Experimental Thermal and Fluid Science: S.S. Pawar, Vivek K. Sunnapwar
Experimental Thermal and Fluid Science: S.S. Pawar, Vivek K. Sunnapwar
Experimental Thermal and Fluid Science: S.S. Pawar, Vivek K. Sunnapwar
a r t i c l e i n f o a b s t r a c t
Article history: Experimental studies on isothermal steady state and non-isothermal unsteady state conditions were car-
Received 17 July 2012 ried out in helical coils for Newtonian and non-Newtonian fluids. Water, glycerol–water mixture as New-
Received in revised form 20 September 2012 tonian fluids and dilute aqueous polymer solutions of sodium carboxymethyl cellulose (SCMC), sodium
Accepted 23 September 2012
alginate (SA) as non-Newtonian fluids were used in this study. These experiments were performed for
Available online 4 October 2012
coils with curvature ratios d = 0.0757, 0.064 and 0.055 in laminar and turbulent regimes. An innovative
approach of correlating Nusselt number to dimensionless number ‘M’ for Newtonian fluids based on
Keywords:
experimental data is presented which is not available in the literature. Several correlations for the first
Helical coil
Heat transfer
time are proposed based on heat transfer data generated from the experiments performed for Newtonian
Newtonian fluid fluids under isothermal and non-isothermal conditions (total 138 tests). These developed correlations
Non-Newtonian fluid were compared with the work of earlier investigators and were found in good agreement. Further, com-
parison of overall heat transfer coefficient Uo, and Nusselt numbers for Newtonian and non-Newtonian
fluids under isothermal and non-isothermal conditions (total 276 tests) is presented in this paper. From
the experimental results, it was found that overall heat transfer coefficient and Nusselt numbers for
water are higher than glycerol–water mixture and non-Newtonian fluids. It was also observed from
experimental results that as helix diameter increases, overall heat transfer coefficient and Nusselt num-
bers of both fluids decreases for the same flow rates.
Ó 2012 Elsevier Inc. All rights reserved.
0894-1777/$ - see front matter Ó 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.expthermflusci.2012.09.024
S.S. Pawar, V.K. Sunnapwar / Experimental Thermal and Fluid Science 44 (2013) 792–804 793
Nomenclature
Table 1
Various correlations developed by earlier investigators under different experimental conditions.
Table 2
Comparison of critical Reynolds number in helical coils (values of d are assumed).
Correlations developed for calculating critical Reynolds number inserting thermocouple probe in the wall, it was tightly covered by
by [24–29] are applicable for Newtonian fluids only whereas Eq. stainless steel strip so that it could not leave the contact of the wall
(1) developed by Mujawar and Rao [7] in their research article to surface. Then it was tightly covered by five turns of thick Teflon
calculate M number are applicable to Newtonian as well as non- tape which has low thermal conductivity, to avoid penetration of
Newtonian fluids, is significant. Therefore, it is essential to develop water in the junction. There was no evidence of opening of Teflon
Nusselt number correlations in terms of M number and tests the tape even at 80 °C temperature of water in vessel and it was for
same with published correlations. sure that the thermocouple junctions were properly secured with
these tapes even for repeated readings. This is because as there is
no any disturbance of fluid coming in contact (like fluid movement
3. Materials and methods
due to agitation, etc.) with the thermocouple junctions, as heat
transfer from outside hot water to wall takes place by natural con-
3.1. Experimental setup
vection. Hence, there is no possibility of penetration of water into
the thermocouple junction which will disturb the thermocouple
A schematic diagram of the experimental setup is as shown in
readings. Thus, sufficient care was taken while attaching or fixing
Fig. 1. The physical dimensions of helical coils used in the present
thermocouples on the surface of tube wall to reduce uncertainties
investigation are given in Table 3.
in the wall temperature measurements.
The coil length is calculated by using Lc = pDN. Inside helical
Five thermocouples as shown in Fig. 1 and one thermometer
coil surface is plain and smooth, and therefore roughness has al-
were used to measure hot water vessel temperature. Digital tem-
most insignificant effect. It will not have much bearing on heat
perature indicator was used to record all these thermocouple tem-
transfer phenomenon. Few experiments were repeated to check
peratures whose least count is 0.1 °C and was calibrated to meet
the reproducibility of results and it was found to be excellent.
limits of error ±0.1 °C.
