International Journal for Computational Methods in Engineering Science and Mechanics, 6(3): 161168, 2005

c Taylor & Francis Inc.

Copyright

ISSN: 15502287 print / 15502295 online
DOI: 10.1080/15502280590923649

A CFD Study on the Prediction of Cyclone

Collection Efficiency
Jolius Gimbun1, Thomas S. Y. Choong2, T. G. Chuah2, and A. Fakhrul-Razi2
1

Faculty of Chemical and Natural Resources Engineering, University College of Engineering and
Technology Malaysia, KUKTEM, MEC Town, 25200 Kuantan, Pahang D. M., Malaysia.
Department of Chemical and Environmental Engineering, Faculty of Engineering, Universiti
Putra Malaysia, Selangor D. E., Malaysia.

10

15

20

This work presents a Computational Fluid Dynamics calculation to predict and to evaluate the effects of temperature, operating pressure and inlet velocity on the collection efficiency of gas
cyclones. The numerical solutions were carried out using spreadsheet and commercial CFD code FLUENT 6.0. This paper also
reviews four empirical models for the prediction of cyclone collection efficiency, namely Lapple [1], Koch and Licht [2], Li and
Wang [3], and Iozia and Leith [4]. All the predictions proved to be
satisfactory when compared with the presented experimental data.
The CFD simulations predict the cyclone cut-off size for all operating conditions with a deviation of 3.7% from the experimental
data. Specifically, results obtained from the computer modelling
exercise have demonstrated that CFD model is the best method of
modelling the cyclones collection efficiency.
Keywords

25

30

35

Cyclone, CFD, Efficiency, Temperature, Inlet Velocity,

Cut-Off Size

1. INTRODUCTION
Cyclones are devices that employ a centrifugal force generated by a spinning gas stream to separate particles from the
carrier gas. Their simple design, low capital cost and nearly
maintenance-free operation make them ideal for use as precleaners 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 component materials.
Cyclone collection efficiencies can reach 99% for particles bigger than 5 m [5], and can be operated at very high dust loading.
Cyclones are used for the removal of large particles for both air
pollution control and process use. Application in extreme con-

Received 6 January 2004; in accepted 25 May 2004.

Address correspondence to Jolius Gimbun, Faculty of Chemical
& Natural Resources Engineering, University College of Engineering
& Technology Malaysia, MEC Town, 25200 Kuantan, Pahang D.M.,
Malaysia. E-mail: jolius@kuktem.edu.my

ditions includes the removal of coal dust in a power plant and

the use as a spray dryer or gasification reactor.
Engineers are generally interested in two parameters in order
to carry out an assessment of the design and performance of a cyclone. These parameters are the collection efficiency of particle
and pressure drop through the cyclone. An accurate prediction
of cyclone efficiency is very important because an inaccuracy in
the efficiency prediction may result in an inefficient design of
the cyclone separator. CFD has a great potential to predict the
flow field characteristics and particle trajectories inside the cyclone as well as the pressure drop [6]. 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 pressure drop.
This study presents an application of computational fluid dynamics, in the prediction of cyclone efficiency. This study also
reviews the prediction of four different empirical models for cyclone efficiency, namely Lapple [1], Koch and Licht [2], Li and
Wang [3], and Iozia and Leith [4]. The simulation results are
then compared with experimental data found in the literature
for different inlet flow rates, pressures and temperatures. In this
study, the CFD calculations are carried out using a commercial
finite volume code, FLUENT 6.0, and the empirical models are
performed in Microsoft Excel spreadsheet.
2. CYCLONE DESIGN
Many different types of cyclones have been built but the reverse flow cyclone with tangential inlet in Fig. 1 is most often
used for industrial gas cleaning [3, 7]. In this study, the numerical
simulation deals with the standard case of reverse flow cyclone
with a tangential rectangular inlet. Cyclone dimensions used in
this simulation are as shown in Table 1.
3. COMPUTATIONAL FLUID DYNAMICS APPROACH
FLUENT is a commercially available CFD code that utilizes
the finite volume formulation to carry out coupled or segregated
161

40

45

50

55

60

65

70

162

J. GIMBUN ET AL.

FIG. 1. Tangential cyclone configuration.

