ENHANCEMENT OF MASS TRANSFER WITH SQUARE-GROOVED SERRATED DISC AS TURBULENCE PROMOTER IN CIRCULAR TUBE D.S.Seetha Rama Rajua*, M.

Gangadharb, V Nageswara Raoa, N.Chitti Babua a Department of Chemical Engineering, College of Engineering(A), Andhra University, Visakhapatnam-530 003, Andhra Pradesh, INDIA b Department of Chemical Engineering, G.M.R. Institute of Technology, GMR Nagar, Rajam, Srikakulam-532127, Andhra Pradesh, INDIA *E-mail: sai_dandu@yahoo.co.in ABSTRACT Mass transfer data were obtained in circular conduit using an electrochemical technique with a potassium ferri-ferro cyanide couple. In circular conduit coaxially placed entry region square-grooved serrated disc was used as turbulence promoter. The study comprised of evaluation of mass transfer rates at the outer wall of the electrochemical cell. Mass transfer coefficients were evaluated from the measured limiting currents. The study covered a wide range of geometric parameters such as diameter of the disc (dd), thickness of the disc (dt) and distance of the disc from the entrance of the test section (h). The results revealed that the mass transfer coefficient increased with increase in velocity, diameter of the disc (dd), thickness of the disc (dt) and decreased with increase in distance of the disc from the entrance of the test section (h). Within the range of variables covered, the augmentation achieved in mass transfer coefficients were up to 2.1 fold over the tube flow in absence of promoter. The entire mass transfer data were correlated with Colburn JD factor and roughness Reynolds number. The following correlation was reported out of the study. JD = 128.40 (Re+)-0.9753 (1)0.9224(2)0.1785 (3)0.4109 Where 1=dd/d, 2=dt/d, 3=h/d are dimensionless groups. d is diameter of test section KEYWORDS: Mass transfer, square- grooved serrated disc, turbulence promoter 1. INTRODUCTION: Researchers will always aim at the maximum output with reduced equipment size so as to minimize the unit product cost. This is particularly true in the design of electrolytic cells where the mass transfer limiting conditions exist. Use of turbulence promoter is one such technique that increases the transfer coefficients by several folds over the smooth flow. Disc placed across the flow inside the tube generates wakes and eddies within the tube. This disturbance continues to its effect on either side of disc and continues to a longer distance which in turn influences the wall mass transfer coefficient over the length of the tube. The study was conducted to know the effectiveness of disc in augmenting mass transfer coefficient. Burns and Jachuck[1] determined liquidsolid mass transfer performance on a 30cm diameter spinning disc reactor by use of the limiting current technique. Ravi et al [2] did experiments and obtain mass transfer data on the walls of the fludized beds using an electrochemical technique with a potassium ferro-ferri cyanide couple. The disc promoter assembly was used as internal in fludized beds. Sherwood and Stone [3] determined the flow of a viscous fluid around disc in a pipe and the added mass of the accelerating disc and flow profiles were also presented. Podesta et al [4] obtained a dimensionless correlation for ionic mass transfer at horizontal disc electrodes under longitudinal vibration. Escudie et al [5] carried out experiments in water fluidized binary mixtures of Teflon spheres, discs and rods. Venkateswarlu et al [6] studied the effect of string of discs placed co-axially in a circular conduit on mass and momentum transfer. Data on mass transfer in circular conduit with coaxially placed square-

grooved serrated disc as turbulence promoter has been taken up since the data on the present system has not been reported in literature. Based on the data a semi-theoretical model has been developed. Parameters covered in the present study are shown below in Table 1 Table 1 Variable Minimum Maximum Max/Min Diameter of the disc, (dd), m 0.025 0.045 1.8 Disc thickness, (dt), m Distance of the disc from the entrance of the test section, (h),m Velocity, V, m/s Reynolds number, Re Schmidt number 0.001 0.10 0.03289 1933 982.2 0.005 0.26 0.3289 19337 1040 5 2.6 10 10 1.0588