An insulated hot water vessel of size 550 mm diame-
ter 700 mm height with 4 mm thickness of mild steel material
3.2. Experimental procedure
of construction was used to house the coils. Outside surface of
the entire hot water vessel was provided with insulation of
The test fluids were pumped through the coils using 1 HP cen-
50 mm thickness of thermocol covered by asbestos material. Insu-
trifugal pump (Single phase, 2800 RPM, Head-30 m, Discharge-
lation to the helical coil to the downside of the test section was also
3600 LPH, Laxmi Make) at different inlet temperatures starting
provided for safety while taking readings. The radiation loss at the
from 32 ± 2 °C and it went on increasing maximum up to
exit thermometer is estimated to be less than 0.1% of the total area
42 ± 2 °C for last reading. This happens due to change in helix
of the test section (helical coil). At inlet, there is obviously negligi-
diameter of the helical cooling coil in cold water tank. The test fluid
ble heat loss by radiation as cold water is entering in the coil.
was taken from a solution tank of size 500 500 500 mm3
Hence, there will not be any practical significance on the readings
which is made of mild steel material. All the test fluids were heated
taken by thermometer at inlet and outlet. The water in the vessel
in hot water vessel while passing through helical coil. The test flu-
outside the coil is in a natural convection heat transfer mode and
ids were cooled in helical coil placed in cold water tank of size
it does not have any significant stratified flow pattern. Hence, the
700 600 1500 mm3 and then recirculated through solution
effect of stratification in the water vessel is ignored.
tank as shown in Fig. 1.
Three electrical heaters were used in hot water vessel out of
The constant cold water (at 30 ± 1 °C) flow rate –
which two heaters of total power 8 KW (5 KW and 3 KW) fixed at
32.92 105 m3/s was supplied while entering in cooling tank to
the bottom were switched on for all the time and third heater of
cool the test fluids in helical coil and same flow rate of water
power 3 KW fixed at side was switched on and off as and when
was leaving cooling tank at different temperatures. The flow rate
needed to maintain a constant hot water vessel temperature
of test fluid was measured by two rota meters (Capacity 0 to
62.5 ± 0.2 °C. For non-isothermal unsteady state condition, all the
10 lpm and 10–100 lpm, Star Make, Sr. No. 238). It was verified
three heaters were switched on continuously until maximum possi-
for each experiment by noting the time taken to fill a bucket of
ble temperature of vessel was attained for maximum possible flow
known volume located at the exit of the system to reduce uncer-
rate of corresponding test fluids. For each experiment, the required
tainty in the flow measurements. The purpose of using different
coil was mounted on supporting platform made of angle
flow rates was to change the temperature distribution along the
(25 25 mm) and was fixed to the hot water vessel. The threaded
tube wall and to change the inside heat transfer coefficient.
connections for inlet and outlet of coil ends as shown in Fig. 1 were
Inlet and outlet temperatures of heating and cooling coils were
used to connect suction and discharge pipes. The inner diameter of
measured by mercury thermometers having a range of 10 to
the fittings was kept equal to the inner diameter of the test coil (coil
150 °C and were calibrated to meet limits of error ±0.1 °C. The sys-
wall thickness = 2.3 mm) to prevent disturbance of the flow pattern
tem was allowed to come to a steady state condition (approximate
of the fluid when entering and leaving the coils. The first and the last,
time taken for each flow rate to come to steady state condition was
two turns of the helical coils were provided respectively as upstream
15–20 min for isothermal steady state conditions) before the tem-
and downstream calming sections in accordance with the equation
peratures were noted. For non-isothermal unsteady state condi-
of Austin and Seader [30] to get fully developed flow given by:
tion, the first reading was taken when vessel temperature
reached at 54 ± 0.2 °C and then remaining all readings were taken
cD ¼ 49½Deða=RÞ1=3 ð3Þ
for every 15 min of time interval. For each test fluid, the experi-
Temperature measurements were made using type K (nickel– ment was repeated for few readings to minimize the uncertainty
chromium) thermocouples of 14-gauge extension wire, tempera- in measured experimental parameters and reproducibility were
ture range 50 to 300 °C, (Digital System and Automation Make, found to be within (an average) ±1.57% for temperatures, ±5.6%
SE-221) and were calibrated to meet limits of error ±0.1 °C. Six for flow rates of test fluids.
