Sunteți pe pagina 1din 11

PTEC-14831; No of Pages 11

Powder Technology xxx (2019) xxx

Contents lists available at ScienceDirect

Powder Technology

journal homepage: www.elsevier.com/locate/powtec

High-efficiency industrial cyclone separator: A CFD study


Dzmitry Misiulia a,⁎, Sergiy Antonyuk b, Anders Gustav Andersson c, Tord Staffan Lundström c
a
Department of Machines and Apparatus for Chemical and Silicate Production, Belarusian State Technological University, 13a Sverdlova str., 220006 Minsk, Belarus
b
Institute of Particle Process Engineering, University of Kaiserslautern, D-67653 Kaiserslautern, Germany
c
Division of Fluid and Experimental Mechanics, Department of Engineering Sciences and Mathematics, Luleå University of Technology, SE-971 87 Luleå, Sweden

a r t i c l e i n f o a b s t r a c t

Article history: The flow within an industrial scroll-inlet high-efficiency cyclone separator has been studied using RSM and LES
Received 20 January 2019 simulations. Of particular interest is the effect of the gas outlet configuration, i.e. outlet scroll and radial bend, on
Received in revised form 10 September 2019 the flow pattern, pressure drop and cyclone efficiency. A surprising phenomenon is that the inner vortex splits
Accepted 15 October 2019
into two vortices for the cyclone with a conventional outlet pipe while if the cyclone is equipped with an outlet
Available online xxxx
scroll or radial bend there is no split. The outlet scroll and radial bend increase the pressure losses by 5.1% and
Keywords:
6.4%, respectively. These installations, moreover, significantly destabilize the pressure losses and the amplitude
Cyclone separator of instantaneous pressure drop oscillations increases from 0.65% to 16.2% and 33.96%, respectively. The investi-
Computational fluid dynamics gated outlet scroll and radial bend have practically no effects on the cyclone efficiency since the flow in the
Vortex breakdown main separation zone is not affected by the gas outlet configuration.
Pressure drop © 2019 Elsevier B.V. All rights reserved.
Collection efficiency

1. Introduction scroll, the gas would undergo an abrupt expansion with an instanta-
neous loss of kinetic energy. Thus, a well-designed outlet scroll will
Gas cyclones are the most popular tools for the separation of solid imply a pressure recovery and this is the main reason of installing it. A
particles from gases in industrial applications. Their popularity is due small pressure drop reduction can also be reached with a radial bend at
to their simplicity, reliability, relatively low manufacturing costs and the outlet [4] but this is not its main purpose.
ability to work under high pressures and temperatures. Several experimental and numerical works on the cyclone aerody-
The usage of high-efficiency cyclone separators is increasing espe- namics and its performance with different gas outlet configurations
cially those that function without bag filters due to very high separation have been carried out. Muschelknautz (as it is quoted in [5]) investi-
capability, i.e. ACS Hurricane cyclones and ReCyclone systems [1,2]. gated the effect of a number of pressure recovery configurations includ-
Gas cyclones are used as a pre-separator of a two or three-stage sys- ing an outlet scroll, and found that the outlet scroll reduced the pressure
tem, as the only separation stage or as the final separation stage. In the drop by 12% as compared to the pressure drop with a conventional cy-
first case, when two or more cyclones are connected in series, an outlet lindrical, sharp-edged vortex finder.
scroll is normally installed directly after and on top of the first stage cy- Funk [3] experimentally investigated the pressure drop of a cyclone
clone. Sometimes, due to practical reasons, a radial bend is attached to with rectangular and radial evasés at the outlet. One conclusion was
the gas outlet pipe (vortex finder) instead of a scroll. In the second case that the pressure drop in a cyclone can be reduced by between 8.7
when cyclone exhausts directly into the atmosphere they commonly and 11.9% with the addition of a radial evasé assuming that the separa-
contain just a cylindrical outlet pipe or one that is covered with a shallow tion efficiency is not affected since the evasé is outside of and down-
cone to keep rain out a so-called rain hat [3]. Outlet scrolls may also be stream from the cyclone.
installed on any cyclone wherein the exhaust gas is to undergo a signif- Kahrimanovic et al. [6] studied a cyclone with two configurations of
icant reduction in velocity, i.e. its expansion to the atmosphere. The the radial diffuser at the gas outlet, a “horizontal” one and an “inclined”
smooth expansion of the gas as it passes through the scroll reduces the one which was arranged at an angle of 30° with respect to the horizontal
tangential velocity of the gas flow and converts a part of its kinetic energy diffuser. These two gas outlet configurations were compared with the ra-
stored with the tangential velocity into static pressure. Without such a dial bend which was referred to as a normal outlet. With the “horizontal”
radial diffuser 30% of pressure was recovered while with the more effi-
cient but complicated “inclined” geometry the enhancement in pressure
Abbreviations: CFD, Computational Fluid Dynamics; DNS, Direct Numerical recovery was only 2–3%. A small enhancement with the “inclined” dif-
Simulations; LES, Large Eddy Simulations; RANS, Reynolds-averaged Navier-Stokes;
RSM, Reynolds stress model.
fuser was explained by the larger radius of the diffuser plate. It was also
⁎ Corresponding author. concluded that the particle separation in the cyclone was not strongly af-
E-mail address: dzmitry.misiulia@belstu.by (D. Misiulia). fected by the radial diffuser placed downstream the vortex finder.

