Experimental Thermal and Fluid Science 31 (2007) 301308 www.elsevier.

com/locate/etfs

Experimental studies on heat transfer and friction factor characteristics of laminar ow through a circular tube tted with regularly spaced helical screw-tape inserts
P. Sivashanmugam *, S. Suresh
Department of Chemical Engineering, National Institute of Technology, Tiruchirappalli 620 015, Tamil Nadu, India Received 15 February 2006; received in revised form 1 May 2006; accepted 1 May 2006

Abstract Experimental investigation of heat transfer and friction factor characteristics of circular tube tted with full-length helical screw element of dierent twist ratio, and helical screw inserts with spacer length 100, 200, 300 and 400 mm have been studied with uniform heat ux under laminar ow condition. The experimental data obtained are veried with those obtained from plain tube published data. The eect of spacer length on heat transfer augmentation and friction factor, and the eect of twist ratio on heat transfer augmentation and friction factor have been presented separately. The decrease in Nusselt number for the helical twist with spacer length is within 10% for each subsequent 100 mm increase in spacer length. The decrease in friction factor is nearly two times lower than the full length helical twist at low Reynolds number, and four times lower than the full length helical twist at high Reynolds number for all twist ratio. The regularly spaced helical screw inserts can safely be used for heat transfer augmentation without much increase in pressure drop than full length helical screw inserts. 2006 Elsevier Inc. All rights reserved.
Keywords: Augmentation; Laminar ow; Regularly spaced helical screw inserts; Twist ratio; Heat transfer

1. Introduction The technique of improving the performance of heat transfer system is referred to as heat transfer augmentation or intensication. This leads to reduce the size and cost of the heat exchanger. Heat transfer enhancement technology has been developed and widely applied to heat exchanger applications; for example, refrigeration, automotives, process industry, chemical industry etc. Many techniques of active and passive techniques are available for augmentation. These details are discussed in detail by Bergles. Also heat augmentation techniques play a vital role for laminar ow, since the heat transfer coecient is generally low in plain tubes. Bergles [1,2] presented a comprehensive survey on heat transfer enhancement by various techniques.
*

Corresponding author. E-mail address: psiva@nitt.edu (P. Sivashanmugam).

Among many techniques (both passive and active) investigated for augmentation of heat transfer rates inside circular tubes, tube tted with full length twisted tape inserts (also called as swirl ow device) has been shown to be very eective, due to imparting of helical path to the ow. Vast literature is available on heat transfer augmentation studies in tube tted with twisted tape in laminar ow. Helical screw-tape swirl ow generators shown in Fig. 1 is a modied form of a twisted tape wound on a single rod gives single way smooth direction of ow like screw motion. In the previous paper (Sivashanmugam and Suresh) heat transfer and friction factor characteristics of laminar ow through a circular tube tted with helical screw-tape inserts has been reported. The present paper reports the heat transfer and friction factor characteristics of turbulent ow through a circular tube tted with full length helical screw inserts (twist) of various twist ratio, and regularly spaced helical screw inserts with dierent spacer length.

0894-1777/$ - see front matter 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.expthermusci.2006.05.005

302

P. Sivashanmugam, S. Suresh / Experimental Thermal and Fluid Science 31 (2007) 301308

Nomenclature AI Ao Cp Di Do f hi k kw L Nu Q Pr R Re inside surface area of test section area, m2 outside surface area of test section area, m2 specic heat at constant pressure, KJ/kg C inside diameter of test section, mm outside diameter of test section, mm friction factor, dimensionless average convective heat transfer coecient, W/m2 C thermal conductivity of uid, W/m C thermal conductivity of the tube wall, W/m C length of the test section, m Nusselt number, dimensionless Nu = hiDi/kw heat transfer rate, W Prandtl number dimensionless Pr = Cpl/k resistance of the heating element, ohm (X) Reynolds number based on internal diameter of the tube, dimensionless Tf Tin Tout Two um Ui Uo V Y average of uid temperature in the test section, C inlet bulk temperature of uid, C outlet bulk temperature of uid, C average wall surface temperature outside test section, C bulk average uid velocity, m/s over all inside heat transfer coecient, W/m2 C over all heat transfer coecient, W/m2 C voltage output from the Auto-transformer, V twist ratio (length of one twist/ diameter of the twist), dimensionless

