ORIGINAL PAPER Experimental and numerical

0 downloads 0 Views 422KB Size Report
coefficient and static pressure difference in a tangential ... Fig. 1. Experimental set-up. calculated the pressure drop from the friction losses in the cyclone body using a wall friction coefficient based on the ... tal set-up shown in Fig. 1. A plexiglas material was used for the cylindrical part of the cyclone to enable monitoring of ...
Chemical Papers 66 (11) 1019–1025 (2012) DOI: 10.2478/s11696-012-0214-7

ORIGINAL PAPER

Experimental and numerical investigation of pressure drop coefficient and static pressure difference in a tangential inlet cyclone separator a

a Department b Department

Fuat Kaya*, b Irfan Karagoz

of Mechanical Engineering, Nigde University, 51240, Nigde, Turkey of Mechanical Engineering, Uludag University, 16059, Bursa, Turkey

Received 3 January 2012; Revised 27 April 2012; Accepted 30 April 2012

The aim of this study was to investigate the pressure drop coefficient and the static pressure difference related to the natural vortex length and to evaluate the results for gas–particle applications. CFD simulations were carried out using a numerical technique which had been verified previously. Results obtained from the numerical simulations were compared with the experimental data. Analysis of the results showed that the pressure drop coefficient decreases with the increasing inlet velocity, becoming almost constant above a certain value of the inlet velocity. The reason is that the effect of viscous forces decreases at high Reynolds numbers. The pressure drop coefficient also decreases with the increasing exit pipe diameter and decreasing exit pipe length. c 2012 Institute of Chemistry, Slovak Academy of Sciences  Keywords: swirling flows, CFD, pressure drop, pressure drop coefficient

Introduction Cyclones are widely used for various purposes, mainly for separation of the dense phase in a twophase flow. There are many parameters such as geometrical dimensions and operation conditions influencing this separation process. Dominant parameters in a cyclone separation process are cyclone dimensions and cyclone inlet velocity. These parameters have been evaluated in the literature to improve separation efficiencies of cyclone separators and to decrease the cyclone pressure drop. Most of these studies were essentially based on experiments due to the difficult theoretical analyses of the highly complex flow occurring in a cyclone separator (Stairmand, 1951; Kim & Lee, 1990; K¨ onig et al., 1991; Lapple, 1951; Moore & McFarland, 1993; Upton et al., 1994; Kenny & Gussman, 1997). Although extensive research has been carried out to investigate the influence of different geometrical parameters such as cyclone length, inlet and outlet pipe

geometries etc. on the performance of cyclones, there has been little work done concerning the detailed research on the cyclone exit pipe. Stairmand (1951) presented one of the most popular cyclone designs and suggested that cylinder height and exit pipe length of a high efficiency cyclone should be 1.5 and 0.5 times the cyclone body diameter, respectively. Obermair et al. (2003) performed cyclone tests with five different dust outlet geometries to investigate the influence of the dust outlet geometry on the separation process. They showed that separation efficiency can be improved significantly by changing the dust outlet geometry, and they concluded that further research is needed to clarify the effects of dust outlet geometry. The effect of a dipleg on the cyclone performance was also suggested and investigated by several researchers (Hoffmann et al., 1996; Xiang et al., 2001; Cortés & Gil, 2007; Kaya & Karagoz, 2009). There are also several studies concerning the pressure drop coefficient. The majority of these studies include theoretical models. Karagoz and Avci (2005)

*Corresponding author, e-mail: [email protected]

Unauthenticated Download Date | 9/24/15 11:26 PM

1020

F. Kaya, I. Karagoz/Chemical Papers 66 (11) 1019–1025 (2012)

Table 1. Geometrical dimensions of the cyclone used a/m

b/m

D2 /m

S/m

h/m

L/m

D3 /m

D1 /m

0.070

0.175

0.047–0.101

0.085–0.120–0.175–0.285–0.350

0.575

1.045

0.075

0.190

used for the cylindrical part of the cyclone to enable monitoring of the flow inside the cyclone. All parts of the cyclone were manufactured as modular. Geometric dimensions of the cyclone used in the experiments are given in Table 1, where a is the inlet height and b is the inlet width, D2 and S are the exit pipe diameter and length, respectively, D1 is the cyclone diameter, h is the cylinder height, L is the cyclone height, and D3 is the cone apex diameter. A single phase fluid flow of air at ambient conditions was used in the study. The velocity profiles were obtained using the Pitot tube measurements and the static pressure difference (∆P) was measured by an inclined manometer. The velocity profiles were obtained using the Pitot tube with an inclined manometer:  2ρm gh sin θ V = (1) ρ where θ is the incline angle of the inclined manometer, ρm is the density of liquid in the manometer, ρ is the density of air, and h is the difference between the manometer heights. The absolute error at a given velocity was calculated by the Taylor expansion as follows:

