ISSN 0104-6632

Printed in Brazil



Brazilian Journal
of Chemical
Engineering
Vol. 21, No. 01, pp. 93 - 101, January - March 2004


A NEW ROLE FOR REDUCTION IN PRESSURE
DROP IN CYCLONES USING COMPUTATIONAL
FLUID DYNAMICS TECHNIQUES

D. Noriler, A. A. Vegini
2
, C. Soares
1
, A. A. C. Barros
1
, H. F. Meier and M. Mori
2*

1
Laboratrio de Fluidodinmica Computacional. Departamento de Engenharia Qumica, Centro de Cincias Tecnolgicas,
Universidade Regional de Blumenau, Cx. P. 1507, CEP 89.010-971, Blumenau - SC, Brasil.
E-mail: [email protected], E-mail: [email protected]
2
Departamento de Processos Qumicos, Faculdade de Engenharia Qumica, UNICAMP,
Cx. P. 6066, CEP 13.081-970, Campinas - SP, Brasil.
E. mail: [email protected].

(Received: April 15, 2003 ; Accepted: July 11, 2003)

Abstract - In this work a new mechanical device to improve the gas flow in cyclones by pressure drop
reduction is presented and discussed. This behavior occurs due to the effects of introducing swirling
breakdown phenomenon at the inlet of the vortex finder tube. The device consists of a tube with two gas inlets
in an appositive spiral flux that produces a sudden reduction in the tangential velocity peak responsible for
practically 80 % of the pressure drop in cyclones. In turn, peak reduction causes a decrease in pressure drop
by a breakdown of the swirling, and because of this the solid particles tend to move faster toward the wall ,
increasing collection efficiency. As a result of this phenomenon the overall performance of cyclones is
improved. Numerical simulations with 3-D, transient, asymmetric and anisotropic turbulence closure by
differential Reynolds stress for Lapple and Stairmand standard geometries of 0.3 m in diameter, show a
reduction in pressure drop of 20 % and a shift of the tangential velocity peak toward the wall. All
numerical experiments were carried out with a commercial CFD code showing numerical stability and good
convergence rates with high-order interpolation schemes, SIMPLEC pressure-velocity coupling and other
numerical features.
Keywords: computational fluid dynamics (CFD), cyclones, pressure prop and turbulence.



INTRODUCTION

Cyclones have been widely used in chemical
processes since the beginning of the twentieth century
for downstream and upstream unit operations. In the
fluid catalytic cracking process (FCC process), for
example, cyclones are very important equipment to
guarantee performance and rentability. Currently,
researchers and engineers in the petroleum industries
are striving for an overall collection efficiency near
99.999 % in systems composed of a network of
primary and secondary cyclones. Pressure drop is also
very important for process savings. In the preheater
tower at a cement industry for example, a reduction in
pressure drop is fundamental for reduction in the cost
of cement where each kcal of energy saved per
kilogram of product plays an important role in the
success of the cement plant.
This broad application of cyclones is due to their
economical operation and maintenance and because
they can also be used under severe pressure and
temperature conditions for particles ranging from 1
to 100 microns in diameter. Despite this simplicity,
the fluid dynamics of swirling flow in cyclones is
very complex, with phenomena such as circulating
zones, high turbulence intensity, downflow,
reversion flow, high vorticity preservation, and
anisotropic behavior of the Reynolds stress, among

*To whom correspondence should be addressed








94 D. Noriler, A. A. Vegini, C. Soares, A. A. C. Barros, H. F. Meier and M. Mori


others. Engineers and scientists throughout the world
are dedicating more and more time to attempting to
understand the fluid dynamics of swirling turbulent
flow, making it possible to seek alternatives for
improving cyclone performance. Optimization of
geometry and optimal operation are the main aspects
observed in the current developments and research.
Dyakowski and Williams (1993), Cristea et al.
(1994), Meier and Mori (1999), and Meier et al.
(2002) developed numerical studies with experimental
data on bench and industrial scales with geometries
like Lapple and Stairmand for verification and
validation in CFD 3-D, by using transient and
turbulent models with the capability to represent the
main phenomena of swirling flow in gas cyclones.
The authors observed in all cases a poor
representation of conventional turbulence closures
using the isotropic behaviour of the Reynolds stress
in the standard k- form. Thus, it was clear that the
anisotropy of Reynolds stress has an important role
in predicting the tangential velocity profiles, and two
closures with two approaches were considered: the
hybrid model composed of a combination of the k-
model and the mixture length theory of Prandtl (k--
Prandtl model) in a symmetric 3-D physical domain
and a second-order closure like Reynolds stress
modelling (differential stress model - DSM ) in an
asymmetric 3-D domain. Both models and
approaches were tested with a large variation in
geometry and operational conditions showing good
agreement with experimental data from the literature.
The anisotropic symmetry approach is very similar
to the perfect mixture in chemical reactors with ideal
behaviour. On the contrary, the anisotropic
asymmetry approach shows deviation from ideal
behaviour. Of course, the second approach requires
more computational time and must be carefully used
by engineers in practical situations.
The main objective of this work is to apply the
computational fluid dynamics techniques in the
solution of a 3-D transient and asymmetric model
with Reynolds stress second-order turbulence
closure, in an attempt to analyse the new device for
reduction in pressure drop in cyclones.
This kind of methodology can be more useful than
traditional experimental studies. The cost of numerical
implementation is much lower than the cost of
building an experimental device, and different
geometric modifications can be analysed at the same
time. The purpose of this work is the implementation
and analysis of a mechanical device in order to study
the reduction in pressure drop in parallel with
interpretation of numerical results and scientific
visualization of the swirling flow in cyclones by using
a rigorous phenomelogical model coupled with
computational fluid dynamics techniques.



