Engineering Applications of Computational Fluid ...

This article was downloaded by: [Universiti Malaysia Pahang UMP] On: 19 March 2015, At: 06:11 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Engineering Applications of Computational Fluid Mechanics Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/tcfm20

CFD Simulation of Aerocyclone Hydrodynamics and Performance at Extreme Temperature a

Jolius Gimbun a

Faculty of Chemical and Natural Resources Engineering, Universiti Malaysia Pahang, UMP, MEC Town, 25200 Kuantan, Pahang D.M., Malaysia E-Mail: Published online: 19 Nov 2014.

To cite this article: Jolius Gimbun (2008) CFD Simulation of Aerocyclone Hydrodynamics and Performance at Extreme Temperature, Engineering Applications of Computational Fluid Mechanics, 2:1, 22-29, DOI: 10.1080/19942060.2008.11015208 To link to this article: http://dx.doi.org/10.1080/19942060.2008.11015208

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content. This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http:// www.tandfonline.com/page/terms-and-conditions

Engineering Applications of Computational Fluid Mechanics Vol. 2, No. 1, pp. 22–29 (2008)

CFD SIMULATION OF AEROCYCLONE HYDRODYNAMICS AND PERFORMANCE AT EXTREME TEMPERATURE Jolius Gimbun

Downloaded by [Universiti Malaysia Pahang UMP] at 06:11 19 March 2015

Faculty of Chemical and Natural Resources Engineering, Universiti Malaysia Pahang, UMP, MEC Town, 25200 Kuantan, Pahang D.M., Malaysia E-Mail: [email protected] ABSTRACT: This work presents a Computational Fluid Dynamics calculation to predict and to evaluate the hydrodynamics and performance of a cyclone operating at extreme temperature. The numerical solutions were carried out using commercial CFD code FLUENT 6.1. The simulation was realised using a Reynolds stress model (RSM) for turbulent modelling and discrete phase model (DPM) for particle trajectories calculation. The kinetic theory was employed to predict the physical gas properties, i.e., ρ, Cp and μ as they vary with the operating temperature. The CFD simulations predict excellently the cyclone performance and gas properties at extreme operating conditions with a deviation of about 5% from the experimental data. The physical mechanism for cyclones operating under high temperature has been also successfully elucidated. Results obtained from the computer modelling exercise have demonstrated that CFD is a reliable method of modelling the cyclone performance at extreme temperature. Therefore, similar method can be applied to examine the effects of operating temperature on the cyclone performance. This method of analysis is almost certainly less expensive than experiment, and represents a cost-effective route for design optimisation. Keywords:

cyclone, high temperature, collection efficiency, pressure drop, hydrodynamics

complicated fluid and particulate flows in cyclones, and the lack of a well established theory. The development of new cyclones still largely relies on empirical formulae, or simple modifications of the old designs. Therefore, very costly experimental assessment and a hefty trial-and-error procedure are usually unavoidable. The working principles of a cyclone chamber operating under extreme temperature are far from being fully understood, mainly owing to the extreme complexity of the swirling turbulent flow field inside the device. The complexity of the flow pattern inside the chamber is due to the high turbulence level, strong anisotropy, threedimensionality and possible non-stationary features typical of highly swirling flows, so that both experimental analysis and numerical simulations become notably difficult. The complicated swirling turbulent flow in a cyclone places great demands on the numerical techniques and the turbulence models employed in the CFD codes when modelling the cyclone flow field and performance. CFD has a great potential to predict the flow field characteristics and particle trajectories inside the cyclone as well as the pressure drop (Slack et al.,

1. INTRODUCTION Aerocyclones or gas cyclones are used in many technical applications (Gupta, Lilley and Syred, 1984), such as physical separation processes (especially dust from a gas stream) and chemical reactions (combustion of solid low calorific value fuels). Their simple design, low capital cost and nearly maintenance-free operation make them ideal for use as pre-cleaners for more expensive final control devices such as baghouses or electrostatic precipitators. Cyclones are particularly well suited for high temperature and pressure conditions because of their rugged design and flexible components materials. Cyclones are used for the removal of large particles for both air pollution controls and process use. Applications in an extreme condition include the removal of dust in power generation plants and incineration plants and the use as a spray dryer or gasification reactor. The design of cyclones operating at high temperatures is still a difficult problem as practical experience shows that the pressure drop and collection efficiency really obtained are mostly different from those calculated. It is due to the very Received: 4 Jun. 2007; Revised: 11 Jul. 2007; Accepted: 27 Aug. 2007 22