2. EXPERIMENTAL PROCEDURE: The schematic diagram of experimental setup is shown in figure 1. It is similar in layout to that was used in earlier studies [6,8,12]. It essentially consists of a storage tank (TS), centrifugal pump (P), rotameter (R), entrance calming section (E1), test section (T) and exit-calming section (E2). The storage tank is cylindrical copper vessel of 100 liter capacity with a drain pipe and a gate valve (V1) to facilitate periodical cleaning. A copper coil (H) with perforations is provided to bubble nitrogen through the electrolyte. The tank is connected to the pump of 2 HP motor with a 0.025m diameter copper pipe on the suction line of the centrifugal pump. The suction line is also provided with a gate valve (V2). The discharge line from the pump splits into two. One served as a bypass line and controlled by valve (V3). The other connects the pump to the entrance calming section (E1) through rotameter. The rotameter is connected to a valve (V4) for adjusting the flow at the desired value. The rotameter has a range of 0-166x10-5 m3/s. The entrance calming section consisted of 0.05 m ID circular copper pipe with a flange and is closed at the bottom with a gland nut (G). The up-stream side of the entrance calming section is filled with capillary tubes to damp the flow fluctuations and to facilitate steady flow of the electrolyte through the test section. Test section is made of a graduated perspex tube of 0.36m length with point electrodes fixed flush with the inner surface of the tube. The point electrodes are made out of a copper rod and machined to the size. They are fixed flush with the inner surface of the test section at equal spacing of 0.01m. Exit calming section is also of the same diameter copper tube of 0.5 m long, and it is provided with a flange on the upstream side for assembling the test section. It has gland nuts (G) at the top and bottom ends to hold the central tube. Two thermo wells (t1, t2) are provided, one at upstream side of the entrance calming section and the other at the downstream side of exit calming section for measurement of temperature of the electrolyte. Square- grooved serrated disc serving as turbulence promoter is made of copper of various sizes with a provision to fix it rigidly within the test section. The serrated disc is placed concentrically in the test section. The limiting current measuring equipment consisted of multimeter of Motwane make that has 0.01mA accuracy and vacuum tube voltmeter is used for potential measurements. The other equipment used in circuit is rheostat, key, commutator, selector switch, and a lead acid battery as the power source. The commutator facilitated the measurement of limiting currents for oxidation and reduction process under identical operating conditions by the change in polarity while the selector switch facilitated the

measurements of limiting currents at any desired electrode along the test section. The circuit diagram used for the measurement of limiting currents is shown in the figure 3.7. Data on limiting currents for the case of reduction of ferricyanide ion is obtained for fluid flow in circular conduits in the presence of square- grooved serrated disc as insert promoter. The following electrode reaction is involved. Cathodic reduction of ferricyanide ion: [Fe(CN)6]-3 + e[Fe(CN)6]-4 ...... (1)