Fig. 1. Experimental set-up.

calculated the pressure drop from the friction losses in the cyclone body using a wall friction coefficient based on the surface roughness and the Reynolds number. They showed that the pressure drop coefficient increases considerably with Re at low Re values and remains almost constant at high Re values. Recently, with the rapid development of computer and numerical techniques, the use of computational fluid dynamics (CFD) has received much attention in the simulation of the cyclone flow and the prediction of the cyclone performance at different geometrical and operational parameters (Hoekstra et al., 1999; Gil et al., 2002; Gong & Wang, 2004; Gimbun et al., 2005; Chuah et al., 2006; Qian et al., 2006; Karagoz & Kaya, 2007, 2009; Kaya & Karagoz, 2008). This research deals mainly with the optimization of the exit pipe geometry in a tangential inlet cyclone. The effects of exit pipe geometry on the pressure drop coefficient and the static pressure difference related to the natural vortex length were investigated experimentally and numerically.

Experimental Experiments were carried out using the experimental set-up shown in Fig. 1. A plexiglas material was

∆V = V (h, θ) − V (h, θ)n = ∆h

∂V ∂V + ∆θ ∂h ∂θ

(2)

The pressure drop through the cyclone, ∆Pk , was obtained as:   2 Vi − Vo2 ∆Pk = ∆P + ρ + ρg(zi − zo ) (3) 2 and the pressure drop coefficient, K, is a dimensionless number calculated as: K=

∆Pk 0.5ρVi2

(4)

where the subscripts i and o refer to the inlet and outlet sections, respectively. The pressure drop coefficient obtained in the experiments is an overall coefficient, including the entrance, friction, change of direction, and exit effects. Numerical simulations for different geometric and operational parameters were carried out using the mesh structure shown in Fig. 2. The governing equations characterizing the flow inside a cyclone can be written in the Cartesian tensor notation as: ∂ui =0 (5) ∂xi

Unauthenticated Download Date | 9/24/15 11:26 PM

1021

F. Kaya, I. Karagoz/Chemical Papers 66 (11) 1019–1025 (2012)

Fig. 2. CFD surface mesh for the cyclone.

Fig. 4. Comparison of pressure losses obtained experimentally at different exit pipe lengths, S/b: 0.486 (∗), 0.686 ( ), 1 (), 1.629 ( ), 2 ( ), and different inlet velocities (D2 = 0.101 m).



Fig. 3. Comparison of pressure losses obtained experimentally at different exit pipe lengths, S/b: 0.486 (∗), 0.686 ( ), 1 (), 1.629 ( ), 2 ( ), and different inlet velocities (D2 = 0.047 m).



    ∂p ∂ ∂uj ∂ui ∂ui µ − ρui uj = ρgi − + + ∂xj ∂xi ∂xj ∂xj ∂xi (6) Details of the numerical technique employed in this study can be found in the related literature (Kaya & Karagoz, 2008, 2009; Karagoz & Kaya, 2009).

ρuj

Results and discussion This study deals with the effects of exit pipe geometry on the pressure drop coefficient and the static pressure difference related to the natural vortex length and the evaluation of the results for gas-particle applications. Numerical technique verified in previous studies (Kaya & Karagoz, 2008, 2009; Karagoz & Kaya, 2009) was used to optimize the pressure drop coefficient. Values of pressure drop obtained for different inlet velocities and different exit pipe dimensions are given in Figs. 3 and 4. As it can be seen from the diagrams, the pressure drop increased with the increasing cyclone inlet velocity; cyclone dimensions are undoubtedly important parameters influencing the pressure drop. High pressure drop occurred when the diame-

Fig. 5. Comparison of pressure losses obtained experimentally for Vi = 5.1 m s−1 at different exit pipe lengths and diameters, D2 /m: 0.047 ( ), 0.101 ().



ter of the exit pipe was small and the length of the exit pipe was large. While high pressure drop was obtained at small exit pipe diameters, it can be seen that the pressure drop decreased considerably at large exit pipe diameters which also provides a large exit opening resulting in the centrifugal forces effect decrease. In addition, the pressure drop increased with the increasing exit pipe length due to the increase in the friction surfaces. In order to investigate the effects of geometrical dimensions, pressure drop measurements were carried out at different exit pipe diameters (D2 ) and lengths (S). The experimental results for Vi = 5.1 m s−1 are given in Fig. 5. From this figure it can be clearly seen that the pressure drop depends considerably on the exit pipe diameter. The pressure drop for D2 = 0.047 m was about 6–7 times higher than that for D2 = 0.101 m. The effect of the exit pipe dimensions on the pressure loss was more obvious for D2 = 0.047 m and it was negligible for D2 = 0.101 m.