75

80

85

90

95

calculations (with reference to the conservation of mass, momentum and energy equations). It is ideally suited for incompressible to mildly compressible flows. The conservation of
mass, momentum and energy in a fluid flow are expressed in
terms of non-linear partial differential equations that generally
defy solution by analytical means. The solution of these equations has been made possible by the advent of powerful workstations, opening avenues towards the calculation of complicated
flow fields with relative ease.
For the turbulent flow in a cyclone the key to the success of
CFD lies with the accurate description of the turbulent behavior
of the flow [6]. To model the swirling turbulent flow in a cyclone
separator, there are a number of turbulence models available in
FLUENT. These range from the standard k-
model to the more
complicated Reynolds stress model (RSM). The comparison of
the different RANS-based turbulence models available in FLUENT 6.0 is presented in Table 2. The k-
model involves the solution of transport equations for the kinetic energy of turbulence
and its dissipation rate and the calculation of a turbulent contribution to the viscosity at each computational cell. The standard
k-
, RNG k-
and Realizable k-
models were not optimized for
the strongly swirling flows typically found in cyclones [8, 9].
Turbulence may be stabilized or destabilized in the parts of flow
domain where strong streamline curvature is present. However,
to reduce the computational effort, the RNG k-
model can be
used with about 12% deviation on experimental data [6]. The
numerical studies carried out by Fredriksson [10] reveal that the
TABLE 1
Cyclone geometry used in this simulation
Geometry

a/D b/D De /D S/D h/D H/D B/D

Stairmand High
0.5 0.2
0.5
0.5 1.5
4
0.375
Efficiency
Kim and Lee
0.33 0.225 0.257 1.157 1.447 3.05 0.482
(1990) cyclone I
Bohnet (1995)
0.533 0.133 0.333 0.733 0.693 2.58 0.333

RNG k-
model underestimates the variation of the axial velocity profile across the radial direction and also overestimates the
magnitude of the tangential velocity and the cyclone pressure
drop.
The Reynolds stress model requires the solution of transport
equations for each of the Reynolds stress components as well
as for dissipation transport without the necessity to calculate an
isotropic turbulent viscosity field. The Reynolds Stress turbulence model yields an accurate prediction of swirl flow pattern,
axial velocity, tangential velocity and pressure drop on cyclone
simulations [810].
The finite volume method has been used to discretize the
partial differential equations of the model using the SIMPLE
method for pressure-velocity coupling and the Second Order
Upwind scheme to interpolate the variables on the surface of
the control volume. The segregated solution algorithm was selected. The Reynolds stress (RSM) turbulence model was used
in this model due to the anisotropic nature of the turbulence in
cyclones. Standard Fluent wall functions were applied and high
order discretization schemes were also used.
Under the RSM second order upwind for discretization there
is a difficulty to reach the convergence in simulation [11].
The residuals may exhibit cyclic tendencies, which means that
the transient pattern occurs. In this instance, the solver must
be changed to a transient solver and this makes the 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.
For the simulation using RNG k-
model the steady state solver
is sufficient to reach convergence.
To calculate the trajectories of particles in the flow, the discrete phase model (DPM) was used to track individual particles
through the continuum fluid. The particle loading in a cyclone
separator is typically small (35%), and therefore it can be safely
assumed that the presence of the particles does not affect the
flow field (one-way coupling). The equation of motion for an
individual particle can be written as Crowe et al., [12]
dv
f
= (u v) + g
dt
v