thermocouples were attached to measure the wall temperatures
of the coils at six different locations. These six thermocouples were 3.3. Preparation of non-Newtonian solutions
attached opposite to each alternate coil turn opposite to inner and
outer side of coil. Holes up to 0.5 mm depth were drilled in the wall Non-Newtonian test solutions which were prepared in water by
to the size of thermocouple probe to be inserted into the wall. After dissolving a known amount of powders SCMC and Sodium Alginate
796 S.S. Pawar, V.K. Sunnapwar / Experimental Thermal and Fluid Science 44 (2013) 792–804
Table 3
Physical dimensions of helical coils used.
Table 4
Types of coils, test fluids, flow rates and experimental conditions used in this experimental work.
K t ¼ 0:00127 þ 0:00000317ðtÞ for 10% glycerol ð4Þ DTo is the temperature difference between the vessel temperature
and an average wall temperature of the coil.
K t ¼ 0:00118 þ 0:00000284ðtÞ for 20% glycerol ð5Þ
4. Results and discussion
t is the temperature of mixture. The thermo-physical properties of
mixtures used in this experimental work are given in Table 5.
Results and discussion is divided into two main sections such
Heat loss from the hot water in the vessel to the cold water
as: (1) results and discussion for Newtonian fluids and (2) heat
flowing in the coil is calculated from:
transfer results and discussion for non-Newtonian fluids.
_ p ðT out T in Þ
Q ¼ mC ð6Þ
4.1. Results and discussion for Newtonian fluids
All quantities were known except the heat transfer rate. The
heat transfer rate calculated from above equation was then used 4.1.1. Development of correlations based on experimental data
to calculate the overall heat transfer coefficient, Uo as: Total 138 experiments were conducted for Newtonian fluids
under isothermal steady state and non-isothermal unsteady state
Table 5
Thermo-physical properties of the mixtures at mean bulk temperature of test fluids.
Type of mixture Tbulk (°C) k (W/mK) cp (J/kg k) q (kg/m3) m (m2/s) Experimental conditions
10% Glycerol 47 0.5499 4010 1021 7.49e7 Isothermal
20% Glycerol 46.77 0.549 3838 1051 9.94e7 Isothermal
50 0.5526 3845.6 1049.8 9.42e7 Non-isothermal
0.5% and 1% SCMC, 47 0.6368 4177 989 – Isothermal
50 0.6474 4178 988 – Non-isothermal
0.5% and 1% S.A 47 0.6376 4175 989 – Isothermal
50 0.6474 4178 988 – Non-isothermal
798 S.S. Pawar, V.K. Sunnapwar / Experimental Thermal and Fluid Science 44 (2013) 792–804
conditions with laminar to turbulent flow regimes as given in Ta- The least-squares power-law fit on experimental data points
ble 4. The desired form of the correlation is as follows: yields the following correlation where the correlation coefficient
R = 94.94%.
0
Nu ¼ a0 ðMÞb ð11Þ
Nu ¼ 0:0023ðMÞ1:2354 Pr0:4 ð13Þ
0 0
where a and b are constants to be evaluated experimentally by
applicable for the range of: 907 6 M 6 3348, 3.83 6 Pr 6 7.3 and
using least-squares method of curve fitting based on experimental
0.055 6 d 6 0.0757. Exponent of Prandtl number is assumed 0.4
data. The experimental data points of three coils (total 90 tests)
for heating process which is also used by most of the earlier inves-
for water and glycerol–water mixture (10% and 20% glycerol in
tigators. Both Eqs. (12) and (13) are developed for isothermal steady
water) under isothermal steady state temperature condition in ves-
state temperature condition in heating vessel whereas Eq. (14) is
sel are presented in Fig. 2, by plotting Nu against M on a logarithmic
developed for combination of isothermal steady state and non-iso-
scale that results the following correlation; where the correlation
thermal unsteady state conditions. Total 138 (90 tests for isother-
coefficient is R = 92.83%.
mal steady state and 48 tests for non-isothermal unsteady state)
experimental data points are presented in Fig. 4, by plotting Nu
Nu ¼ 0:0065ðMÞ1:1798 ð12Þ against M on a logarithmic scale which is resulted into following
correlation; where the correlation coefficient is R = 89.43%.