Unauthenticated Download Date | 9/24/15 11:26 PM

1022

F. Kaya, I. Karagoz/Chemical Papers 66 (11) 1019–1025 (2012)

Fig. 6. Comparison of pressure contour obtained numerically at different exit pipe lengths (S = 0.175 m and 0.185 m) and diameters (D2 = 0.047 m and 0.101 m) for Vi = 3.63 m s−1 and 6.5 m s−1 , respectively.

Fig. 7. Comparison of pressure loss coefficients obtained experimentally at different exit pipe lengths, S/b: 0.486 (∗), 0.686 ( ), 1 (), 1.629 ( ), 2 ( ), and different inlet velocities for D2 = 0.047 m.

Fig. 8. Comparison of pressure loss coefficients obtained experimentally at different exit pipe lengths, S/b: 0.486 (∗), 0.686 ( ), 1 (), 1.629 ( ), 2 ( ), and different inlet velocities for D2 = 0.101 m.

The pressure contours obtained numerically for Vi = 3.63 m s−1 , D2 = 47 mm and for Vi = 6.5 m s−1 , D2 = 0.101 m in case of S = 0.175 m and 0.185 m are given in Fig. 6. The obtained results are in good agreement with the experimental data. The maximum deviations for D2 = 0.047 m and 0.101 m were 7.4 % and 1.9 %, respectively. The pressure drop coefficient decreased with the increasing cyclone inlet velocity as given by Karagoz and Avci (2005) for both 0.047 m and 0.101 m exit pipe diameters at large exit pipe lengths, and it was almost constant at small exit pipe lengths (Figs. 7 and 8). The reason is that at low inlet velocities, viscous effects and pressure drop coefficient of the fluid flow are high. Viscous effects decrease with the increasing inlet velocity decreasing thus the effects of surface friction

and the pressure drop coefficient which finally become almost constant. Figs. 9 and 10 show the static pressures obtained numerically and experimentally at locations z = 0.375 m, 0.700 m, 0.800 m, and 0.900 m, shown in Fig. 6, along the cyclone wall. It can be clearly seen that the static pressure decreased with the decreasing inlet velocity and exit pipe length for both exit pipe diameters. It can be concluded that the natural vortex end reaches the bottom of the cyclone due to the constant static pressure difference between the locations at high inlet velocities (Fig. 9). Thus, the particle collection efficiency increases due to a better separation in gas–particle applications. The static pressure line for S = 0.085 m in Fig. 9 as well as all lines in Fig. 10 indicate that





Unauthenticated Download Date | 9/24/15 11:26 PM

1023

F. Kaya, I. Karagoz/Chemical Papers 66 (11) 1019–1025 (2012)

s−1

Fig. 9. Comparison of static pressures for Vi = 3.63 m and D2 = 0.047 m obtained experimentally and calculated at different exit pipe lengths, S/m: 0.085 (∗), 0.120 ( ), 0.175 (), 0.175 (cal.) (), 0.285 ( ), 0.285 (cal.) ( ), 0.350 ( ), and different locations (z = 0.375 m, 0.700 m, 0.800 m, and 0.900 m)

Fig. 11. Comparison of axial velocity profiles obtained experimentally and calculated for Vi = 3.63 m s−1 at different exit pipe lengths, S/m: 0.175 (cal.) ( ), 0.175 ( ), 0.285 (cal.) (), 0.285 (), and for D2 = 0.047 m.







Fig. 12. Comparison of axial velocity profiles obtained experimentally and calculated for Vi = 6.5 m s−1 at different exit pipe lengths, S/m: 0.175 (cal.) ( ), 0.175 ( ), 0.285 (cal.) (), 0.285 (), and for D2 = 0.101 m.