100

105

110

115

120

125

130

[1]

where the other contributions to the force on the particle (buoy- 135
ancy, virtual mass and Basset term) are negligible because of
the small fluid-to-particle density ratio. The response time of
the particle, v is defined in terms of the particle density, particle
diameter and the viscosity of the air as:

v =

p d 2p
18

[2]

The drag factor f is defined as:

f =

C D Rer
24

140

[3]

A CFD STUDY ON THE PREDICTION OF CYCLONE

163

TABLE 2
Comparison of the different turbulence models in FLUENT 6.0
Model
Standard k-

RNG k-

Realizable k-
RSM

Strength

Weaknesses

Robust, economical and reasonably accurate

Many sub-models available, i.e. combustion,
buoyancy, compressibility, etc.
Good for moderately complex behavior like jet
impingement, separating flows, swirling flows, and
secondary flows.
Offers largely the same benefits as RNG; resolves
round jet anomaly.
Physically most complete model (history, transport and
anisotropy of turbulent stresses are all accounted
for). Most suitable for complex 3D flows with strong
streamline curvature, swirl and rotation.

with

Mediocre results for complex flow involving severe

pressure gradients, strong streamline curvature,
swirl and rotation.
Subjected to limitations due to isotropic eddy
viscosity assumption.
Subjected to limitations due to isotropic eddy
viscosity assumption.
Requires more CPU effort (23 times); limited near
wall modelling options; tightly coupled momentum
and turbulence equations.

Pentium IV 2.8 GHz HP workstation XW8000 with 512 cachememory, 1 GB RAM-memory, and 110 GB hard-disc memory. 160
Rer =

g d p |u v|
g

[4]

where Rer is the relative Reynolds number and C D is the drag coefficient. In FLUENT, the drag coefficient for spherical particles
is calculated by using the correlations developed by Morsi and
145 Alexander [13]. For non-spherical particles, the correlation was
developed by Haider and Levenspiel [14]. The ordinary differential equation (Eq. (1)) was integrated along the trajectory of an
individual particle. Collection efficiency statistics were obtained
by releasing a specified number of monodispersed particles at
150 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).
The numerical calculation was made with a fine numerical
155 grid as shown in Fig. 2. The numerical grid of cyclone A, B
and C contains 28871, 33056, and 18045 nodes respectively, to
yield a reasonable prediction. The details of the CFD setting are
presented in Table 3. The CFD simulation was performed with a

FIG. 2.

4. CYCLONE EFFICIENCY EMPIRICAL MODELS

4.1. Iozia and Leith Model
Iozia and Leith [4] logistic model is a modified version of
Barth [15] Model, which is developed based on force balance.
The model assumes that a particle carried by the vortex endures 165
the influence of two forces: a centrifugal force, Z and a flow
resistance, W . The collection efficiency i of particle diameter
d pi can be calculated from
1
[5]
i =
1 + (d pc /d pi )
is an expression for slope parameter derived based on the
statistical analysis of experimental data of a cyclone with D = 170
0.25 m given as

2





d pc
ab
ab
+
1.05
ln
+ 5.21 ln
= 0.62 0.87 ln
100
D2
D2
[6]

CFD surface mesh of cyclone used in the simulations.

164

J. GIMBUN ET AL.

TABLE 3
Detail on CFD setting
Boundary condition
Inlet
Outlet
Cyclone wall
Viscous
Turbulence

Velocity inlet
Outflow
Standard wall function
Reynolds stress model (RSM)
RNG k-

Discretization
Pressure
Pressure-velocity coupling
Momentum
Turbulence kinetic energy
Turbulence dissipation rate
Reynolds stresses
Discrete phase modelling
Assumption
Maximum number of step
(phase integration)

d pc =

Presto!
SIMPLE
2nd order upwind
2nd order upwind
2nd order upwind
2nd order upwind

20000

i = 1 exp{1 }

0.25

De
D

1 = 2(S + L)/a

0.5

185

[9]