Eighty liters of equimolal solution of 0.01M Potassium ferricyanide and Potassium ferrocyanide with 0.5N NaOH as indifferent electrolyte were prepared. The electrolyte is analyzed for ferrocyanide ion concentration by volumetric titration method using standard potassium permanganate solutions (1) and for ferricyanide ion using idometric method (2). The viscosity of the solution at different temperatures was measured with Ostwald Viscometer and densities were measured using specific gravity bottle. The point electrodes are polished using four zero emery to get a smooth surface without protuberances. The size of the electrode is measured with a two-tank micrometer and also with a traveling microscope. Initially blank runs are conducted with indifferent electrolyte (sodium hydroxide solution) alone to ensure that the limiting currents obtained in the subsequent runs are due to diffusion of reacting ions (Ferri cyanide ion) only. The electrolyte is pumped at a desired flow rate (through the test section) by operating the control and by-pass valves. After steady state is attained, potentials are applied across the test electrode and wall electrode in small increments of potentials (100mV) and the corresponding currents are measured for each increment. In view of the large area of the counter electrode in relation to the test electrode nearly constant potential is maintained at the test electrode. Since the potential values are not of criteria in the present study, the limiting currents are determined from the measurements of applied potentials and currents as has been done in several earlier works [ 6,8,12]. The attainment of limiting current is indicated by the constancy of current with a large increase in the potential. Mass transfer coefficients are computed from the measured limiting currents by the following equation: iL kL = (2) n.F . A.C 0 Pressure drop measurements are also taken simultaneously using a U-tube manometer with Carbon tetrachloride as manometric liquid. The turbulence promoters are shown in figure 2. 3. RESULTS Mass transfer coefficients obtained were arithmetically averaged and average mass transfer coefficients are computed. Mass transfer coefficients were influenced by dynamic and geometric parameters. The geometric parameters covered in the present study are square- grooved serrated disc diameter (dd), serrated disc thickness (dt) and distance of the serrated disc from entrance of the test section (h). Effect of disc diameter: Disc diameter has strong influence on mass transfer coefficient KL. KL versus velocity (V) was drawn for different disc diameters and is shown in figure 3.Various disc diameters used in the present

study are dd=0.025m, 0.030m, 0.035m, 0.040m and 0.045m.Mass transfer coefficient increases with increase in disc diameter. The augmentation in mass transfer coefficient is 1.25 times over the smooth tube [7] for disc diameter 0.025m at a velocity of 0.3289m/s while the augmentation is 2.08 times over the smooth tube [7] for the disc diameter 0.045m at the same velocity 0.3289m/s. Effect of disc thickness (dt): In figure 4, mass transfer coefficient was plotted against velocity for the different disc thickness values of dt=0.001m, 0.002m, 0.003m, 0.004m and 0.005m by keeping all the other geometric parameters constant at dd=0.045m,h=0.18m.From the figure it was revealed that the mass transfer increases from 1.87 times 2.63 times over Lin et al[7] for the smooth tube as the thickness of the disc (dt) increases from 0.001m to 0.005m at the velocity of 0.3289m/s. Effect of distance of the serrated disc from the entrance of the test section (h): As the distance from the entrance of the test section (h) varied, the extent of turbulence also varied because of the change in circulatory pattern. Mass transfer coefficient versus velocity data for different values of h were plotted and shown in figure 5.Mass transfer coefficients are decreased from 1.48 times to 2.28 times over smooth tube of Lin et al[7] while increasing h from 0.10m to 0.26m at the velocity of 0.3289m/s. Table 2 indicates the exponent on velocity of the present study together with the other works. The exponent on velocity is comparable to that in the studies on mass transfer with different turbulence generating systems. Table 2 Comparison with other works: Author Promoter System Exponent Range of Re on velocity Klaczack [10] Sujatha [12] Venkateswarlu [6] Sitaraman [11] Sarveswara Rao [8] Changal Raju [9] Nageswara Rao [13] Present work Spiral coil Tapes mounted on a rod String of discs String of spheres String of cones Wires wound on a rod Single disc Square- Grooved Serrated disc Mass transfer Mass transfer Mass transfer Mass transfer Mass transfer Mass transfer Mass transfer Mass transfer 0.520 0.490 0.498 0.556 0.431 0.490 0.405 0.431 1700-20000 1348-30605 3300-18650 100-34000 690-20200 1500-20000 1300-12000 1933-19337

4. CORRELATION DEVELOPED: The data on mass transfer with single serrated disc as turbulence promoter could well be calculated in the lines done in earlier studies [8,12].Correlation of data using Colburn JD factor with roughness Reynolds number have yielded the following equation JD = 91.846(Re+)-0.9723 .....................(3)

Average deviation = 20.93, Standard deviation = 24.15 By incorporating dimensionless geometrical groups, the following correlations are yielded JD = 128.40 (Re+ )-0.9753 (1)0.9224(2)0.1785 (3)0.4109 ...(4) Average deviation = 1.69, Standard deviation = 2.14 Where 1=dd/d, 2=dt/d, 3=h/d, which are dimensionless groups and Sc is Schmidt number. Correlation graph for equation 4 is given in the figure 6.