https://doi.org/10.1016/j.powtec.2019.10.064
0032-5910/© 2019 Elsevier B.V. All rights reserved.

Please cite this article as: D. Misiulia, S. Antonyuk, A.G. Andersson, et al., High-efficiency industrial cyclone separator: A CFD study, Powder
Technol., https://doi.org/10.1016/j.powtec.2019.10.064
2 D. Misiulia et al. / Powder Technology xxx (2019) xxx

The main goal with this work is to investigate the flow pattern and where Cs1, Cs2, Cr1, Cr2, Cr3, Cr4, Cr5 are the model coefficients; aij is an an-
performance of a high-efficiency industrial cyclone separator with a isotropy tensor; Sij is a mean strain rate tensor; and Θij is a vorticity ten-
conventional gas outlet pipe and reveal the effects of adding an outlet sor. These tensors are given by
scroll or a radial bend using Computational Fluid Dynamics (CFD)
simulations. u0i u0j 2
aij ¼ − δij ð8Þ
k 3
2. Numerical modelling !
1 ∂ui ∂u j
2.1. Governing equations Sij ¼ þ ð9Þ
2 ∂x j ∂xi

A Eulerian-Lagrangian approach was applied to model two-phase !


1 ∂ui ∂u j
gas-solid flow in a cyclone. The gas phase was treated as an incompress- Θij ¼ − ð10Þ
2 ∂x j ∂xi
ible isothermal flow using an Eulerian approach and the Lagrange
method was applied for the solid particles. These particles were tracked
through the cyclone. A transport equation for turbulence dissipation rate is written as
  
2.1.1. Governing equations for continuous phase νt ∂ε
∂ νþ
The Reynolds-averaged Navier-Stokes (RANS) equations for the in- ∂ε ∂ε ε σ εRS ∂xk
þ uk ¼ ðC ε1 P k −C ε2 εÞ þ ð11Þ
compressible isothermal gas phase can be written as: ∂t ∂xk k ∂xk