The radial particle velocity and the radial concentration

profile are not constant for uncollected particles within
the cyclone.
Boundary conditions with the consideration of turbulent diffusion coefficient and particle bounce reentrainment on the cyclone wall are:
c = c0 , at = 0
c
Dr
= (1 )wc, at r = D/2
r

[15]

[16] 195

4.3. Koch and Licht Model

Koch and Licht [2] collection theory recognized the inherently turbulent nature of cyclones and the distribution of gas
residence times within the cyclone. Koch and Licht described
particle motion in the entry and collection regions with the ad- 200
ditional following assumptions:

1.4

4.2. Li and Wang Model

The Li and Wang [3] model includes particle bounce or reentrainment and turbulent diffusion at the cyclone wall. A twodimensional analytical expression of particle distribution in the
cyclone is obtained. Li and Wang model was developed based
180 on the following assumptions:

[14]

[7]

175

[13]

where

ab
dc = 0.47D
D2

(1 )K ww
Dr rwn

190

The resultant expression of the collection efficiency for particle

of any size is given as


HS
[(dc /B) 1] for dc > B [8a]
(D/B) 1
for dc < B
[8b]

zc = H S

(1 n)( p g )d 2 Q


18 b rw1n rn1n

Spherical particle

9 Q
p z c vt2max

[12]

where

and

where core length, z c , and core diameter, dc , are given as

z c = (H S)

The concentration distribution in a cyclone is given as:

 

1
c0 (rw rn ) exp K (1+n)
r 1+n


c(r, ) =
 rw
1
1+n dr
rn exp K (1+n) r
K =

and d pc is the 50% cut size given by Barth [15]



The tangential velocity is related to the radius of cyclone by: u R n = constant.

[10]
[11]

The tangential velocity of a particle is equal to the tangential velocity of the gas flow, i.e. there is no slip in the
tangential direction between the particle and the gas.
The tangential velocity is related to the radius of cyclone by: u R n = constant.

205

A force balance and an equation on the particles collection yields

the grade efficiency i

0.5/(n+1) 
Gi Q
i = 1 exp 2
(n + 1)
D3


[17]

where
8K c
K a2 K b2



(12D)0.14
T + 460 0.3
n = 1 1
2.5
530
2
p d pi
i =
18

G=

[18]
[19]
[20]

A CFD STUDY ON THE PREDICTION OF CYCLONE

165

FIG. 3. Particle trajectories from CFD simulation of different particle size in the Bohnet cyclone at T = 1073 K.

210

G is a factor related to the configuration of the cyclone, n is

related to the vortex and is the relaxation term.

4.4. Lapple Model

Lapple [1] model was developed based on force balance without considering the flow resistance. Lapple assumed that a par215 ticle entering the cyclone is evenly distributed across the inlet
opening. The particle that travels from inlet half width to the
wall in the cyclone is collected with 50% efficiency. The semi
empirical relationship developed by Lapple [1] to calculate a
50% cut diameter, d pc , is

d pc =
220

9 b
2 Ne vi ( p g )

 12

[21]

where Ne is the number of revolutions

Ne =



1
H h
h+
a
2

[22]

The efficiency of collection of any size of particle is given by

i =

1
1 + (d pc /d pi )2

[23]

5. RESULT AND DISCUSSION

5.1. Grade Efficiency Prediction under Ambient
Temperature and Pressure
An accurate prediction of cyclone efficiency under ambi- 225
ent temperature and pressure is important since there are a lot
of applications of cyclone under these conditions. Application
of the cyclone under room temperature includes the removal
of sawdust, grain dust and rock dust. Kim and Lee [16] and
Dirgo and Leith [17] presented experimental data obtained at 230
room temperature. The calculated trajectories of 1, 2, 2.5 and
6 m particles in the Bohnet cyclone are shown in Fig. 3.
While, the CFD flow field simulation on Bohnet cyclone is
presented in Fig. 4. The comparisons between the presented

FIG. 4. CFD flow field simulation on Bohnet cyclone (vi = 8 m/s, T = 293 K).

