar flow. The
effect of all the factors discussed above is to distribute the temper-
ature of the fluid uniformly around the circumference of the inside
tube wall and thus to damp the thermal oscillations. The uncer-
tainty in experimental results is calculated by the method of Kline
and Mc Clintock as cited by Holman [34]. In this method the uncer-
tainty in results is based on the uncertainty in the primary exper-
imental measurements. The instruments were calibrated in order
to meet the specifications by taking repeated trials. The calculation
of uncertainty was based on Eq. (3.2) of Holman [34]. The calcu-
lated value of uncertainty was ±3% for heat transfer rate (Q), ±5%
for inside film heat transfer coefficient (hi), ±5% for Nusselt number
(Nu) and ±2.5% for volumetric flow rate (V). The overall effect of
these uncertainties on the correlations developed may be consid-
ered negligible.
3.1. Interpretation
In the present investigation considering the geometry of the
coils the effect of Reynold’s number on the Nusselt number of
the fluid flowing through coils (1,2,3) is less when compared to
their average coil radius. An increase in the curvature ratio
increases the heat transfer to the oil because of increase in cen-
trifugal force as the fluid flows through a coil of high curvature
ratio. The centrifugal force which is the ratio of the product of mass
of the flowing fluid and the square of the fluid velocity to the radius
of spiral coil increases, as the mean radius of the coil decreases.
Centrifugal force increases the magnitude of secondary flow and
also because of the variable magnitude of centrifugal force exists
in spiral coils, the particles of the fluid mix intensely which
 R.H. Patil / Experimental Thermal and Fluid Science 84 (2017) 144–155
enhance the heat transfer. The effect of pitch on heat transfer in
spiral coils is negligible this is also in case of helical coils as
observed by many investigators Shah [13]. The effect of the num-
ber of turns in coils (7,8,9) on the heat transfer to the oils is less
than the effect of curvature ratio of the coil but their effect cannot
be neglected for the completeness of the correlation equations.
There is no need of an involvement of a separate curvature ratio
term or Prandtl number term in the correlations involving R0 num-
ber. For curved coils only R0 number is enough to accurately predict
the heat transfer coefficients in industrial spiral coil heat exchang-
ers. This also eliminates the need to assume the exponent of
Prandtl number as 0.4 which is done in literature. R0 number is
thus, more appropriate to characterize the heat transfer phe-
nomenon in curved coils.
4. Conclusion
Experimental set-up is fabricated to study the heat transfer to
Newtonian fluid flowing through spiral coiled tubes under isother-
mal steady state and a combination of isothermal steady state and
non-isothermal unsteady state condition for laminar flow condi-
tion. An innovative approach of correlating Nusselt numbers with
a newly introduced R0 number is presented on the basis of experi-
mental findings. R0 number is first proved by Buckingham’s Pi the-
orem and then correlations considering the effect of R0 number, not
available in literature are developed. The dimensionless R0 number
is found very useful to characterize the heat transfer phenomenon
occurring in curved coiled tubes. Also R0 number aims at simplify-
ing the existing correlations on curved coils to a great extent since
the number of variables involved in the formation of correlations
decrease considerably. An overall effect of increase in R0 number
in spiral coils is to increase the heat transfer but as the average
radius of the spiral coil increases the heat transfer decreases
because of decrease in centrifugal force. A possible explanation
to neglect the effect of thermal oscillations in the thermal entrance
region in the correlations proposed is given. The correlations
involving R0 numbers, developed for spiral coiled tubes on the basis
of experimental data are validated with the works of earlier
researchers and predicted results are found to be satisfactory.
The correlations thus can be used for the industrial design of spiral
coil heat exchangers. R0 number is an interesting dimensionless
number and application of this number to other curved coils
involving variable and constant curvature ratios will definitely
prove useful for future investigations.
References
[1] J.H. Grindley, A.H. Gibson, On the frictional resistance to the flow of air through
a pipe, Proc. Roy. Soc. A 80 (1908) 114–139.
[2] J. Eustice, Flow of water in curved pipes, Proc. Roy. Soc. A-84 (1910) 107–118.
[3] J. Eustice, Experiments of streamline motion in curved pipes, Proc. Roy. Soc. A-
85 (1911) 119–131.
[4] W.R. Dean, Note on the motion of fluid in a curved pipe, Philos. Mag. Ser. 4 (20)
(1927) 208–223.
[5] W.R. Dean, The stream-line motion of fluid in a curved pipe, Philos. Mag. Ser. 5
(1928) 673–695.
[6] E.N. Seider, G.E. Tate, Heat transfer and pressure drop of liquids in tubes, Ind.
Eng. Chem. 73 (1951) 803–809.
[7] L.R. Austin, J.D. Seader, Fully developed viscous flow in coiled pipes, Am. Inst.
Chem. Eng. 19 (1973) 85–94.
[8] V. Kubair, N.R. Kuloor, Heat transfer to Newtonian fluids in spiral coils at
constant tube wall temperature in laminar flow, Indian J. Technol. 3 (1965)
144–146.
155
[9] V. Kubair, N.R. Kuloor, Comparison of performance of helical and spiral coil
heat exchangers, Indian J. Technol. 4 (1965) 1–4.
[10] V. Kubair, N.R. Kuloor, Flow of Newtonian fluids in Archimedean spiral tube
coils: correlation of the laminar, transition and turbulent flows, Indian J.
Technol. 4 (1966) 3–8.
[11] V. Kubair, N.R. Kuloor, Heat transfer to Newtonian fluids in coiled pipes in
laminar flow, Int. J. Heat Mass Transfer 9 (1966) 63–75.