∂ui where Cε1 and Cε2 are model coefficients; σεRS is a turbulent Schmidt
¼0 ð1Þ number; νt is the turbulent kinematic viscosity
∂xi
! 2
k
∂ui ∂u 1 ∂p ∂ ∂ui νt ¼ C μRS ð12Þ
þ uj i ¼ − þ ν −u0i u0j ð2Þ ε
∂t ∂x j ρ ∂xi ∂x j ∂x j
where CμRS is a model coefficient.
where ui is the time-averaged (mean) gas velocity; u0i is the fluctuating The RSM model developed by Speziale, Sarkar and Gatski [8] was
gas velocity; u0i u0j is the fluctuating Reynolds stress contribution; ν is the used in this study. This RSM model uses a quadratic relation for the
pressure-strain correlation and is more accurate than the other RSM
kinematic viscosity of the gas; ρ is the gas density; p is the mean static
models particularly for swirling flows [9]. The RSM model coefficients
pressure.
are listed in Table 1.
The Reynolds stress model (RSM) (as reported in [7]) has been
For a proper prediction of the flow field at the wall a scalable wall
proven as the most appropriate RANS turbulence model for the predic-
function [10] was applied.
tion of highly swirling flows in a cyclone separator and it was therefore
applied in this study. A separate differential Reynolds stress transport
2.1.2. Governing equations for dispersed phase
equation was solved for each Reynolds stress components according to:
The dispersed phase was calculated using particle transport model-
! 0 0! ling where representative spherical particles were tracked through the
2 k
2 ∂ui u j
∂ ν þ Cs flow. The tracking was carried out by forming a set of ordinary differen-
∂u0i u0j ∂u0i u0j 3 ε ∂xk 2 tial equations in time for each representative particle, consisting of
þ uk − ¼ P ij − δij ε þ Φij ð3Þ
∂t ∂xk ∂xk 3 equations for position and velocity using forward Euler integration
[10,11] according to:
1 0 0
where Cs is an isotropic diffusion coefficient; k ¼ u u is the turbulence
2 i j xnp ¼ xop þ uop Δt ð13Þ
kinetic energy; ε is turbulence dissipation rate; Pij is the exact produc-
tion term; δij is the Kronecker delta (δij = 1 if i = j and δij = 0 if i ≠ j); dxp
Φij is a pressure-strain correlation. up ¼ ð14Þ
dt
The exact production term is defined as:
where xp was a particle position; up was a particle velocity; the super-
∂u j ∂u scripts o and n referred to old and new values respectively; t was
P ij ¼ −u0i u0k −u0j u0k i ð4Þ
∂xk ∂xk time; Δt was a time step.
In forward integration, the particle velocity calculated at the start of
The pressure-strain correlation acts to drive turbulence towards an the time step was assumed to prevail over the entire step. At the end of
isotropic state by redistributing the Reynolds stresses and can be split the time step, the new particle velocity was calculated using the analyt-
into two parts, a slow, also known as the return-to-isotropy, term and ical solution to the particle momentum equation as:
a rapid term which are defined as:
! ! !
π 3 d up
dp ρp ¼ FD þ FG ð15Þ
Φij ¼ Φij;1 þ Φij;2 ð5Þ 6 dt

   where dp was the particle diameter; ρp was the particle density; FD was
1
Φij;1 ¼ −ε C s1 aij þ C s2 aik ajk − amn amn δij ð6Þ an aerodynamic drag force acting on the particle; FG was a net force due
3

1 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi Table 1
Φij;2 ¼ − C r1 P ij aij þ C r2 kSij −C r3 kSij amn amn RSM model coefficients.
2
  ð7Þ CμRS σεRS Cs Cs1 Cs2 Cr1 Cr2 Cr3 Cr4 Cr5 Cε1 Cε2
2  
þC r4 k aik Sjk þ ajk Sik − akl Skl δij þ C r5 k aik Θjk þ ajk Θik 0.1 1.363 0.22 1.7 1.05 0.9 0.8 0.65 0.625 0.2 1.44 1.83
3

Please cite this article as: D. Misiulia, S. Antonyuk, A.G. Andersson, et al., High-efficiency industrial cyclone separator: A CFD study, Powder
Technol., https://doi.org/10.1016/j.powtec.2019.10.064
D. Misiulia et al. / Powder Technology xxx (2019) xxx 3

to gravity being equal to the gravitational force subtracted with the configurations were investigated, an ordinary outlet pipe, an outlet
buoyant force. scroll and a radial bend as shown in Fig. 1. The geometrical dimensions
The aerodynamic drag force was defined as: of the cyclone are given in Table 2.