166

235

240

245

250

255

J. GIMBUN ET AL.

FIG. 5. Calculated and measured collection efficiencies for Kim and Lee [16]
cyclone (P = 1 Bar, T = 293 K, vi = 4.25 m/s, D = 0.311 m). Data point
from Kim and Lee (1990).

FIG. 7. Calculated and measured collection efficiencies for Stairmand high

efficiency cyclone (P = 1 Bar, T = 293 K, vi = 5 m/s, D = 0.305 m). Data
point from Dirgo and Leith [17].

experimental data, empirical models and CFD prediction are

shown in Figs. 5 to 7.
The Li and Wang empirical model prediction is found to
agree much better with the data from Kim and Lee, and Dirgo
and Leith, compared to the other models developed by Koch and
Licht, Iozia and Leith, and Lapple (Figs. 5 to 7). Lapples model
yields less accurate fitting to the experimental data (curves are
flatter at higher particle size), as does the Koch and Licht model.
Both models considerably underestimate efficiency for large particles and overestimate efficiency for small particles. The Lapple
model is unable to fit well with any experimental data. This is
possibly because the Lapple model simply assumes that particles that enter the cyclone are evenly distributed across the inlet
opening and a particle that travels from the inlet half width to
the cyclone wall is collected with 50% efficiency. Unjustified assumptions of complete and uniform mixing of uncollected dust
at any height in the cyclones may also contribute to the discrepancy between the experimental data and the Koch and Licht
predictions. Mothes and Loffler [18] experimental findings further support the fact that there is indeed a concentration gradient
in the radial direction of the cyclones.
Iozia and Leith logistic model predicted the efficiency satisfactory for cyclone of diameter 0.305 m as shown in Fig. 6 and
7. For smaller cyclone diameters, the prediction of the Iozia and
Leith model is not satisfactory. It considerably overestimates

the grade efficiency for D = 0.0311 m, as shown in Fig. 5. The

reason for this disagreement may be caused by the generalized
form of core length, z c in the Iozia and Leith model, which is developed based on the statistical analysis of experimental cyclone
data from cyclone of D = 0.25 m. Therefore, the prediction of
the model is only satisfactory for cyclone diameter around this
range.
The CFD simulations yielded very good predictions on cyclone collection efficiency under ambient temperature and pressure operating condition, as shown in Figs. 5 to 7. The accuracy of the CFD prediction on cyclone collection efficiency is
comparable to the Li and Wang model in all types and size of
cyclones evaluated in this study. There is a slight discrepancy
on the CFD prediction as shown in Fig. 5. However, the CFD
result still yielded an accurate prediction on cut size diameter,
D pc , of each cyclone under ambient temperature and pressure
condition (Table 4).

FIG. 6. Calculated and measured collection efficiencies for Stairmand high

efficiency cyclone (P = 1 Bar, T = 293 K, vi = 15 m/s, D = 0.305 m). Data
point from Dirgo and Leith [17].

FIG. 8. Calculated and measured collection efficiencies for Stairmand high

efficiency cyclone (P = 1.7 Bar, T = 293 K, vi = 11 m/s, D = 0.4 m). Data
point from Ray et al. [19].

260

265

270

275

5.2. Grade Efficiency Prediction under Different

Operating Conditions
Ray et al. [19] and Bohnet [20] have done an experiment
under high temperature and pressure operating conditions. The
comparison between the experimental data, CFD and the four 280
selected empirical model predictions is shown in Figs. 8 and 9.
The prediction of the Li and Wang model under high pressure

A CFD STUDY ON THE PREDICTION OF CYCLONE

167

TABLE 4
Comparison of measured and predicted cut-off size of different cyclones
Models
Cyclone type and experiment value
Kim and Lee [16]
Dirgo and Leith [17] 5 m/s
Dirgo and Leith [17] 15 m/s
Ray et al. [19]
Bohnet [20] 873 K
Bohnet [20] 1073 K
Average deviation (%)