[12] Z. Zapryanov, C. Christov, E. Toshev, Fully developed laminar flow and heat
transfer in curved tubes, Int. J. Heat Mass Transfer 23 (1980) 873–880.
[13] R.K. Shah, S.D. Joshi, Convective heat transfer in curved ducts, in: S. Kakaç, R.K.
Shah, W. Aung (Eds.), A Handbook of Single Phase Convective Heat Transfer,
first ed., Wiley Interscience, Hoboken, NJ, 1987.
[14] W.G. Andrew, H.B. Williams, Applied Instrumentation in the Process
Industries, third ed., Gulf Publicating Company, Elsevier Publication,
Houston, Texas, 1993.
[15] D. Kern, Process Heat Transfer, first ed., Tata McGraw-Hill, NewDelhi, 1997
(Chapter 3).
[16] W.M. Rohsenow, J.P. Hartnett, Y.I. Cho, Handbook of Heat Transfer, third ed.,
McGraw Hill, New York, 1998 (Chapter 5).
[17] P. Naphon, S. Wongwises, Investigations of the performance of a spiral-coil
finned tube heat exchanger under dehumidifying conditions, J. Eng. Phys.
Thermophys. 76 (2003) 71–79.
[18] P. Naphon, S. Wongwises, A review of flow and heat transfer characteristics in
curved tubes, Renew. Sustain. Energy Rev. 10 (2004) 463–490.
[19] P. Naphon, J. Suwagrai, Effect of curvature ratios on the heat transfer and flow
developments in the horizontal spirally coiled tubes, Int. J. Heat Mass Transfer
08 (2006) 002.
[20] P. Naphon, S. Wongwises, An experimental study in the in-tube convective
heat transfer coefficients in a spiral coils heat exchanger, Int. Commun. Heat
Mass Transfer 29 (2002) 797–809.
[21] S. Kakaç, H. Liu, A. Pramuanjaroenkij, Heat Exchangers, third ed., CRC Press,
Boca Raton, 2012 (Chapter3).
[22] S.S. Pawar, V.K. Sunnapwar, Experimental studies on heat transfer to
Newtonian and non-Newtonian fluids in helical coils with laminar and
turbulent flow, Exp. Thermal Fluid Sci. 44 (2013) 792–804.
[23] P.H. Oosthuizen, D. Nayler, An Introduction to Convection Heat Transfer
Analysis, first ed., WCB/McGraw-Hill, Singapore, 1999 (Chapter 4).
[24] Geun-Jong Yoo, Hoon-Ki Choi, Wa-ryong Dong, Fluid flow and heat transfer
characteristics of spiral coiled tube: effects of Reynolds number and curvature
ratio, J. Cent, South Univ. 19 (2012) 471–476.
[25] Z. Altaҫ, Ӧ. Altun, Hydrodynamically and thermally developing laminar flow in
spiral coil tubes, Int. J. Therm. Sci. 77 (2014) 96–107.
[26] V. Kubair, N.R. Kuloor, Non-isothermal pressure drop data for spiral tube coils,
Indian J. Technol. 3 (1965) 382–383.
[27] B.A. Mujawar, M. Raja Rao, Flow of non-Newtonian fluids through helical coils,
Ind. Eng. Chem. Process Des. Dev. 17 (1978) 22–27.
[28] H. Ito, Friction factors for turbulent flow in curved pipes, ASME J. Basic Eng.
Trans. 81 (1959) 123–132.
[29] P.S. Srinivasan, S.S. Nandapurkar, F.A. Holland, Pressure drop and heat transfer
in coils, Chem. Eng. J. 218 (1968) CE113–CE119.
[30] P.S. Srinivasan, S.S. Nandapurkar, F.A. Holland, Friction factors for coils, Trans.
Inst. Chem. Eng. 48 (1970) TI56–TI61.
[31] V. Kubair, C.B.S. Varrier, Pressure drop for liquid flow in helical coils, Trans.
Indian Inst. Chem. Eng. 14 (1961-1962) 93–97.
[32] E.F. Schmidt, Warmeubarang and Druckverlust in Rohrshlangen, Chem. Eng.
Technol. 13 (1967) 781–789.
[33] M. Ghobadi, Y. Stephan Muzychka, A review of heat transfer and pressure drop
correlations for laminar flow in curved circular ducts, Heat Transfer Eng. 37
(10) (2016) 815–839.
[34] J.P. Holman, Experimental Methods for Engineers, fifth ed., McGraw Hill,
Singapore, 1989 (Chapter 3).
[35] S. Rainieri, F. Bozzoli, G. Pagliarini, Experimental investigation on the
convective heat transfer in straight and coiled corrugated tubes for highly
viscous fluids: preliminary results, Int. J. Heat Mass Transfer 55 (2012) 498–
504.
[36] R.A. Seban, E.F. Mclaughlin, Heat transfer in tube coils with laminar and
turbulent flow, Int. J. Heat Mass Transfer 6 (1963) 387–395.
[37] A.N. Dravid, K.A. Smith, E.W. Merrill, P.L.T. Brian, Effect of secondary fluid
motion on laminar flow heat transfer in helically coiled tubes, AIChE J. 17
(1971) 1114–1122.
[38] L.A.M. Janssen, C.J. Hoogendoorn, Laminar convection heat transfer in helical
coiled tubes, Int. J. Heat Mass Transfer 21 (1978) 1197–1206.
[39] J.M. Tarbell, M.R. Samuels, Momentum and heat transfer in helical coils, Chem.
Eng. J. 5 (1973) 117–127.
[40] M. Akiyama, K.C. Cheng, Laminar forced convection in the thermal entrance
region of curved pipes with uniform wall temperature, Can. J. Chem. Eng. 52
(1974) 234–240.