Fig. 10. Comparison of static pressures for Vi = 6.5 m s−1 and D2 = 0.101 m obtained experimentally and calculated at different exit pipe lengths, S/m: 0.085 (*), 0.120 ( ), 0.175 (), 0.175 (cal.) (), 0.285 ( ), 0.285 (cal.) ( ), 0.350 ( ), and different locations (z = 0.375 m, 0.700 m, 0.800 m, and 0.900 m)



the vortex end is located between z = 0.750 m and 0.900 m. The main reason is probably the distortion of the vortex structure. In this case, the particle collection efficiency decreases due to the decreasing separation space in gas-particle applications. The maximum deviation between numerical results and experimental data was 6.8 %. Axial and tangential velocity profiles were also obtained numerically and experimentally at different axial locations. Two different vortex regions are clear as can be seen from the axial velocity profiles in Figs. 11



and 12. A vortex moves downwards along the wall and a second vortex moves upwards in the center. The downward moving vortex was smaller and the upward moving one was larger in case of increased exit pipe diameter and decreased exit pipe length for the same inlet velocities. The main reason is that the centrifugal forces decrease and, therefore, the flow turns upward before reaching the bottom of the cyclone. It can be seen from Figs. 13 and 14 that the maximum tangential velocity was about 2.8 times the inlet velocity for D2 = 0.047 m, S = 0.285 m and about 1.2 times the inlet velocity for D2 = 0.101 m at the same exit pipe length. Therefore, high tangential velocities provide higher separation efficiency in gas–particle applications. Fig. 15 shows the velocity magnitude profiles and the absolute error of the velocity magnitude at z =

Unauthenticated Download Date | 9/24/15 11:26 PM

1024

F. Kaya, I. Karagoz/Chemical Papers 66 (11) 1019–1025 (2012)

Conclusions

Fig. 13. Comparison of tangential velocity profiles obtained experimentally and calculated for Vi = 3.63 m s−1 at different exit pipe lengths, S/m: 0.175 (cal.) ( ), 0.175 ( ), 0.285 (cal.) (), 0.285 (), and for D2 = 0.047 m.





Fig. 14. Comparison of tangential velocity profiles obtained experimentally and calculated for Vi = 6.5 m s−1 at different exit pipe lengths, S/m: 0.175 (cal.) ( ), 0.175 ( ), 0.285 (cal.) (), 0.285 (), and for D2 = 0.101 m.



The effect of exit pipe dimensions on the pressure drop coefficient and the static pressure difference in a tangential inlet cyclone separator was investigated experimentally and numerically. The static pressure decreased with the decreasing inlet velocity and exit pipe length for all exit pipe diameters. It can also be concluded that the vortex end reached the bottom of the cyclone at high inlet velocities. Two different vortex regions can be clearly seen from the axial velocity profiles. A vortex moves downwards along the outer wall and a second vortex moves upwards in the center of the cyclone. The downward moving vortex became relatively smaller and the upward moving vortex became larger in case of increased exit pipe diameters and decreased exit pipe lengths for the same inlet velocity. The maximum tangential velocity decreased with the increasing exit pipe diameter, while all other parameters remained constant. High pressure drop occurred when the diameter of the exit pipe was small and the length of the exit pipe was large. While high pressure drop was obtained at small exit pipe diameters, it can be seen that the pressure drop decreased considerably at large exit pipe diameters providing a large exit opening resulting in the centrifugal forces effect decrease. In addition, pressure drop increased with the increasing exit pipe length due to the increase in the friction surfaces. The pressure drop coefficient decreased and then became almost constant with the increasing inlet velocity, and it also decreased with the increasing exit pipe diameter and decreasing exit pipe length.



Fig. 15. Velocity magnitude profiles and the absolute error of the velocity magnitude at z = 0.375 m for S = 0.175 m and D2 = 0.101 m at the inlet velocity Vi = 9.5 m s−1 .

0.375 m for S = 0.175 m and D2 = 0.101 m at the inlet velocity of 9.5 m s−1 .