2
1 πdp

  2.3. Boundary conditions and numerical settings
FD ¼ CDρ
u−up
u−up ð16Þ
2 4
A velocity profile for a fully developed air flow was applied at the
where CD was the drag coefficient derived from the Schiller Naumann inlet. All cyclones were investigated at an area-averaged inlet velocity
correlation according to [10]: of ≪win≫ = 24.5 m s−1, that being the optimal velocity for this type
 of cyclone (according to [4]), see Fig. 2. The recommended inlet velocity
24
range for the investigated cyclone is 16.8–24.5 m s−1. Therefore, the cy-
C D ¼ max 1 þ 0:15 Rep0:687 ; 0:44 ð17Þ
Rep clone with an ordinary outlet pipe was also investigated at area-
averaged inlet velocities of 16.8 and 20.7 m s−1. The area-averaged
where the particle Reynolds number was: inlet velocity was computed as the ratio of the volumetric flow rate to
the inlet cross-sectional area as:
dp jus j
Rep ¼ ð18Þ
ν Q
≪win ≫ ¼ ð20Þ
ab
The net force due to gravity was in its turn given by

π 3 where Q was the volumetric flow rate; a and b were the inlet width and
FG ¼ d ρ −ρ g ð19Þ height.
6 p p
At the outlet, an opening boundary condition with zero pressure
where g was gravity. normal gradient that allows the air to cross the boundary surface in
both directions was used. The air density and dynamic viscosity were
2.2. Cyclone geometry set to 1.205 kg m−3 and 1.831 × 10−5 Pa s, respectively. A medium
turbulence intensity of 5% was set both at the inlet and the outlet. A
A high-efficiency industrial gas cyclone STsN-40 [4] with an internal non-slip smooth wall boundary condition was applied on all other
diameter of 0.3 m was studied. As previously described the three outlet boundaries. In two-phase simulations, 25,000 representative particles

Fig. 1. Cyclone with different gas outlet configurations.

Please cite this article as: D. Misiulia, S. Antonyuk, A.G. Andersson, et al., High-efficiency industrial cyclone separator: A CFD study, Powder
Technol., https://doi.org/10.1016/j.powtec.2019.10.064
4 D. Misiulia et al. / Powder Technology xxx (2019) xxx

Table 2 Table 3
Cyclone dimensions (normalized by cyclone diameter D). Numerical settings.

a b De d d1 d2 H h hc h1 h2 l s Case Gas outlet Number of mesh Numerical Δt, hCoi ≪y+≫


No configuration elements, N, 106 model 10−5
0.38 0.16 0.4 0.2 0.7 0.15 3.3 1.6 1.6 0.8 0.1 0.5 0.4
sec

1 Pipe 0.327 RSM 5 0.447 104.1


0.2−4 μm in diameter (1000 particles for 25 different diameters) with a 2 Pipe 0.551 RSM 5 0.471 90.29
3 Pipe 0.871 RSM 5 0.491 91.00
density of 1930 kg m−3, random initial positions and zero slip velocity
4 Pipe 1.211 RSM 5 0.505 91.51
were injected at the inlet. For a proper particle tracking solution, 100 in- 5 Pipe 3.601 RSM 2 0.260 54.91
tegration steps per element were applied [10,12,13]. To reduce the com- 6 Scroll 0.999 RSM 5 0.526 95.10
putational time, the particles reaching the side wall and bottom of the 7 Bend 0.975 RSM 5 0.534 93.48
dust hopper were terminated by applying the parallel and perpendicu- 8 Pipe 7.25 LES 1 0.164 34.78
9 Pipe 7.25 RSM 1 0.181 33.8
lar coefficients of restitution of 0.0 on these surfaces. For all other walls,
these coefficients were set to 0.8 which is commonly applied to many
different materials [14]. The effects of the grid size on the time-averaged tangential and axial
The time step Δt was 1 × 10−5 –5 × 10−5 s that resulted in a mean velocities in the cyclone body along the x-axis (y = 0, z = 0) can be seen
volume-averaged Courant number of 0.16–0.51 for all cases (Table 3). in Fig. 3. The velocities are normalized with the area-averaged inlet ve-
The root-mean-square scaled residuals were set below 2 × 10−5 locity ≪win≫.
which required, in average, 5 loop iterations within a time step. All meshes captured the inverted “W”-shape of the tangential veloc-
The simulations were performed on a 64-bit Linux cluster at Luleå ity profile, but the axial velocity profiles predicted with the coarsest
University of Technology using a commercial solver ANSYS CFX 15.0 mesh significantly differed from the others. The maximum tangential
(ANSYS, Inc., Canonsburg, Pennsylvania, USA). During simulations, tan- velocity increased with the increase in number of mesh elements
gential velocities at some points in the cyclone body, area-averaged (Fig. 3). However increasing the number of elements in the mesh
total and static pressure at different sections, total pressure drop, from 0.871 million had negligible effects on the tangential and axial ve-
volume-averaged turbulence kinetic energy and turbulence eddy dissi- locity distributions. To conclude, the flow pattern in a cyclone can be
pation as well as area-averaged y+ were monitored to be sure that the quite accurately captured with a mesh consisting of 0.871 million
flow inside the cyclone had reached its fully developed state (statisti-
cally steady-state). After that, an arithmetic averaging was initiated.