applicable for the range of: 907 6 M 6 3348, 3.83 6 Pr 6 7.3 and
0.055 6 d 6 0.0757. Nu ¼ 0:0096ðMÞ1:1202 ð14Þ
The solid line presents the fitting curve through experimental
data points using a least-squares power law fit. This correlation applicable for the range of: 880 6 M 6 4642, 3.54 6 Pr 6 7.3 and
is proposed without considering the effect of Prandtl number as 0.055 6 d 6 0.0757.
a heat transfer parameter which is important when test fluid is Here value of correlation coefficient R is lower as compared to
highly viscous. Hence, Nusselt number is correlated for the same Eqs. (12) and (13) due to lower values of Nusselt numbers for
condition and same test data to M number and Prandtl number non-isothermal unsteady state condition. Fig. 5 presents the plot
by plotting M on x-axis and Nu/Pr0.4 on y-axis using logarithmic of Nu/Pr0.4 vs. Gz for experimental data points obtained from 78
scale as shown in Fig. 3. tests of water and 10% glycerol–water mixture under isothermal
Fig. 2. Plot of Nu vs. M for 90 tests data obtained from water and glycerol–water mixture under isothermal condition to develop correlation (12).
Fig. 3. Plot of Nu/Pr0.4 vs. M for 90 tests data obtained from water and glycerol–water mixture under isothermal condition to develop correlation (13).
S.S. Pawar, V.K. Sunnapwar / Experimental Thermal and Fluid Science 44 (2013) 792–804 799
Fig. 4. Plot of Nu vs. M for 138 tests data obtained for water and glycerol–water mixture under isothermal and non-isothermal conditions to develop correlation (14).
Fig. 5. Plot of Nu/Pr0.4 vs. Gz for 78 tests data obtained for water and 10% glycerol–water mixture under isothermal condition to develop correlation (15).
condition to get Eq. (15) by least-squares power-law fit on log–log zero, the pipe becomes horizontal and when a/R tends to infinity,
scale as: the pipe becomes vertical. Hence, based on the nature of the corre-
lations available in the present literature, it is proposed to develop a
Nu ¼ 0:9366ðGzÞ0:8177 Pr0:4 ð15Þ correlation based on experimental data of three coils for Newtonian
applicable for the range of: 3.83 6 Pr 6 5.149, 0.055 6 d 6 0.0757 fluids. Experimental data points of all coils (90 tests) are presented
and 19 6 Gz 6 152. Coiled tubes are usually characterized by Dean in Fig. 6, by plotting Nu/Pr0.4 against De on a logarithmic scale. The
number. Fluid flow and heat transfer data analysis is done in liter- solid line presents the fitting curve through the experimental data
ature using Dean number which is: De = Re (a/R)1/2. The physical points using a least-squares power law fit that results the following
significance of Reynolds number is inertia force by viscous force correlation; where the correlation coefficient is R = 98.15%.
and clearly gives the idea of fluid flowing through straight pipe Nu ¼ 0:0472De0:8346 Pr0:4 ð16Þ
about its flow pattern such as laminar, transition and turbulent.
However, Dean number does not give any idea about the flow pat- valid for the range of: 586 6 De 6 4773, 0.055 6 d 6 0.0757, and
tern of the fluid flowing through helical coils. When a/R tends to 3.83 6 Pr 6 7.3.
800 S.S. Pawar, V.K. Sunnapwar / Experimental Thermal and Fluid Science 44 (2013) 792–804
Fig. 6. Plot of Nu/Pr0.4 vs. De for 90 tests data obtained for water and 10%, 20% glycerol–water mixture under isothermal condition to develop correlation (16).
Fig. 7. Plot of Nu vs. M for 48 tests data obtained for water and 20% glycerol–water mixture under non-isothermal condition to develop Eq. (17).
Fig. 8. Plot of Nu vs. Re for comparison of developed correlations (Eqs. (12)–(17)) based on present experimental data (for Re = 3498–17347, Pr = 3.83 and d = 0.0757) with
works of earlier researchers for coil-I.