2.86
6.24
3.06
2.61
2.52
3.12
0

CFD

Li and Wang

Iozia and Leith

Koch and Licht

Lapple

2.91
6.14
3.27
2.54
2.75
3.12
3.67

3.05
5.91
3.06
2.67
3.38
3.83
11.85

1.7
6.73
3.34
2.84
1.85
1.96
21.69

0.82
4.72
2.43
2.46
1.54
1.91
33.28

2.52
8.22
4.19
3.57
2.48
2.48
23.24

operating conditions is good compared to the experimental data

as shown in Fig. 8. CFD results and the Iozia and Leith model
285 also yield a reasonably good prediction on cyclone efficiency
under this operating condition.
The data presented by Bohnet [20] concerns experiments at
temperatures above 1000 K. It appears that the CFD code shows
good prediction of cyclone efficiency under extremely high tem290 peratures, as shown in Fig. 9. The model of Dirgo is found to
overestimate the cyclone collection efficiency under the high
temperature operating condition (Fig. 9). The models of Koch
and Licht, and Lapple still show a reasonably good prediction
under this extreme condition. Meanwhile, Li and Wang model is
295 found to underestimate the cyclone collection efficiency under
the extreme operating temperatures.
5.3. Cut-Off Size Prediction
Cyclones have been characterized by a cut size (d50 ), which
defines the particle size for which the cyclone collection effi300 ciency is 50%. It is important to know the cyclone cut-off diameter under certain operational conditions and geometry. The
comparison between the experimental data, CFD and the four
selected empirical models prediction is shown in Table 4.
The simulation results obtained from the computer modelling
305 exercise have demonstrated that CFD code is the best method of
modelling the cyclones cut-off size with the average deviation of

3.7% to the measured value. The Li and Wang, Lapple, Iozia and
Leith, and Koch and Licht models were found to be inconsistent
in the cut-off size prediction with the deviation ranging from 310
11.9 to 33.3% from the measured value.
6. CONCLUSIONS
The Li and Wang model and CFD code both predict very
well the cyclone efficiency and cut-off size for any operational
conditions. The Li and Wang model and FLUENT CFD code
produce a better fit to the Ray, Dirgo and Leith, and Kim and
Lee experimental data respectively. In all operating conditions
and cyclone types the FLUENT CFD and Li and Wang model
were found to be much closer to the experimental measurement.
However, only the FLUENT CFD code is consistently predicts
the cyclone cut-off size. Therefore, both the Li and Wang model
and FLUENT CFD code can be used to evaluate the collection
efficiency in the cyclone design except for the extreme operating temperatures, which is Li and Wang model is less accurate.
The Lapple and Koch and Lich models considerably underestimate the efficiency for large particles and overestimate efficiency
for small particles. Iozia and Leith logistic model show a good
agreement with an experimental data for the cyclone size range
of D = 0.250.4 m, but it is unable to predict correctly the efficiency for small cyclone (D < 0.1 m). Iozia and Leith model
is only suitable for efficiency prediction of cyclone diameter
around 0.25 m.

FIG. 9. Separation efficiency of Bohnet (1995) cyclone at high temperature (P = 1 Bar, vi = 8.61 m/s, D = 0.15 m). Data point from Bohnet [20].