3. Grid independence study

A grid independence study was performed for a cyclone separator


with an ordinary gas outlet pipe. Five high quality grids with 0.327,
0.551, 0.871, 1.211 and 3.601 million elements were generated, No
1–5 in Table 3. For a proper near the wall treatment with scalable wall
function in RSM, the thickness of the first grid nearest the wall should
result in y+ from 11.06 to 300. The mean area-averaged values of y+
for all 5 grids were in this range, see Table 3.

Fig. 3. Predicted time-averaged tangential and axial velocities along the x-axis (y = 0, z =
Fig. 2. Inlet velocity contour plot. 0) normalized by area-averaged inlet velocity.

Please cite this article as: D. Misiulia, S. Antonyuk, A.G. Andersson, et al., High-efficiency industrial cyclone separator: A CFD study, Powder
Technol., https://doi.org/10.1016/j.powtec.2019.10.064
D. Misiulia et al. / Powder Technology xxx (2019) xxx 5

Fig. 4. Meshes for a cyclone with the investigated gas outlet configurations.

elements. Using a finer mesh will not lead to significant changes in ve-
locity profiles but will increase the computational time. Therefore,
meshes with the same element size were created for the cyclones
with an outlet scroll (No 6 in Table 3) and radial bend (No 7 in
Table 3), as shown in Fig. 4. 2
Due to a lack of information in the literature regarding pressure and
velocity distributions in the investigated cyclone with its scroll inlet,
there are no reliable data to perform a validation study. Based on the re- 1.5
sults of previous studies of a reverse-flow cyclone separator with a
helical-roof inlet [7,10–13] that showed good agreement between the
numerical and experimental data, it is likely that the numerical model
1
and settings can adequately predict the flow pattern in this cyclone as
well. Nevertheless, in order to exclude some uncertainty, Large Eddy
Simulations (LES) of a cyclone with an ordinary gas outlet pipe were
performed using the dynamic Smagorinsky-Lilly model [15] and the re- 0.5 RSM
sults were compared with the RSM simulations (No 8 and 9 in Table 3). LES
LES simulations require much finer mesh and the mesh size should
preferably be in the dissipation range of the turbulent length scale 0
where motions experience viscous effects. Based on mean volume- -1 -0.5 0 0.5 1
averaged turbulence kinetic energy and eddy dissipation the length 0.2
scales and ranges for the cyclone were defined according to [11,16]
and are sketched in Fig. 5.
The simulations were performed with a mesh consisting of 7.25 mil- 0.1
lion elements. It can be seen from Fig. 5 that the average mesh element
size for the cyclone was within the dissipation range of the universal
equilibrium range. 0
The tangential and axial velocity profiles predicted with the RSM and
LES models are shown in Fig. 6. The tangential velocity predicted with
-0.1

-0.2

-0.3
-1 -0.5 0 0.5 1

Fig. 6. Predicted with RSM and LES time-averaged tangential and axial velocities along the
Fig. 5. Eddy sizes for the turbulent flow within the cyclone. x-axis (y = 0, z = 0) normalized by area-averaged inlet velocity.