Engineering Applications of Computational Fluid Mechanics Vol. 2, No. 1 (2008)

Many different types of cyclones have been built but the reverse flow cyclone with the tangential inlet in Fig. 1 is most often used for industrial gas cleaning (Li and Wang, 1989; Altmeyer et al., 2004). In this study, the numerical simulation deals with the reverse flow cyclone with a tangential rectangular inlet. The key geometrical features of this type of cyclone are identified in Fig. 1 and detailed dimensions given in Table 1.

2000; Griffith and Boysan, 1996). CFD is also capable of predicting the cyclone performance at various operating conditions as demonstrated in our previous work (Gimbun et al., 2005a & 2005b). This study is performed in an attempt to uncover the effect of temperature on cyclone hydrodynamics as well as to evaluate numerically the cyclone performance under extreme operating temperature.


a

ar

le

B De

h S

ad

b

D

H

Fig. 1

Tangential aerocyclone dimensions.

Table 1 Cyclone geometry used in this simulation. Geometry (mm)

a

b

De

S

h

H

le

ad

ar

B

D

Bohnet (1995)

80

20

50

104

110

387

245

210

44

50

150

2. CFD APPROACH The FLUENT software version 6.1 was used for the CFD simulations. The cyclone chamber was represented using a hexahedral cooper mesh consisted of 20,846 nodes (Fig. 2(a)). A grid refinement was applied in the central region of the cyclone to ensure a better prediction of cyclone hydrodynamics (Fig. 2(b)). A velocity inlet boundary was used to specify an air inlet velocity ranging from 4.62 to 14.16 m/s. The simulations were carried out with a zero underflow component. The underflow was therefore represented using a wall boundary. An outflow boundary condition was used to represent the cyclone overflow. The air properties such as viscosity and density at predefined temperature are modelled via the kinetic theory available in FLUENT.

(a)

(b)

Fig. 2

The CFD computational mesh of the cyclone. 23


solving transport equations for the individual Reynolds stresses, together with a transport equation for the dissipation rate. RSM has a greater potential to give accurate predictions for complex flows, as it takes into account the effects of streamline curvature, swirl, rotation, and rapid changes in strain rate in a more rigorous manner than two-equation models such as k-ε. The foundation of RSM is the exact set of transport equations:


The finite volume method has been used to discretise the partial differential equations of the model using the SIMPLE method for pressurevelocity coupling and the second order scheme to interpolate the variables on the surface of the control volume. The Reynolds stress (RSM) turbulence model was used in this model due to the anisotropic nature of the turbulence in cyclones. RSM model abandons the assumption of the isotropic eddy-viscosity hypothesis to close the Reynolds-averaged Navier-Stokes equations, by

(

)

(

[

)

)]

(

∂ ∂ ∂ ρui′u ′j + ρu k ui′u ′j = − ρui′u ′j u k′ + p δkj ui′ + δik u ′j ∂ ∂ ∂ t x x k k 14243 144244 3 14444442444444 3

Time derivative

+

Cij = Convection

DT,ij = Turblent diffusion

∂u j ⎤ ⎛ ∂u i ⎞ p ⎛⎜ ∂u i′ ∂u ′j ⎞⎟ ∂ ⎡ ∂ ⎟+ ui′u ′j ⎥ − ρ⎜⎜ ui′u k′ + u ′j u k′ + ⎢μ ∂xk ⎣ ∂xk ∂xk ∂xk ⎟⎠ ρ ⎜⎝ ∂x j ∂xi ⎟⎠ ⎦ ⎝ 144 42444 3 14444244443 1442443

( )

DL,ij = Molecular diffusion

Pij =Stress production

(

(1)

φij = Pr essure strain

)

∂u ′ ∂ui′ ′ εikm + ui′u m′ ε jkm − 2 ρ Ω k u ′j u m − 2μ i 3 ∂xk ∂xk 144444244444 14243 Fij = Pr oduction by system rotation εij = Dissipation

The dissipation tensor is assumed isotropic and is approximated by:

Of the various terms in these exact equations, Cij , DL, ij , Pij , and Fij do not require any modelling.