S.S. Pawar, V.K. Sunnapwar / Experimental Thermal and Fluid Science 44 (2013) 792–804 801
A correlation for Newtonian fluids under non-isothermal un- Jayakumar et al. [10] had the prediction of 0.32%, 7.52%, 6.81%,
steady state conditions, Eq. (17) is developed by using least- 3.18%, 3.15%, and 16.27% for the same. It can be noted that all
squares power-law fit on log–log scale as shown in Fig. 7 for 48 these predicted values stand higher than the corresponding values
tests of water and 20% glycerol which is resulted as: calculated by developed Eqs. (12)–(17). Developed correlations
from Eqs. (12)–(16) based on experimental data under isothermal
Nu ¼ 0:0062ðMÞ1:1631 ð17Þ steady state condition shows acceptable range of deviations within
±10%. Eq. (17) which is developed under non-isothermal unsteady
applicable for the range: 880 6 M 6 4642. Comparison of this corre-
state condition gives little higher deviation as compared to Eqs.
lation is shown in Fig. 8. It predicts lowest values of Nusselt num-
(12)–(16). Hence, all these correlations are recommended for the
bers as compared to developed correlations (12)–(16) and with
design of industrial helical coil heat exchangers for Newtonian
the works of earlier researchers. All the developed correlations
fluids.
(12)–(17) are having reliability margins within ±10%.
The correlations developed from Eqs. (12)–(16) based on vari-
ous experimental data which are generated for Newtonian fluids 4.1.2. Comparison of overall heat transfer coefficient
covering Reynolds number range from 2227 to 17,347 under iso- Overall heat transfer coefficient of three helical coils for Newto-
thermal steady state condition and from 2351 to 28,904 for 48 nian fluids under isothermal steady state and non-isothermal un-
tests under non-isothermal unsteady state condition. Fig. 8 pre- steady state conditions is shown in Fig. 9. It is observed from
sents a comparison of the Nusselt numbers predicted by Eqs. Fig. 9 that as Reynolds number increases, overall heat transfer coef-
(12)–(17) with (i) Xin and Ebadian [9] and (ii) Jayakumar et al. ficient increases for water as well glycerol–water mixture but it de-
[10]. It can be seen from Fig. 8 that present correlations devel- creases as helix diameter of coil increases. From Eq. (7), overall
oped are in good agreement with Nusselt numbers predicted by heat transfer coefficient is directly proportional to heat transfer
correlations by [9,10]. It is also observed from Fig. 8 that, the rate, Q and inversely proportional to outer surface area of coil, Ao
present correlation (16) gives almost same values of Nusselt and DTo. Though there is increase in heat transfer rate (due to more
numbers with correlation developed by Jayakumar et al. [10] for residual time for temperature penetration from hot fluid in vessel
all the range of Reynolds number. However, there is a marginal to cold fluid flowing in coil for higher length of coil) as coil length
difference of Nusselt numbers in turbulent regime as compared increases but this is dominated by larger outer surface area of coil
to laminar regime for all the correlations except Eq. (16) and which decreases value of Uo for higher helix diameter. For coil-I,
Jayakumar et al. [10]. In this experimental work, critical Reynolds outer surface area is 0.65396 m2 and for coil-III, it is
number is calculated as per criteria given by Mujawar and Rao 0.89212 m2. Value of DTo also increases as helix diameter in-
[7]. Similarly, comparisons for other coils and conditions have creases. Value of outside heat transfer coefficient is inversely pro-
been done (not shown here). The range of M number of Newto- portional to the length of coil. Also, secondary motion becomes
nian fluids, for all experiments (138 tests) lies in between 850 weaker for higher helix diameter (which decreases value of inner
and 4642 which correspond to Reynolds number 2227 to heat transfer coefficient, hi) due to decrease in centrifugal force act-
28,904. Measurements up to M = 2100 correspond to laminar flow ing on moving fluid element inside the coil. Due to all these rea-
regimes and, M > 2100 or Re > 10,000 as per Ito [24,27] measure- sons, overall heat transfer coefficient decreases as helix diameter
ments correspond to turbulent flow as seen from earlier correla- increases. Also, it is observed from results that, for coil-I, an aver-
tions developed by investigators. age value of Uo of water is found to be 15.51% higher than 10% glyc-
The values of Nusselt numbers predicted by Xin and Ebadian [9] erol–water mixture and 16.73% for 20% glycerol–water mixture
are 9.18%, 15.74%, 15.55%, 14.4%, 6.01%, and 23.71%. However, under isothermal steady state condition.