References Chuah, T. G., Gimbun, J., & Choong, T. S. Y. (2006). A CFD study the effect of cone dimensions on sampling aerocyclones performance and hydrodynamics. Powder Technology, 162, 126–132. DOI: 10.1016/j.powtec.2005.12.010. Cortés, C., & Gil, A. (2007). Modeling the gas and particle flow inside cyclone separators. Progress in Energy and Combustion Science, 33, 409–452. DOI: 10.1016/j.pecs.2007.02.001. Gil, A., Cortés, C., Romeo, L. M., & Velilla, J. (2002). Gasparticle flow inside cyclone diplegs with pneumatic extraction. Powder Technology, 128, 78–91. DOI: 10.1016/s00325910(02)00215-2. Gimbun, J., Chuah, T. G., Fakhru’l-Razi, A., & Choong, T. S. Y. (2005). The influence of temperature and inlet velocity on cyclone pressure lose: a CFD study. Chemical Engineering and Processing, 44, 7–12. DOI: 10.1016/j.cep.2004.03.005. Gong, A. L., & Wang, L. Z. (2004). Numerical study of gas phase flow in cyclones with the repds. Aerosol Science and Technology, 38, 506–512. DOI: 10.1080/02786820490449548. Hoekstra, A. J., Derksen, J. J., & Van Den Akker, H. E. A. (1999). An experimental and numerical study of turbulent swirling flow in gas cyclones. Chemical Engineering Science, 54, 2055–2065. DOI: 10.1016/s0009-2509(98)00373-x. Hoffmann, A. C., De Groot, M., & Hospers, A. (1996). The effect of the dust collection system on the flowpat-

Unauthenticated Download Date | 9/24/15 11:26 PM

F. Kaya, I. Karagoz/Chemical Papers 66 (11) 1019–1025 (2012)

tern and separation efficiency of a gas cyclone. The Canadian Journal of Chemical Engineering, 74, 464–470. DOI: 10.1002/cjce.5450740405. Karagoz, I., & Avci, A. (2005). Modelling of the pressure drop in tangential inlet cyclone separators. Aerosol Science and Technology, 39, 857–865. DOI: 10.1080/02786820500295560. Karagoz, I., & Kaya, F. (2007). CFD investigation of the flow and heat transfer characteristics in a tangential inlet cyclone. International Communications in Heat and Mass Transfer, 34, 1119–1126. DOI: 10.1016/j.icheatmasstransfer.2007.05. 017. Karagoz, I., & Kaya, F. (2009). Evaluations of turbulence models for highly swirling flows in cyclones. Computer Modeling in Engineering & Sciences, 43, 111–130. DOI: 10.3970/cmes.2009.043.111. Kaya, F., & Karagoz, I. (2008). Performance analysis of numerical schemes in swirling turbulent flows in cyclones. Current Science, 94, 1273–1278. Kaya, F., & Karagoz, I. (2009). Numerical investigation of performance characteristics of a cyclone prolonged with a dipleg. Chemical Engineering Journal, 151, 39–45. DOI: 10.1016/j.cej.2009.01.040. Kenny, L. C., & Gussman, R. A. (1997). Characterizations and modelling of a family of cyclone aerosol preseparators. Journal of Aerosol Science, 28, 677–688. DOI: 10.1016/s00218502(96)00455-7. Kim, J. C., & Lee, K. W. (1990). Experimental study of particle collection by small cyclones. Aerosol Science and Technology, 12, 1003–1015. DOI: 10.1080/02786829008959410.

1025

K¨ onig, C., B¨ uttner, H., & Ebert, F. (1991). Desing data for cyclones. Particle & Particle Systems Characterization, 8, 301–307. DOI: 10.1002/ppsc.19910080155. Lapple, C. E., (1951). Processes use many collector types. Chemical Engineering, 58, 144–151. Moore, M. E., & McFarland, A. R. (1993). Performance modelling of single-inlet aerosol sampling cyclones. Environmental Science & Technology, 27, 1842–1848. DOI: 10.1021/es00046a012. Obermair, S., Woisetschläger, J., & Staudinger, G. (2003). Investigation of the flow pattern in different dust outlet geometries of a gas cyclone by laser Doppler anemometry. Powder Technology, 138, 239–251. DOI: 10.1016/j.powtec.2003.09. 009. Qian, F. P., Zhang, J. G., & Zhang, M. Y. (2006). Effects of the prolonged vertical tube on the separation performance of a cyclone. Journal of Hazardous Materials, 136, 822–829. DOI: 10.1016/j.jhazmat.2006.01.028. Stairmand, C. J. (1951). The design and performance of cyclone separators. Transaction of the Institution of Chemical Engineers, 29, 356–383. Upton, S. L., Mark, D., Douglass, E. J., Hall, D. J., & Griffiths, W. D. (1994). A wind tunnel evaluation of the physical sampling efficiencies of three bioaerosol samplers. Journal of Aerosol Science, 25, 1493–1501. DOI: 10.1016/00218502(94)90220-8. Xiang, R. B., Park, S. H., & Lee, K. W. (2001). Effects of cone dimension on cyclone performance. Journal of Aerosol Science, 32, 549–561. DOI: 10.1016/s0021-8502(00)00094-x.

Unauthenticated Download Date | 9/24/15 11:26 PM