Brazilian Journal of Chemical Engineering




Figure 1: Geometric details of the device for reduction in pressure drop in cyclones.




A New Role for Reduction in Pressure Drop in Cyclones 95

The mechanical device is based on the principle
that 80% of the pressure drop is directly influenced
by tangential velocity peak. Therefore, the idea is to
decrease the velocity peak in the vortex finder of the
cyclone by using the fluid dynamics behavior of the
swirling flow, i.e.,., the main vortex is divided into
two opposite vortices in the finder by two inlets in a
spiral format which produces a shock between the
streamlines, thereby increasing the static pressure of
the system. Figure 1 shows in detail the format of
the device and its location in the vortex finder tube.
The mathematical modeling using in this work is
presented in the following section. It is also
important to point out that the model has been
previously submitted to extensive corroboration
studies (Meier et al., 2000).


MATHEMATICAL MODELLING

The transport equation, better known as the
Navier-Stokes equations used in this work to predict
the swirling, turbulent single-phase flow in cyclones,
can be written as follows:

Mass conservation equation

( )
. v 0
t

+ =

(1)

Momentum conservation equation,

( )
( ) ( )
v
. vv g . - v'v
t

+ = +

' (2)

where is the stress tensor and can be defined as

( )
T 2
pI v I v v
3
| |

= + +
|

\ .
.
(

(3)

where is the fluid density, v is the velocity field,
is the molecular viscosity, and I is the identity
tensor.
The last term on the right of Equation 2, v'v' , is
the time average of the dyadic product of fluctuation
velocities, and it is referred to as the Reynolds
turbulent stress.
The turbulence closure applied in this work is a
second-order closure based on the conservation
equations for each Reynolds stress component. The
model is well known as the differential stress model
(DSM) and for incompressible flows can be
expressed as

( )
( ) ( )
T
s
DS
v'v'
C k
. v'v'v . v'v' v'v'
t
2
P I.
3

(
+ =
(


+
(4)

P represents shear stress production and can be
expressed as

( ) ( )
T
P v'v' v v v'v'
( (
= +
(
(5)

and is a correlation for pressure force, which in
incompressible flow can be written as follows:

1 2
= + (6)

where

1 1S
2
C v'v'- kI
k 3
|
=

\ .
|
|
(7)

and

2 2S
2
C P P I
3
| |
=
|
\ .
. (8)

In this case, P is the trace of the tensor P, and C
S
,

DS
, C
1S
and C
2S
are constants of the model.
It is necessary to include an additional equation
for the rate of dissipation of kinetic energy that
appears in Equation (4).
( )
( )
( )
S
2
1 2
C k
. v . v'v'
t
C P C
k k

(
+ =
(



+
(9)

and k is obtained directly by its definition,

2
k 1/ 2 v' = (10)

In Table 1 the values of the DSM constants
applied in this work are presented.

Brazilian Journal of Chemical Engineering Vol. 21, No. 01, pp. 93 - 101, January - March 2004








96 D. Noriler, A. A. Vegini, C. Soares, A. A. C. Barros, H. F. Meier and M. Mori



Table 1: Constants of the DSM.