However, DT, ij , φij , and εij need to be modelled to

2μ

close the equations. The reason is simply because the averaging procedure of ui′u ′j u k′ will generate a lot

∂ ∂xk

⎡ μt ∂ui′u ′j ⎤ ⎢ ⎥ ⎢⎣ σ k ∂xk ⎥⎦

(2)

The pressure strain term is modelled as: p ⎛⎜ ∂ui′ ∂u ′j + ρ ⎜⎝ ∂x j ∂xi

(4)

Standard FLUENT wall functions, after Launder and Spalding (1974), were applied and second order discretisation schemes were also used. Under the RSM second order upwind scheme for discretisation there is a difficulty to reach convergence in simulation. The residuals may exhibit cyclic tendencies which mean that the transient pattern occurs. In this instance, the solver must be changed to a transient solver with a small time step something in the region of 0.025 seconds or a tiny fraction of the residence time of the cyclone. The simulation is then solved with a coupling of unsteady and steady state solvers in FLUENT. Having obtained the solution for the fluid flow, the particulate flow simulation has been performed by numerically tracing the motion of the particles in a Lagrangian frame of reference via the discrete phase model (DPM). The particle loading in a

of unknown variables and it becomes difficult to solve them directly. Turbulent diffusivity transport term is modelled as a simplified form of generalized gradient diffusion hypothesis as: DT, ij =

∂ui′ ∂ui′ 2 = δij ε ∂xk ∂xk 3

⎞ ⎟ = −C1 ε ⎡ui′u ′j − 2 δij k ⎤ − C 2 ⎡ Pij − 2 δij P ⎤ ⎥ ⎢ ⎥ ⎟ 3 3 k ⎢⎣ ⎦ ⎣ ⎦ ⎠

(3) where P = 0.5Pij and the constants are C1 = 1.8 and C2 = 0.6.

24


particle movement throughout the cyclone chamber. It should be noted that more than 100 particles were injected at the inlet surface for the calculation of collection efficiency.

sampling cyclone is typically small (3–5% volume fraction), and therefore, it can be safely assumed that the presence of the particles does not affect the flow field. The equation of motion for an individual particle can be written as follows (Crowe, Sommerfeld and Tsuji, 1998):


dv f (u − v ) + g = dt τ v

(5)

where the other contributions to the force on the particle (buoyancy, virtual mass and Basset term) are negligible because of the small fluid-to-particle density ratio. The response time of the particle, τv, is defined in terms of the particle density, particle diameter and the viscosity of the air as: τv =

ρ p d p2

(6)

18 μ g

The drag factor f is defined as: f =

C D Re r 24

(7)

where Re r =

ρg d p u − v μg

(8) Fig. 3

where Rer is the relative Reynolds number and CD is the drag coefficient. In FLUENT, the drag coefficient for spherical particles is calculated by using the correlations developed by Morsi and Alexander (1972). For non-spherical particles, the correlation was developed by Haider and Levenspiel (1989). The ordinary differential equation (Eq. (5)) was integrated along the trajectory of an individual particle. Collection efficiency statistics were obtained by releasing a specified number of monodispersed particles at the inlet of the cyclone and by monitoring the number escaping through the underflow. Collisions between particles and the walls of the cyclone were assumed to be perfectly elastic (coefficient of restitution is equal to 1). Effect of the instantaneous turbulent velocity fluctuations on the particle trajectories cannot be ignored especially in this case where the particle size ranges from 0.6 to 10 μm. Therefore, the stochastic tracking model was enabled in this simulation. Fig. 3 shows the particle trajectories of a single injection at the inlet surface. It shows clearly the effect of turbulent dispersion on the

Particle trajectories from a single injection point at cyclone inlet.