Fig. 9. Comparison of overall heat transfer coefficients for Newtonian fluids as a function of Reynolds number under isothermal and non-isothermal conditions (86 tests
data).
802 S.S. Pawar, V.K. Sunnapwar / Experimental Thermal and Fluid Science 44 (2013) 792–804
Fig. 10. Comparison of experimental Nusselt numbers for coil-I with straight pipe using Pigford’s Eq. (18).
Fig. 11. Comparison of Nusselt numbers vs. Gz numbers for non-Newtonian fluids under isothermal and non-isothermal conditions with coil-I.
4.1.3. Comparison of Nusselt number Nusselt numbers for 0.5% and 1% sodium alginate are higher than
Comparison of Nusselt numbers for coil-I with Pigford’s correla- SCMC solutions having same content in water under isothermal
tion [33] applicable for Newtonian fluids for straight pipes given by and non-isothermal conditions. Also, increase in percentage of
Eq. (18) is presented in Fig. 10. non-Newtonian powders in water decreases heat transfer coeffi-
cients as viscosity of mixture increases. In isothermal condition,
Nu ¼ 1:75½ð3n þ 1Þ=4n1=3 Gz1=3 ð18Þ heat transfer coefficients are higher than non-isothermal condition
for the same type of fluids because of temperature difference at
For water, n = 1.
starting of an experiment. Similar plots of Nu vs. Gz are plotted
Enhancement in Nusselt numbers for helical coil over straight
for other two coils, d = 0.064 and d = 0.055 (not shown here). From
pipe by using Eq. (18) is found to be 30.84%. Similarly, comparison
all these plots of different coil curvature ratios for non-Newtonian
for other coils and fluids are done (not shown here).
fluids, it is observed that the heat transfer coefficient decreases as
helix diameter increases.
4.2. Heat transfer results and discussion for non-Newtonian fluids Fig. 12 presents comparison of overall heat transfer coefficient
of three coils for non-Newtonian fluids under isothermal steady
For non-Newtonian fluids, Graetz number is the correct charac- state as well as non-isothermal unsteady state conditions. Overall
teristic dimensionless group to correlate the heat transfer coeffi- heat transfer coefficient increases for all non-Newtonian fluids
cients (Rajasekharan et al. [23]). From Fig. 11, it can be seen that shown in Fig. 12 as Graetz number increases. From experimental
S.S. Pawar, V.K. Sunnapwar / Experimental Thermal and Fluid Science 44 (2013) 792–804 803
Fig. 12. Comparison of overall heat transfer coefficients for non-Newtonian fluids as a function of Gz number under isothermal and non-isothermal conditions (120 tests).
results, it is observed that overall heat transfer coefficient of non- Hence, these developed correlations are recommended in the de-
Newtonian fluids is lower than Newtonian fluids for same condi- sign of industrial helical coil heat exchangers for said conditions.
tions. This is due to higher viscosities of non-Newtonian fluids than M number is found to be significant to characterize the hydrody-
Newtonian fluids. Higher viscosity fluid forms thicker thermal namics of flow in helical coil for Newtonian fluids for any coil cur-
boundary layer. This increases thermal resistance which reduces vature ratios and heat transfer phenomena. The results in this work
inner heat transfer coefficient and subsequently overall heat trans- showed that overall heat transfer coefficient is higher for smaller
fer coefficient. An average overall heat transfer coefficient for So- helix diameter as compared to larger helix diameter due to signif-
dium Alginate solution was found to be higher than SCMC icant effect of centrifugal force on secondary flow in coil. It is also
solution for same experimental conditions. observed from results that heat transfer coefficients for pure water
The determination of overall heat transfer coefficient is an are higher than glycerol–water mixture and non-Newtonian fluids
important step to calculate surface area required for process heat for the same conditions. In future work, Nusselt number need to be
transfer in the design of helical heat exchangers. In future analysis, correlated to M number based on experimental data for any type of
Nusselt number need to be correlated to M number for non-New- power-law fluids and any coil curvature ratios, and shall be vali-
tonian fluids based on experimental data and shall be validated dated with the work of earlier investigators.
with the published work.