Constant DSM
C

0.115
C
1
1.440
C
2
1.920

k
1.000

1.000
C
S
0.220
C
1S
2.500
C
2S
0.550

DS
1.000



NUMERICAL METHODS

The numerical method applied to solve the model
was the finite volume method (MVF) with a
multiblock structure in a generalized co-ordinate
system for the numerical grid. The numerical
algorithm SIMPLEC was applied to the pressure-
velocity coupling, with a high-order interpolation
scheme (HIGHER UPWIND).
The Rhie Chow algorithm with the AMG
procedure (algebraic multi-grid) for solution of the
system of equations was used to prevent numerical
errors like check-boarding and zig-zag due to
the collocated grid. In this case, all variables are
calculated in the centre of each cell, and errors due to
the non-orthogonality of cells are generated during
the build of the numerical structure.
Details about the numerical method can be found
in Meier (1998), Meier and Mori (1998 e 1999),
Maliska (1995), Patankar (1980) and CFX4-4 User
Guide (2001).
Details about the numerical experiments
procedure for obtaining a stable solution with DSM
are available in Meier et al. (2000).

TEST CASES

The cyclones chosen for analyses of perturbation
caused by additional new devices are of two types,
Lapple and Stairmand. This choice was based on the
fact that a large quantity of data related to these types
of cyclones can be found in the literature.
Table 2 contains the geometric features of Lapple
and Stairmand cyclones.
Analysisof cyclones with a diameter of 0.3
meters (D
C
) was based on the operational conditions
and physical properties of the fluid, as described in
Table 3.


Table 2: Standard geometric features of
Lapple and Stairmand cyclones.

Cyclone

Lapple Stairmand
B
C
/D
C
0.250 0.200
D
O
/D
C
0.500 0.500
H
C
/D
C
0.500 0.500
L
C
/D
C
2.000 1.500
S
C
/D
C
0.620 0.500
Z
C
/D
C
2.000 2.500
D
U
/D
C
0.250 0.370



Brazilian Journal of Chemical Engineering








A New Role for Reduction in Pressure Drop in Cyclones 97


Table 3: Physical properties and operational
conditions of the test cases.

Physical Properties Operational Conditions
= 1.000 kg/m
3
V
in
= 15.000 m/s
= 1.800.10
-5
kg/m.s P

= 101325.000 Pa


RESULTS AND ANALYSIS

All numerical solutions were obtained using a PC
PENTIUM IV, 1.6 Ghz and 1.0 Gb RAM with CFX
4.4 from AEA Technology for Windows NT.
Figure 2 shows the numerical grid used in this
study, which offers a solution independent of the
number of cells in the grid.
The CPU time for each case was 72 hours using a grid
with about 40000 cells. To guarantee reproducibility of the
results, the numerical grids for all Lapple and Stairmand
cyclones were kept the same with or without the device.
Results obtained on pressure drop in all test cases
are compared in Figure 3. A substantial decrease in
the pressure drop (about 20%) is obtained when the
proposed device is implemented.







Figure 2: Numerical grid.


Brazilian Journal of Chemical Engineering Vol. 21, No. 01, pp. 93 - 101, January - March 2004








98 D. Noriler, A. A. Vegini, C. Soares, A. A. C. Barros, H. F. Meier and M. Mori


Lapple without
Device
Lapple with
Device
Stairmad without
Device
Stairmand with
Device
0
100
200
300
400
500
600
700
P
r
e
s
s
u
r
e

D
r
o
p

(
P
a
)




-22.7 %
-23.8 %

Figure 3: Pressure drop in cyclones.

Figures 4 and 5 illustrate the tangential velocity
profile at 15 cm below the device and at the exit of
finder, respectively.
In Figure 4 we can see that the negative effects of
the mechanical device cannot be observed in the
tangential velocity fields. A deviation in the
tangential velocity peak in the direction of the wall
was observed, but this represents a positive effect,
mainly for the efficiency analysis. It can be observed
in Figure 5 that the velocity peaks decrease in the
vortex finder region when the device is used. This
behaviour confirms the decrease in pressure drop.
Figure 6 illustrates the pressure maps for all test
cases, and it is possible to observe that the pressure
differential between the centre of the cyclone and the
peripheries decreases, the high-pressure region is
concentrated near the wall, and the low-pressure
region is concentrated exclusively in the vortex
finder region. Assuming the pressure difference as
the driving force for particle drag, it is possible that
the new device increases cyclone efficiency.
Figure 7 shows the tangential velocity fields, and
it is evident that the high-velocity regions are
transferred to the region near the wall, concentrating
the centrifugal field that is responsible for the
increase in collection efficiency. It is possible to note
that the swirling flow is symmetrical, explaining the
good results obtained when a 3D axisymmetric
modification, 2D domain is applied to the
representation of fluid dynamic patterns in cyclones.



Figure 4: Tangential Velocity 15 cm below the Device (Cylindrical Region).

Brazilian Journal of Chemical Engineering








A New Role for Reduction in Pressure Drop in Cyclones 99


Figure 5: Tangential velocity at the exit of the vortex finder tube.