315

320

325

330

168

ACKNOWLEDGEMENT
The authors would like to thank Dr. Tom Fraser, Fluent India
and Fluent Europe UK for their guidance and support.

u, v
Rer
CD
RANS

NOMENCLATURE

Greek Letters
375
v
particle response time (s)
g
gas viscosity (m2 /s)

slope parameter

relaxation time (s)

380

angular coordinate

particle bounce or re-entrainment coefficient

characteristic value
grade efficiency of particle size at mid-point of ini
ternal i (%)
g
gas density (kg/m3 )
385
p
particle mass density (kg/m3 )

L
a
b
D
De
340 H
h
S
B
c0 , c1
345 d p
Dr
d pc
335

n
Q
r
R
T
w
355 wn , ww
d pi
g
G
i
360 K a
Kb
Kc
Ne
350

365

J. GIMBUN ET AL.

vi
K

zc
dc
370 vt max

natural length (m)

cyclone inlet height (m)
cyclone inlet width (m)
cyclone body diameter (m)
cyclone gas outlet diameter (m)
cyclone height (m)
cyclone cylinder height (m)
cyclone gas outlet duct length (m)
cyclone dust outlet diameter (m)
particle inlet and outlet concentration (kg/m3 )
particle diameter (m)
radial turbulent diffusion coefficient
cut particle diameter collected with 50% efficiency
(m)
cyclone vortex exponent (0.5 < n < 1)
volumetric gas flow rate (m3 /s)
radial dimension, rw = D/2 and rn = De /2(m)
radius (m)
absolute temperature (K)
radial particle velocity (rad/s)
radial particle velocity at r = rn and r = rw (rad/s)
diameter of particle in size range i(m)
gravity acceleration (m/s2 )
cyclone configuration factor
subscript donates interval n particles size range
a/D
b/D
cyclone volume constant
number of revolutions Ne of gas spins through a in
the outer vortex
inlet velocity (m/s)
cyclone configuration and operating condition constant
core length (m)
core diameter (m)
maximum tangential velocity (m/s)

Velocity magnitude (m/s)

relative Reynolds number
drag coefficient
Reynolds Average Navier Stokes

REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.

11.
12.
13.
14.
15.
16.
17.
18.
19.
20.

Lapple, C. E., Chem. Eng. 58, 144 (1951).

Koch, W. H., and Licht, W., Chem. Eng., 7, 80 (1977).
Li Enliang, and Wang Yingmin, A.I.Ch.E. J. 35, 666 (1989).
Iozia, D. L., and Leith, D., Aerosol Sci. Technol. 12, 598 (1990).
Silva, P. D., Briens, C., Bernis, A., Powder Technol. 131, 111 (2003).
Griffiths, W. D., and Boysan, F., J. Aerosol Sci. 27, 281 (1996).
Altmeyer, S., Mathieu, V., Jullemier, S., Contal, P., Midoux, N., Rode, S.,
and Leclerc, J.-P., Chem. Eng. Process 43, 511 (2004).
Reddy, M., Fluent India, Personal Communication, mxr@fluent.co.in
(2003).
Fraser, T., Personal Communication, fraserta1@cf.ac.uk, www.cfdonline.com (2003).
Fredriksson, C., Exploratory experimental and theoretical studies of cyclone gasification of wood powder. Doctoral thesis, Lulea University of
Technology, Sweden (1999).
Gimbun, J., Chuah, T. G., Fakhrul-Razi, A., and Thomas S. Y. Choong,
Chem. Eng. Process (2004) (in press).
Crowe, C. T., Sommerfeld, M., and Tsuji, Y., Multiphase Flow with Droplets
and Particles. CRC Press, Boca Raton (1998).
Morsi, S. A., and Alexander, A. J., J. Fluid Mech. 55, 193 (1972).
Haider, A., and Levenspiel, O., Powder Technol. 58, 63 (1989).
Barth, W., Brennstoff-Warme-Kraft, 8, 1 (1956).
Kim, J. C., and Lee, K. W., Aerosol Sci. Technol. 12, 1003 (1990).
Dirgo, J., and Leith, D., Aerosol Sci. Technol. 4, 401 (1985).
Mothes, H., and Loffler, F., J. Aerosol Sci. 13, 184 (1982).
Ray, M. B., Hoffmann, A. C., and Postma, R. S., J. Aerosol Sci. 31, 563
(2000).
Bohnet, M., Chem. Eng. Process 34, 151 (1995).

390

395

400

405

410

415