Comparison of correlations:
For a selected set of geometric parameters correlation factor in mass transfer (Y3) is plotted against Re+. For comparison of data with the other studies, Nageswara Rao data [13] having comparable geometric parameters, is taken and computed with present method and shown in figure 7. The data falls close to the present study indicating correlation presented in the present work is comparable and better correlating. Figure 8 shows efficiency of promoter against Reynolds number for the same set of data points which augment best of present study and compared with data of Nageswara Rao [13]. The figure shows efficiency of promoter is increasing with increase in Reynolds number of present study. More over efficiency of promoter of present study is found to be higher when compared to the studies of Nageswara Rao [13] 5. CONCLUSIONS: Mass transfer coefficient increases with increase in velocity, serrated disc diameter (dd), thickness of the serrated disc (dt) and decreases as distance of the disc (h) from the entrance of the test section increases. In the present study, it is found that disc diameter (dd) of 0.045m, disc thickness=0.005m, distance of the disc from the entrance of the test section (h) =0.10m, offered maximum augmentation. Maximum augmentation of 2.1 folds is observed over smooth tube flow. This is valid for the range of variables covered in this study, but probably capable of predicting data at higher ranges. The data generated and correlation developed is helpful in the design of efficient electrolytic reactors for organic synthesis and other applications. Dimensionless Groups: JD = Mass Transfer Factor (kL/V).SC2/3 Re = Reynolds number = dV/ + Re = Roughness Reynolds number = (dd / d).Re.(f/2) St = Stanton number = kL/V = Stanton number for conduit without internals Sto Sc = Schmidt number /DL Sh = Sherwood number kL.d/DL u+ = dimensionless velocity, u/u* y+ = dimensionless radial distance from the wall, y u*/ Nomenclature: d = Diameter of test section, m DL = Diffusivity of reacting ion, m2/sec dd = Disc diameter, m f = Friction factor, p d gc /2LV2 P = Pressure difference, N/m2 F = Faradays constant = 96,500 coulombs/g-mol g = Acceleration due to gravity, m/sec2

gc = Gravitational constant. iL = Limiting current, amp kL = Mass Transfer coefficient, m/s ko = Mass transfer coefficient of the empty conduit, m/s L = Length of test section, m n = Number of electrons transferred Q = Volumetric flow rate, m3/s dd = diameter of the disc, m dt = thickness of the disc, m h = Location of the disc from the entrance of the test section, m u = Local velocity, m/s * u = Friction velocity = (w gc /) , m/s V = Average velocity, m/s y = Radial distance from the wall, m Y3 = JD / (dd/d)0.9224(dt/d)0.1785 (h/d)0.4109 Greek letters: D c w = = = = = = = Eddy viscosity, m2/s Eddy diffusivity, m2/s Viscosity of fluid, Kg/m. sec Kinematic viscosity, m2/s Density of manometer fluid, Kg/m3 Density of fluid, Kg/m3 Shear stress , N/m2

References:
[1] J.R.Burns and R.J.J.Jchuck, Determination of liquid solid mass transfer coefficient for a spinning disc reactor using limiting current technique, Int.J.Heat and Mass Transfer,48,2005,2540-2547 [2] T.Ravi,B.Srinivasa Rao,P.Gopala Krishnaand P.Venkateswarlu Ionic mass transfer studies in fluidized beds with co-axially placed discs on a rod as internal Chem.Engg.Process, 35, 1996, 187-193 [3] J.D.Sherwood ,H.A.Stone Added mass of a disc accelerating within a pipe Phys.Fluids., 9,1997,3141-3148 [4] J.J.Padesta,G.S.Paus,A.J.Arvia Ionic mass transfer at horizontal disc electrode under longitudinal vibration Electrochem.Acta, 19, 1974, 583-589 [5] R.Escudie,N.Epstein,.R.Grace,and H.T.Bi Effect of particle shape on liquid fluidized beds of binary solids mixtures, Chem.Engg.Science., 61,2006,1528-1539 [6] P.Venkateswarlu,T.Gopichand,G.J.V.J.Raju Increased mass transfer in a circular column in the presence of disc promoter J.Energy Heat Mass Transfer, 2000, 22,195-203 [7] Lin C.S.,E.B.Denton,H.S.Gaskill and G.L.Putnam Diffusion controlled reaction, Ind.Engg.Chemistry, 1951, 43, 2136-2143