Greek letters q density of uid, kg/m3 l viscosity of uid, N s/m2 DP pressure drop of uid, N/m2

2. Technical details of helical screw-tape inserts The geometrical conguration of helical screw-tape inserts is shown in Fig. 1. The helical screw-tape inserts with various twist ratio is made by winding uniformly a strip of 8.5 mm width over a 8 mm rod, and coated with chromium by electroplating to prevent corrosion. The twist ratio Y dened as the ratio of length of one twist to diameter of the twist is varied from 1.95 to 4.89. This range was selected with reference to very widely reported literature [4,5] value between 2 and 7 for twisted tape inserts. The helical screw twist inserts are designated as A, B, C, and D for the twist ratio values of 1.95, 2.93, 3.91, and 4.89 respectively as shown in Fig. 1. The helical screw inserts with the spacer length 100 mm, 200 mm, 300 mm, and 400 mm are designated as A1, A2, A3, and A4, B1, B2, B3, and B4, C1, C2, C3, and C4, and D1, D2, D3, and D4 respectively for Y = 1.95, 2.93, 3.91 and 4.89 as shown in Figs. 1a1d.

3. Experimental set up The experimental set up used for the present study is same as that used in previous paper [3] and is explained in this section. The schematic diagram of the experimental set up is shown in Fig. 2a. It consists of calming section, test section, Rota meters, inlet tank with the capacity 1 cubic meter for supplying water, outlet mixing section, water collection tank capacity 1 cubic meter for receiving water from test section. Calming section with the dimension 2000 mm long, 25.45 mm ID, 33.33 mm OD made of stainless steel tube is used to eliminate the entrance eect. The test section is of smooth stainless steel pipe with the dimension 1800 mm long with 25.45 mm ID, 33.33 mm OD. The outside surface of the tube is brazed with thermowell made of stainless steel having dimensions 6.36 mm ID, 1 mm thick, 120 mm length at distance of 150 mm all along the length in a straight line as shown in Fig. 2b. The test section tube is wound with ceramic beads coated electrical

Fig. 1. Helical screw inserts of dierent twist ratio.

P. Sivashanmugam, S. Suresh / Experimental Thermal and Fluid Science 31 (2007) 301308

303

Fig. 1a. Helical screw inserts of twist ratio 1.95 with various spacer length.

Fig. 1b. Helical screw inserts of twist ratio 2.93 with various spacer length.

Fig. 1c. Helical screw inserts of twist ratio 3.91 with various spacer length.

Fig. 1d. Helical screw inserts of twist ratio 4.89 with various spacer length.

304

P. Sivashanmugam, S. Suresh / Experimental Thermal and Fluid Science 31 (2007) 301308

Fig. 2a. Diagram of experimental set up.

Fig. 2b. Diagram of test section.

SWG Nichrome heating wire of resistance 37 X. Over the electrical winding, two layers of asbestos rope tape is wound. Over the asbestos tape winding approximately 50 mm thickness of glass wool is lined and over which, another two layers of asbestos rope tape is wound to minimize heat loss. The terminals of the Nicrome wire are attached to the Auto-transformer, by which heat ux can be varied by varying the voltage. The Auto-transformer is connected to the servo voltage regulator to minimize the voltage uctuations. Calibrated RTD PT 100 type temperature sensors of 0.1 C accuracy with digital indicator are placed in the thermowell to measure the outside wall temperatures of test section. One end of the electrical heating test section is attached with the calming section, and while the other end is attached with the mixing section (length 200 mm, ID 25.45 mm, OD 33.33 mm), where two baes are provided inside the pipe at a distance of 100 mm from the ange connection for ecient mixing of outlet uid. Flanges are used for attachments, and 50.90 mm thick non-conducting polypropylene disc is placed in-between the anges to prevent the heat conduction ow to the calming section, and the mixing section. Two RTD PT 100 type temperature sensors one just before the test section, and the other after the mixing section are placed to measure the inlet and out let temperature of uid. The inlet tank used for supplying water is tted with porta-