Acknowledgements
4.3. Uncertainty analysis of the results
The authors would like to express their cordial thanks to Dr. B.A.
Experimental uncertainty was calculated by using Eq. (3.2) gi-
Mujawar for introducing the problem and his excellent help during
ven in Holman [34]. Probable uncertainties involved in the mea-
experimentation and data analysis. Authors also acknowledge Uni-
surement of various parameters are taken into consideration. To
versity of Mumbai and the Management of Lokmanya Tilak College
verify the repeatability of the experiments, few runs for each test
of Engineering, Navi-Mumbai for the financial support to this re-
fluids were repeated which yielded excellent results. The calcula-
search work. Further, the authors would like to gratefully acknowl-
tions made showed that the uncertainties taking part in the mea-
edge assistance provided by Dr. V.K. Yakkundi in editing and proof
surements were approximately ±(3–5)% for heat transfer rate,
reading.
overall heat transfer coefficient, volumetric flow rate and ±3% for
Nusselt number respectively. Hence, it will be insignificant effect
on the correlations deduced from the experimental data. References
[1] P.C. Mukesh Kumar, J. Kumar, S. Suresh, Heat transfer and friction factor
5. Conclusion studies in helically coiled tubes using Al2O3/water nanofluid, European Journal
of Scientific Research 82 (2012) 161–172.
An experimental setup is fabricated to study fluid to fluid heat [2] S.K. Das, N. Putra, P. Thiesen, W. Roetzel, Temperature dependence of thermal
conductivity enhancement for nanofluids, ASME Transactions, Journal of Heat
transfer in a helically coiled heat exchanger in laminar to turbulent Transfer 125 (2003).
regime under isothermal steady state and non-isothermal unstea- [3] H. Jeschke, Warmeubergang un Druckverlust in Rohrschlagen, VDI Zeitschrift
dy state conditions for Newtonian as well as non-Newtonian fluids. VDI 69 (1925) 24–28.
[4] R.A. Seban, E.F. McLaughlin, Heat transfer in tube coils with laminar and
For the first time, an innovative approach of correlating Nusselt
turbulent flow, International Journal of Heat and Mass Transfer 6 (1963) 387–
number to dimensionless number ‘M’ for Newtonian fluids based 395.
on experimental data is presented which is not available in the lit- [5] G.F.C. Rogers, Y.R. Mayhew, Heat transfer and pressure loss in helically coiled
erature. Correlations developed based on experimental data (Eqs. tubes with turbulent flow, International Journal of Heat and Mass Transfer 7
(1964) 1207–1216.
(12)–(17)) are compared with work of earlier researchers and the [6] Y. Mori, W. Nakayama, Study on forced convective heat transfer in curved
results predicted by these equations are found to be satisfactory. tubes, (3rd report, theoretical analysis under the conditions of uniform wall
804 S.S. Pawar, V.K. Sunnapwar / Experimental Thermal and Fluid Science 44 (2013) 792–804
temperature and practical formulae), International Journal of Heat and Mass [20] A.B. Metzner, J.C. Reed, Flow of non-Newtonian fluids—Correlations of laminar,
Transfer 10 (1967) 681–695. transition and turbulent flow regions, A.I.Ch.E. Journal 1 (1955) 434–440.
[7] B.A. Mujawar, M. Raja Rao, Flow of non-Newtonian fluids through helical coils, [21] B.K. Rao, Turbulent heat transfer to power-law fluids in helical passages,
Industrial and Engineering Chemistry Process Design and Development 17 International Journal of Heat and Fluid Flow 15 (1994) 142–148.
(1978) 22–27. [22] T.J. Rennie, G.S.V. Raghavan, Thermally dependent viscosity and non-
[8] S.S. Pawar, V.K. Sunnapwar, B.A. Mujawar, Experimental determination of heat Newtonian flow in a double-pipe helical heat exchanger, Applied Thermal
transfer coefficient of turbulent flow in helical coil with circular c/s and its Engineering 27 (2007) 862–868.
comparison with rectangular c/s, Journal of Cooperation Among University, [23] S. Rajasekharan, V.G. Kubair, N.R. Kuloor, Heat transfer to non-Newtonian
Research and Industrial Enterprises 3 (2010) 13–22. fluids in coiled pipes in laminar flow, International Journal of Heat Mass
[9] R.C. Xin, M.A. Ebadian, The effects of Prandtl numbers on local and average Transfer 13 (1970) 1583–1594.
convective heat transfer characteristics in helical pipes, Journal of Heat [24] H. Ito, Friction factors for turbulent flow in curved pipes, Journal of Basic
Transfer 119 (1997) 467–474. Engineering Transactions of the ASME 18 (1959) 123–134.