[8] S.Sarveswara Rao, G.J.V.J.Raju Studies on Ionic Mass Transfer with co-axially placed cones on a rod in homogeneous fluid and fluidized beds, Paper presented at World Congress III of Chemical Engineers, 1986,Tokyo,p481 [9] D.Changal Raju,C.Bhaskara Sharma,G.J.V.Raju, Ionic mass transfer at the outer wall of concentric annuli in the presence of fluidizing solidsEffect of wires wound on central rod, J.Energy Heat Masss Transfer, 1992, 14, 151-156 [10] Klaczac,A Trans. A.S.M.E., J. Heat Transfer, 1973, 95, 557 [11] Sitaraman, T.S Augmentation of mass transfer with coaxial string of spheres as internals in tubes and fluidized beds, 1977, Ph.D. Thesis, University of Madras, India [12] V.Sujatha,C.Bhaskara Sarma,G.J.V.Jagannadha Raju Studies on ionic mass transfer with coaxially placed helical tapes on a rod in forced convection flow, Chem.Engg.Process.,1997,36, 67-73 [13] Nageswara Rao, V.and Chitti Babu, N. Studies on ionic mass transfer with coaxially placed disc as turbulence promoter in circular Conduits, Caledonian journal of Engineering, Vol.4 (2009)

Figure 1 Schematic Diagram of Experimental Setup

Figure 2 Turbulence Promoters

10 -5

kL(m/s)

dd
0.025

dt
0.001

V
0.18

0.030 0.001 0.18 0.035 0.001 0.18 0.040 0.001 0.18 0.18 0.045 0.001 Lin et al

10 -6 10 -1 V (m /s) Variation of M ass Transfer C oefficient w ith Velocity -Effect of D isc D iam eter

Figure 3

10 -5 kL(m/s)

dd dt 0.045 0.045 0.001 0.001 0.18 0.045 0.045 0.002 0.002 0.18 0.045 0.045 0.003 0.003 0.18 0.045 0.045 0.004 0.004 0.18 0.045 0.045 0.005 0.005 0.18 Lin et Lin al et al 10 -6 10 -1

h 0.18 0.18 0.18 0.18 0.18

V (m /s) V ariation of M ass T ransfer C oefficient w ith V elocity -E ffect of D isc T hickness

Figure 4

10 -5

kL(m/s)

dd 0.045 0.045 0.045 0.045 0.045 Lin et 10 -6 10 -1 V(m/s)

dt 0.001 0.001 0.001 0.001 0.001 al

h 0.26 0.22 0.18 0.14 0.10

Variation of Mass Transfer Coefficient with Velocity -Effect of Distance of Disc from Entrance of the Test Section

Figure 5

Mass Transfer Correlation Factor [CFR(jD)]

10-1

10-2

103 Re Correlation plot for equation (4)

104

Figure 6

102

101 Mass Transfer Correlation Factor

100

10-1

10-2

10-3

Present study Data of Nageswara Rao 103 Re + + 104

10-4

Variation of CFR with Re - Comparision plot

Figure 7

10 1 Augmentation factor (St / St0)

10 0

10 -1

Present data Data of Nageswara Rao 10 -2 10 3 Re Variation of St/St 0 with Reynolds Num ber -Com parision plot 10 4

Figure 8