Please cite this article as: D. Misiulia, S. Antonyuk, A.G. Andersson, et al., High-efficiency industrial cyclone separator: A CFD study, Powder
Technol., https://doi.org/10.1016/j.powtec.2019.10.064
6 D. Misiulia et al. / Powder Technology xxx (2019) xxx

Fig. 7. Time-averaged static pressure, tangential velocity and axial velocity in the cyclone body and in the vortex finder.

Please cite this article as: D. Misiulia, S. Antonyuk, A.G. Andersson, et al., High-efficiency industrial cyclone separator: A CFD study, Powder
Technol., https://doi.org/10.1016/j.powtec.2019.10.064
D. Misiulia et al. / Powder Technology xxx (2019) xxx 7

-50 -50 -50

-40 -40 -40

-30 -30 -30

-20 -20 -20

-10 -10 -10

0 0 0
-0.1 0 0.1 -0.1 0 0.1 -0.1 0 0.1

z=0 z = 0.15 m z = 0.3 m

10 10 10

5 5 5

0 pipe 0 0
scroll
bend
-5 -5 -5

-0.1 0 0.1 -0.1 0 0.1 -0.1 0 0.1

Fig. 8. Time-averaged tangential and axial velocity profiles in the cyclone body along the x-axis at three different z positions (z= 0, 0.15, and 0.3 m).

LES were slightly higher. Both the RSM and LES models revealed the for the three gas outlet configurations are shown in Fig. 7. The radial
same profiles for tangential and axial velocities and agreed well. profiles of the tangential and axial velocity components along the x-
axis at three cut sections along the cyclone axis (z= 0, 0.15 and
4. Results and discussion 0.3 m) are presented in Fig. 8. Since the geometrical axis of the outlet
scroll and radial bend did not coincide with the cyclone geometrical
4.1. Effects of the gas outlet configuration on the flow pattern in the cyclone axis (z-axis) which was used as the axis of rotation for computing
the tangential and axial velocities, the contour plots of the static pres-
The contour plots of the time-averaged static pressure, tangential sure and velocity field in the scroll outlet and radial bend are not
velocity and axial velocity in the cyclone body and in the vortex finder shown in Fig. 7.

Fig. 9. Iso-surfaces with the vorticity around z-axis of −3000 s−1 for the three configurations.

Please cite this article as: D. Misiulia, S. Antonyuk, A.G. Andersson, et al., High-efficiency industrial cyclone separator: A CFD study, Powder
Technol., https://doi.org/10.1016/j.powtec.2019.10.064
8 D. Misiulia et al. / Powder Technology xxx (2019) xxx

Fig. 10. Iso-surfaces with the vorticity around z-axis of −2000, −2500 and −3000 s−1 at area-averaged inlet velocity of 16.8 (a), 20.7 (b) and 24.5 m s−1 (c) respectively.

Static pressure in a cyclone (Fig. 7) has its minimum value in the vor- scroll and radial bend downstream the vortex finder increased this dif-
tex core region, i.e. near or at the geometrical axis of the cyclone and in- ference to 6.3 and 5.1 kPa respectively.
creases towards the periphery where it reaches its maximum value. The The maximum tangential velocity was approximately twice as large
difference between the maximum and minimum values, i.e. the differ- as the inlet velocity and was practically constant along the cyclone axis
ence in static pressure at the cyclone wall and its centerline was about (Fig. 7). The radial profile of the tangential velocity distribution had an
4.6 kPa in the cyclone with the ordinary outlet pipe. Placing the outlet inverted “W”-shape and was almost independent of the z-axis (Fig. 8).

Fig. 11. The instantaneous tangential and axial velocity contour plots obtained with RSM and LES.