[10] J.S. Jayakumar, S.M. Mahajani, J.C. Mandal, P.K. Vijayan, R. Bhoi, Experimental [25] P.S. Srinivasan, S.S. Nandapurkar, F.A. Holland, Friction factor for coils,
and CFD estimation of heat transfer in helically coiled heat exchangers, Transactions of the Institution of Chemical Engineers 48 (1970) T156–T161.
Chemical Engineering Research and Design 86 (2008) 221–232. [26] A. Cioncolini, L. Santini, An experimental investigation regarding the laminar
[11] S.A. Berger, L. Talbot, L.S. Yao, Flow in curved pipes, Annual Review Fluid to turbulent flow in helically coiled pipes, Experimental Thermal and Fluid
Mechanics 15 (1983) 461–512. Science 30 (2006) 367–380.
[12] P. Naphon, S. Wongwises, A review of flow and heat transfer characteristics in [27] H. Ito, Flow in curved pipes, JSME International Journal 30 (1987) 543–552.
curved tubes, Renewable Sustain Energy Review 10 (2006) 463–490. [28] E.F. Schmidt, Warmeubergang and Druckverlust in Rohrschbugen, Chemical
[13] S.S. Pawar, V.K. Sunnapwar, B.A. Mujawar, A critical review of heat transfer Engineering Technology 13 (1967) 781–789.
through helical coils of circular cross section, Journal of Scientific and [29] V. Kubair, N.R. Kuloor, Flow of Newtonian fluids in Archimedian spiral tube
Industrial Research 70 (2011) 835–843. coils: correlation of the laminar, transition and turbulent flows, Indian Journal
[14] N.S. Cheng, Formula for viscosity of glycerol–water mixture, Industrial and of Technology 4 (1966) 3–8.
Engineering Chemistry Research 47 (2008) 3285–3288. [30] L.R. Austin, J.D. Seader, AICHE Journal 20 (4) (1974) 820.
[15] J.B. Segur, H.E. Oberstar, Viscosity of glycerol and its aqueous solutions, [31] CINDAS, Properties of inorganic and organic fluids, in: C.Y. Ho (Ed.), CINDAS
Industrial and Engineering Chemistry 43 (1951) 2117–2120. data series on material properties, Hemisphere publishing corporation, New
[16] P.N. Shankar, M. Kumar, Experimental-determination of the kinematic York, London, vol. V-1, 1988.
viscosity of glycerol water mixtures, Proceedings of the Royal Society of [32] O.K. Bates, Industrial and Engineering Chemistry 28 (1936) 494–498.
London Series a-Mathematical Physical and Engineering Sciences 444 (1994) [33] R.L. Pigford, Non-isothermal flow and heat transfer inside vertical tubes,
573–581. Chemical Engineering Progress Symposium Series 51 (1955) 79–92.
[17] F. Chenlo, R. Moreira, G. Pereira, B. Bello, Kinematic viscosity and water activity [34] J.P. Holman, Experimental Methods for Engineers, fifth ed., McGraw Hill,
of aqueous solutions of glycerol and sodium chloride, European Food Research Singapore, 1989 (Chapter 3).
and Technology 219 (2004) 403–408. [35] C. Pablo, K.P. Sandeep, Heat transfer coefficients in helical heat exchangers
[18] J.R. Jones, Flow of a non-Newtonian fluid in a curved pipe, Quarterly Journal of under turbulent flow conditions, International Journal of Food Engineering 4
Mechanics and Applied Mathematics 13 (1960) 428–443. (2008) 1–12.
[19] S. Rajasekharan, V. Kubair, N.R. Kuloor, Secondary flow of non-Newtonian
fluids in helical coils, Indian Journal of Technology 4 (1966) 33–35.