ble agitator not shown experimental diagram for maintaining constant temperature. Two calibrated Rotameters having ow ranges of (0.1 1 103 m3 per minute, and 112 103 m3 per minute) to cover the full laminar ranges are attached to the calming section to measure the ow. The water at constant temperature is being taken from the inlet tank through centrifugal pump as shown in Fig. 2a. The by-pass valve attached to Rotameter is used to regulate the ow rate to the test section. The two pressure tapes one just before the test section and the another just after test section are provided, and attached to U tube manometer for pressure drop measurement. 4. Experimental procedure To start with the centrifugal pump was switched on, and the water ow rate to the test section was adjusted using by-pass valve. The heat ux was set by adjusting the electrical voltage with the help of Auto-transformer, and the constant heat ux was allowed to continue till the steady state is attained. The steady state was obtained within 1 hour for the rst run, and 25 min for the subsequent runs. The inlet, outlet temperature of water, and the wall temperatures in all RTD temperature sensors were noted after steady state. The electrical heat ux was measured by calibrated Ammeter and Volt meter, and also with the help of Watt meter. The ow rate to the test section and heat ux were varied, and readings were taken after attaining steady state. The ow rate was varied from 0.1 103 m3 per minute to 2.4 103 m3 per minute. Experiments were conducted for plain tube, and subsequently by inserting the helical screw insert of various twist ratio, and helical screw twist with various spacer length. The pressure drop was measured for each ow rate with the help of manometer under isothermal condition of ow.

P. Sivashanmugam, S. Suresh / Experimental Thermal and Fluid Science 31 (2007) 301308

305

5. Pressure drop calculation The pressure drop was determined from the dierences in the level of manometer uid. The fully developed friction factor was calculated from the following equation. f Di =LDP =2qu2 m where DP is the pressure drop over length L. 6. Heat transfer calculation The heat transfer rate in the test section was calculated using [6] Q V 2 =R mC p T out T in U o Ao T wo T f where 1=U o Ao 1=hi Ai lnDo =Di =2pk w L 3 2 1

100

Nusselt Number (Nu)

Plain tube Sieder and Tate

10

1 1 10 100 1000 10000

Reynolds Number (Re)

Fig. 3. Data verication of Nusselt number for plain tube.

The internal convective heat transfer coecient, hi was determined by combining Eqs. (2) and (3). The thermal equilibrium test showed that the heat supplied by electrical winding in the test section was 810% larger than the heat absorbed by the uid. This was caused by thermal loss from the test section. The average value of heat transfer rate obtained by heat supplied by electrical winding, and heat absorbed by the uid was taken for internal convective heat transfer coecient calculation. The Nusselt number was calculated using equation Nu hi Di =k i 4

7.2. Eect of spacer length on heat transfer augmentation Figs. 4, 6, 8 and 10 depicts variation of Nusselt number with Reynolds number for full length helical screw twist and twist with dierent spacer length for the twist ratio 1.95, 2.93, 3.91, and 4.89 respectively. Fig. 4 presents variation of Nusselt number with Reynolds number for twist ratio 1.95 and with dierent spacer length. The Nusselt number for full length helical screw twist is higher than the twist with spacer length of 100, 200, 300 and 400 mm spacer length. However the Nusselt number for the helical screw twist with spacer length 100 mm is very much closer to the full length helical screw twist indicating same performance in terms of heat transfer augmentation. This means that intensity of heat transfer

Nusselt Number (Nu)

All the uid thermophysical properties were determined at the average of the inlet and outlet bulk temperatures, Tf. Experimental uncertainty was calculated following Coleman and Steele method [7] and ANSI/ASME standard [8]. The uncertainties associated with the experimental data are calculated on the basis 95% condence level. The measurement uncertainties used in the method are as follows: bulk uid temperature and wall temperatures 0.1 C, uid ow rate 2%, and uid properties 2%. The uncertainty calculation showed that maximum of 6%, 5%, and 8% for Reynolds number, friction factor, and Nusselt number respectively. 7. Results and discussion 7.1. Plain tube data Fig. 3 shows variation of Nusselt number with Reynolds number for plain tube. The data obtained by the experiment are matching with the Sieder and Tata [9] equation of the form given in Eq. (5) for plain tube with the discrepancy of less than 11 % Nu 1:86Gz
1=3