3. RESULTS AND DISCUSSIONS The predictions on cyclone performance were compared with the result from Bohnet (1995). The CFD predictions on cyclone performance are in good agreement with Bohnet's data within the operating domain investigated in this study (figures 4 to 6). In each case, the cyclone cut-off size and pressure drop were accurately predicted with an average deviation of 5% from the presented experimental data, which probably is in the same magnitude of the experimental error. The CFD prediction of cyclone grade efficiency operating from ambient to extreme temperature also shows a good agreement with Bohnet's data. However, there is some discrepancy on the prediction of grade efficiency due to the assumption that particles are only collected at the bottom or top outlet chamber and the particle deposition at the cyclone wall is not considered. Evidence of the particle collection through deposition on the cyclone wall has been reported previously by Yoshida et al. (2001). 25



Fig. 4

Prediction of cyclone pressure drop. Data points adopted from Bohnet (1995).

Fig. 5

Prediction of cyclone grade efficiency. Data points adopted from Bohnet (1995).

Fig. 6

Prediction of cyclone cut-off size at extreme temperature (T = 293 ~ 1073 K, Q = 60 m3/hr). Data points adopted from Bohnet (1995).

Fig. 7

Position of flow field data.

The experimental data of Bohnet (1995) have been successfully reproduced numerically via CFD modelling. The pressure drop and collection efficiency of the cyclone were successfully predicted numerically, implying that the cyclone flow field was also simulated correctly. In order to understand the mechanism that lies behind cyclones operating at extreme temperature, the flow field data at three different axial positions Z = 0.15, 0.2 and 0.25 m (Fig. 7) at the ambient temperature (293 K) and an extreme temperature (873 K) were extracted from the CFD simulation. There is a significant reduction (> 2 m/s) on the peak axial and tangential velocities as the operating temperature increases from ambient to 873 K as shown in figures 8 and 9. Such phenomena are attributed to the decreases in air density and increases of air viscosity at higher temperature. Since the centrifugal forces are proportional to the gas density, a reduction in the gas density can severely affect the velocity magnitude. It is also interesting to note that the magnitudes of tangential and axial velocities at the vortex core decrease with increasing temperature. This is because the reverse flow at the vortex core of the cyclone becomes weaker as the temperature increases. In Fig. 10, the contour of velocity magnitude inside the cyclone chamber at ambient and extreme temperatures clearly shows the evidence of a lower velocity magnitude at higher temperature. It shows that more than 2 m/s of velocity magnitude are reduced by increasing the operating temperature from ambient to 873 K.

26



Fig. 8

Profiles of axial velocity at various axial positions inside the cyclone chamber at ambient and extreme temperatures.

Fig. 9

Profiles of tangential velocity at various axial positions inside the cyclone chamber at ambient and extreme temperatures.

(a)

(b)

Fig. 10 Contour of velocity magnitude (m/s) inside the cyclone chamber at (a) extreme and (b) ambient temperatures. 27



Fig. 11 Profiles of static pressure at various axial positions inside the cyclone chamber at ambient and extreme temperatures.

Table 2 Prediction of air properties at extreme temperature. μ (N.s/m2)

ρ (kg/m3)

Cp (kJ/kg.K)

CFD Kinetic theory

3.90E-05

0.398

1.05

Incropera & De Witt (2002)

3.91E-05

0.399

1.12

modelling with the kinetic theory can successfully predict the air properties in highly swirling turbulent flow in the cyclone with average deviation of less than 5% from the published data. The CFD simulation presented in this work provides a better understanding on the influence of temperature on the cyclone hydrodynamics, and this might be helpful in design and troubleshooting of cyclones at extreme temperature.