Lapple with device

Lapple without device

Stairmand with device

Stairmand without device



Brazilian Journal of Chemical Engineering Vol. 21, No. 01, pp. 93 - 101, January - March 2004




Figure 6: Pressure maps for Lapple and Stairmand cyclones with and without the mechanical device.




100 D. Noriler, A. A. Vegini, C. Soares, A. A. C. Barros, H. F. Meier and M. Mori



Lapple with device Lapple without device
Stairmand with device Stairmand without device
Figure 7: Tangential velocity maps for Lapple and Stairmand
cyclones with and without the mechanical device.

CONCLUSIONS

Based on the results obtained in this work, it is
possible to conclude that implementation of the
mechanical device generates a decrease in the
pressure drop of about 20%, which causes a
reduction in energy consumption in this separation
process and has positive effects on collection
efficiency.
Although in this study the gas-solid system is not
analysed, there is evidence of an increase in
efficiency, since the tangential velocity peak is
transferred to the wall and the pressure field becomes
more homogeneous with the mechanical device,very
similar to the symmetric behaviour in cyclones.
An experimental study with the objective of
evaluating the geometric conditions and the
efficiency of the cyclone is suggested for
consolidation of this study. It is necessary to
comment that this mechanical device can be installed
in any cyclone, since it is merely an adaptation of
the vortex finder region .


NOMENCLATURE

Latin Letters

B
c
inlet section length, m

Brazilian Journal of Chemical Engineering




D
c
cyclone diameter, m




A New Role for Reduction in Pressure Drop in Cyclones 101

D
o
vortex finder diameter, m
D
u
underflow diameter, m
g gravitational acceleration vector, m/s
2

H
c
height section length, m
I identity tensor
k turbulent kinetic energy, m
2
/s
2

L
c
cylindrical section height, m
p pressure, Pa
P mean shear stress production, Pa/s
P trace of the tensor P, Pa/s
S
c
vortex finder height, m
t time, s
v velocity vector, m/s
Z
c
conical section height, m

Greek Letters

rate of dissipation of turbulent energy, m
2
/s
3

pressure force correlation, Pa
viscosity, kg/ms
density, kg/m
3

normal stress tensor, Pa


REFERENCES

CFX4 for Windows NT. CFX4-4 User Guide, AEA
Technology. Oxfordshire, United Kingdom(2001).
Cristea, E. D., Malfa, E., and Coghe, A., 3-D
Simulation and Experiments of Cement Rotary
Kiln Preheater Top Cyclone. Proceedings of
Fluent European Users Group Meeting,
Harrogate, U. K., C54 (1994).
Dyakowski, T. and Williams, R.A., Modelling
Turbulent Flow in a Hydrocyclone Operating
without an Air Core. Chem. Eng. Sci., vol 48, n
6, pp. 119-128 (1993).
Maliska, C.R., Transferncia de Calor e Mecnica
dos Fluidos Computacional. CTC Editora, Rio de
Janeiro, RJ (1995).
Meier, H.F., Kasper, F.R.S., Peres, A.P., Huziwara,
W.K., and Mori, M., Comparison between
Turbulence Models for 3-D Turbulent Flows in
Cyclones In: 21
st
Iberian Latin American Congress
on Computational Methods in Engineering, Rio de
Janeiro. CILANCE 2000. PUC RIO, vol.1( 2000).
Meier, H.F. and Mori, M., Anisotropic Behavior of
the Reynolds Stress in Gas and Gas-Solid Flows
in Cyclones. Powder Technology, vol.101,
pp.108-119 (1999).
Meier, H.F., and Mori, M. Gas-Solid Flow in
Cyclones: The Eulerian-Eulerian Approach.
Computers and Chemical Engineering., vol.22,
pp.64 -644 (1998).
Meier, H.F., Ropelato, K., Forster, H., Iess, J J.,
Mori, M., Computational Fluid Dynamics (CFD)
zur Berechnung und Auslegung von Zyklonen -
Teil 1. ZKG International,vol.55, n 04, pp.6-75
(2002a).
Meier, H.F., Ropelato, K., Forster, H., Iess, J J.,
Mori, M.,Computational Fluid Dynamics (CFD)
zur Berechnung und Auslegung von Zyklonen -
Teil 2. ZKG International - Cement-Lime-
Gypsum. vol.55, n.06, pp.5 -64 (2002b).
Patankar, S.V., Numerical Heat Transfer and Fluid
Flow. Ed. Hemisphere Pub. Co., New York
(1980).


Brazilian Journal of Chemical Engineering Vol. 21, No. 01, pp. 93 - 101, January - March 2004