100
Plain tube A A1 A2

10

A3 A4

1 1 10 100 1000 10000

Reynolds Number (Re)

Fig. 4. Nusselt number vs Reynolds number for helical screw of twist ratio 1.95 with dierent spacer length.

l=lw

0:14

306

P. Sivashanmugam, S. Suresh / Experimental Thermal and Fluid Science 31 (2007) 301308

100
Plain tube B

due to swirl ow generated in full-length helical twist is almost in the same order of magnitude of heat transfer enhancement for helical with 100 mm spacer length. The decrease in Nusselt member due to decrease in magnitude of swirl generated for helical twist with spacer length 100 mm and helical twist with spacer length 200 is 10%. Similarly the decrease in Nusselt number for the helical twist with subsequent increase in spacer length (200 and 300 mm, 300 and 400 mm) is also within 10%. These observations indicate that for each 100 mm increase of spacer length, the intensity of heat transfer due to swirl ow decreases by 10%. The heat transfer augmentation increases conventionally with Reynolds number indicating higher heat transfer rate. The above mentioned trends were observed for helical twist with twist ratio 2.93 (Fig. 6), 3.91 (Fig. 8), 4.89 (Fig. 10) and with spacer length 100, 200, 300 and 400 mm. 7.3. Eect of spacer length on friction factor

Nusselt Number (Nu)

B1 B2

10

B3 B4

1 1 10 100 1000 10000

Reynolds Number (Re)

Fig. 6. Nusselt number vs Reynolds number for helical screw of twist ratio 2.93 with dierent spacer length.
1

Figs. 5, 7, 9 and 11 presents variation of friction factor with Reynolds number for full length helical screw twist and twist with dierent spacer length for the twist ratio 1.95, 2.93, 3.91, and 4.89 respectively. Fig. 5 shows the variation of friction factor vs Reynolds number. The friction factor for full length helical screw twist is higher than the plain tube as well as helical screw inserts with dierent spacer length. The friction factor for helical screw twist with 100 mm spacer length is nearly two times lower than the full length helical screw twist at low Reynolds number, and four times lower than the full length helical twist at high Reynolds number. The friction factor for helical screw twist with 200 mm spacer length is nearly four times lower than the full length helical screw twist at low Reynolds number, and six times lower than the full length helical screw twist at high Reynolds number. Whereas for 300 and 400 mm spacer ve times and seven

Plain tube B

Friction factor (f)

0.1

B1 B2 B3

0.01

B4

0.001 1 10 100 1000 10000

Reynolds Number (Re)

Fig. 7. Reynolds number vs friction factor for helical screw of twist ratio 2.93 with dierent spacer length.
100
Plain tube

C C1
Plain tube

Friction factor (f)

Nusselt Number (Nu)

C2 C3

0.1
A1 A2 A3

10

C4

0.01

A4

0.001 1 10 100 1000 10000

1 1 10 100 1000 10000

Reynolds Number (Re)

Fig. 5. Reynolds number vs friction factor for helical screw of twist ratio 1.95 with dierent spacer length.

Reynolds Number (Re)

Fig. 8. Nusselt number vs Reynolds number for helical screw of twist ratio 3.91 with dierent spacer length.