A huge fall on the static pressure drop as shown in Fig. 11 is a direct result from a weaker reverse flow at the higher temperature. The pressure drop of a reverse flow cyclone is very much affected by the velocity magnitude especially the axial and tangential velocities. About 20% reduction on the peak axial and tangential velocities at the vortex core region can be observed from figures 8 and 9, and that explains the significant reduction on the cyclone pressure drop at a high operating temperature. The collection efficiency decreases as the temperature increases from 293 K to 1073 K for the same inlet velocity as observed in Fig. 5, and this trend is well predicted by the CFD simulation. This can be explained by the calculated velocities profiles, which show that the tangential velocity decreases as the temperature increases. As the centrifugal forces are proportional to the square of the tangential velocity, any reduction in tangential velocity can lead to a lower collection efficiency. The predicted air properties using the kinetic theory were compared with the published data from Incropera and De Witt (2002) to further confirm the accuracy of the CFD simulation. The comparison is presented in Table 2. It shows that the CFD

4. CONCLUSIONS The cyclone performance at extreme temperature can be modelled accurately via CFD simulation using the combination of RSM for turbulence modelling and kinetic theory for the temperaturedependent gas properties calculation. The flow mechanism inside a cyclone chamber operating at extreme temperature has also been successfully elucidated. The effect of temperature on the gas properties has also been accurately modelled by the kinetic theory. Therefore, the present numerical approach can be used as a very cost effective method of pre-evaluation of the performance of high temperature cyclones before a very costly experimental assessment is carried out. 28


11. Launder BE, Spalding DB (1974). The Numerical Computation of Turbulent Flows. Comput. Methods Appl. Mech. Engrg. 3(2):269–289. 12. Li E, Wang Y (1989). A new collection theory of cyclone separators. AIChE J. 35(4):666–669. 13. Morsi SA, Alexander AJ (1972). An investigation of particle trajectories in twophase flow systems. J. Fluid Mech. 55:193– 208. 14. Slack MD, Prasad RO, Bakker A, Boysan F (2000). Advances in cyclone modelling using unstructured grids. Chem. Eng. Res. Des. 78(8): 1098–1104. 15. Yoshida H, Fukui K, Yoshida K, Shinoda E (2001). Particle separation by Iinoya’s type gas cyclone. Powder Technol. 118(1–2):16–23.

ACKNOWLEDGEMENTS J. Gimbun would like to thank Assoc. Prof. Dr. Thomas SY Choong and Assoc. Prof. Dr. TG Chuah for initiating his strong interest to study the CFD of cyclones. J. Gimbun would also like to acknowledge the support from the UMP under the research grant RDU05/1/18.


REFERENCES 1. Altmeyer S, Mathieu V, Jullemier S, Contal P, Midoux N, Rode S, Leclerc J-P (2004). Comparison of different models of cyclone prediction performance for various operating conditions using a general software. Chem. Eng. Process 43(4):511–522. 2. Bohnet M (1995). Influence of the gas temperature on the separation efficiency of aerocyclones. Chem. Eng. Process 34(3):151– 156. 3. Boysan F, Ewan BCR, Swithenbank J, Ayers WH (1983). Experimental and theoretical studies of cyclone separator aerodynamics. IChemE Symp. Series, No 69, 305–320. 4. Crowe CT, Sommerfeld M, Tsuji Y (1998). Multiphase Flow with Droplets and Particles. CRC Press, Boca Raton. 5. Gimbun J, Choong TSY, Chuah TG, Fakhru’lRazi A (2005a). A CFD study on the prediction of cyclone collection efficiency. Int. J. Comput. Methods Eng. Sci. Mech. 6(3):161–168. 6. Gimbun J, Chuah TG, Fakhru’l-Razi A, Choong TSY (2005b). The influence of temperature and inlet velocity on cyclone pressure drop: a CFD study. Chem. Eng. Process 44(1):7–12. 7. Griffiths WD, Boysan F (1996). Computational fluid dynamics (CFD) and empirical modelling of the performance of a number of cyclone samplers. J. Aerosol Sci. 27(2):281–304. 8. Gupta AK, Lilley DG, Syred N (1984). Swirl Flows. Abacus Press, Tunbridge Wells. 9. Haider AM, Levenspiel O (1989). Drag coefficient and terminal velocity of spherical and nonspherical particles. Powder Technol. 58(1):63–70. 10. Incropera FP, DeWitt DP (2002). Introduction to Heat Transfer. 4th ed. John Wiley & Sons, USA.

29

Engineering Applications of Computational Fluid ...

Engineering Applications of Computational Fluid ...

Suggest Documents