P. Sivashanmugam, S. Suresh / Experimental Thermal and Fluid Science 31 (2007) 301308

1
1
Plain tube

307

Plain tube

Friction factor (f)

0.1

C C1 C2 C3

Friction factor (f)

0.1

D1 D2 D3

0.01
C4

0.01

D4

0.001

0.001

10

100

1000

10000

10

100

1000

10000

Reynolds Number (Re)

Fig. 9. Friction factor vs Reynolds number for helical screw of twist ratio 3.91 with dierent spacer length.
100
Plain tube D D1

Reynolds Number (Re)

Fig. 11. Friction factor vs Reynolds number for helical screw of twist ratio 4.89 with dierent spacer length.

selt number from twist ratio 1.95 to 4.89 is nearly 3040% for all Reynolds number for full length helical twist and helical screw twist with various spacer length, whereas the decrease in friction factor is 4045%. 8. Conclusion Experimental investigation of heat transfer and friction factor characteristics of circular tube tted with fulllength helical screw inserts of dierent twist ratio, and full length with dierent spacer length have been presented. The experimental data obtained were compared with those obtained from plain tube published data. The heat transfer coecient increases with the twist ratio and friction factor also increases with the twist ratio. The decrease in Nusselt number for the helical twist with spacer length is within 10% for each subsequent increase in spacer length. The decrease in friction factor is nearly two times lower than the full length helical twist at low Reynolds number, and four times lower than the full length helical twist at high Reynolds number. The decrease in friction factor is nearly 40% for the twist ratio from 1.95 to 2.93 and decreases in the similar trend for each increase of twist ratio from 2.93 to 3.91, and from 3.91 to 4.89 for all Reynolds number. The decrease in Nusselt number from twist ratio 1.95 to 4.89 is nearly 3040% for all Reynolds number for full length helical screw inserts and helical screw twist with various spacer length, whereas the decrease in friction factor is 4045%. References
[1] A.E. Bergles, Techniques to Augment Heat Transfer, in: W.M. Rohsenow, J.P. Hartnett, E. Ganie (Eds.), Handbook of Heat Transfer Application, McGraw-Hill, New York, 1985. [2] A.E. Bergles, Some perspectives on enhanced heat transfer, secondgeneration heat transfer technology, ASME J. Heat Transfer 110 (November) (1988) 10821096. [3] P. Sivashanmugam, S. Suresh, Experimental studies on heat transfer and friction factor characteristics in laminar ow through a circular

Nusselt Number (Nu)

D2

10

D3 D4

1 1 10 100 1000 10000

Reynolds Number (Re)

Fig. 10. Nusselt number vs Reynolds number for helical screw of twist ratio 4.89 with dierent spacer length.

times lower at low Reynolds number respectively, and six times and times lower than the full length helical screw twist at high Reynolds number. The similar above mentioned observation were made (Figs. 7, 9 and 11) for all other twist set and twist set with dierent spacer length. 7.4. Eect of twist ratio on heat transfer augmentation and friction factor Comparing full length helical screw twist with the twist ratio 1.95 with 2.93, the increase in Nusselt number is only 1012% and increase in the same magnitude for the twist ratio from 2.95 to 3.91 and from 3.91 to 4.89. The decrease in friction factor is nearly 40% for the twist ratio from 1.95 to 2.93 and decreases in the similar way for each increase of twist ratio from 2.93 to 3.92, and from 3.91 to 4.89 for all Reynolds number. The similar behaviors were observed for regularly spaced helical screw inserts. The decrease of Nus-

308

P. Sivashanmugam, S. Suresh / Experimental Thermal and Fluid Science 31 (2007) 301308 [6] Q. Liau, M.D. Xin, Augmentation of convective heat transfer inside tubes with three-dimensional internal extended surfaces and twisted tape inserts, Chem. Eng. J. 78 (2000) 95105. [7] H.W. Coleman, W.G. Steele, Experimental and Uncertainty Analysis for Engineers, Wiley, New York, 1989. [8] ANSI/ASME, 1986, Measurement uncertainty, PTC 19,1 1985, 1986. [9] E.N. Sieder, G.E. Tate, Ind. Eng. Chem. 28 (1936) 1429.

tube tted with helical screw-tape inserts, J. Appl. Therm. Eng., in press. [4] W.J. Marner, A.E. Bergles, Augmentation of highly viscous laminar heat transfer inside the tube with constant wall temperature, Exp. Therm. Fluid Science 2 (1989) 252267. [5] Z. Lejaks, I. Machac, J. Sir, Heat transfer to a Newtonian liquid owing through a tube with an internal helical element, Int. Chem. Eng. 27 (2) (1987) 210217.