Please cite this article as: D. Misiulia, S. Antonyuk, A.G. Andersson, et al., High-efficiency industrial cyclone separator: A CFD study, Powder
Technol., https://doi.org/10.1016/j.powtec.2019.10.064
D. Misiulia et al. / Powder Technology xxx (2019) xxx 9

Fig. 12. Streamlines in the vortex finder and different gas outlet configurations.

The outlet scroll and radial bend had insignificant effects on the tangen- axial direction. Spiral modes of instability were found to cause a lateral
tial velocity field in the cyclone body. expansion of the cross-section of the vortex core, and a corresponding
The axial velocity field in the cyclone with different gas outlet config- drop in axial velocity. When the assumption of axial periodicity was
urations could be divided into two zones, an outer zone with negative neglected, as in a fully three-dimensional DNS, this then led to an axial
velocity values and an inner zone with positive velocity values. The stagnation point – indicative of vortex breakdown. Vortex tip stabiliza-
outer zone determined the downwardly directed outer vortex and the tion has been also explored at the bottom of a cyclone [19–21].
inner one defined the size of the upwardly directed inner vortex. Not Surprisingly, such a “vortex-breakdown” phenomenon does not
the outlet scroll nor a radial bend had a noticeable effect on the axial ve- occur if the outlet scroll or radial bend is placed downstream the vortex
locity distribution in the outer zone but they did effect the inner zone. finder. Moreover, the phenomenon does not appear in other cyclone ge-
This can be explained by some transformation of the inner vortex ometries with the ordinary gas outlet pipe, see for instance [11,22].
which precesses around the cyclone axis. The time-averaged parameters of the flow at the gas outlet are not
Fig. 9 represents the isosurfaces of the vorticity around z-axis of representative since the flow at the gas outlet is unstable and there is
−3000 s−1. A very interesting phenomenon, a “vortex-breakdown a strong precession of the vortex core [11]. Instead, 15 streamlines
with bifurcated double helix” occurred in the cyclone separator with coloured by absolute velocity in the vortex finder, outlet scroll and ra-
an ordinary outlet pipe. The onset of the “vortex-breakdown” was a dial bend are presented in Fig. 12. The flow in the conventional vortex
bit upstream the vortex finder where the upwardly directed inner vor- finder is still quite swirling at the outlet and the maximum absolute ve-
tex divided into two vortices which then entered the vortex finder. As a locity reaches 70 m s−1. In fact, a well-designed outlet scroll should re-
result, there were two vortices in the vortex finder which broke up at duce the velocity due to expansion of the cross-sectional area. However,
the gas outlet section. these simulations showed that the velocities in the outlet scroll and ra-
The revealed vortex-breakdown with bifurcated double helix oc- dial band were higher than in the ordinary outlet pipe and the stream-
curred in the cyclone separator with an ordinary outlet pipe at different lines were not structured, especially in the outlet scroll. With the
inlet velocities over the whole recommended operation range of 16.8– investigated outlet scroll design the velocity at the outlet was not re-
24.5 m s−1. Fig. 10 shows the vortex-breakdown phenomenon at duced and as a result there maybe higher pressure losses at the outlet.
area-averaged inlet velocities of 16.8, 20.7 and 24.5 m s−1. The inlet ve-
locity did not affect the vortex-breakdown but changed the swirl inten- 4.2. Effects of the gas outlet configuration on the cyclone performance
sity in the cyclone and vortex finder.
Such a “vortex-breakdown” phenomenon can be also seen in Fig. 11 Cyclone performance is characterized by pressure drop and collec-
which represents the instantaneous tangential and axial velocity con- tion efficiency.
tour plots in the vortex finder. The contour plots predicted with the Pressure drop is an important cyclone performance characteristic
RSM and LES model are similar and show good agreement. since it determines operation cost of a cyclone which is proportional
Such a vortex breakdown with bifurcated double helix in a swirling to the energy required to overcome these pressure losses. The instanta-
flow was reported by Meliga et al. [17]. Also, Karniadakis and Sherwin neous total pressure drop across the cyclone with the three different gas
[18] identified such an onset of the “vortex-breakdown” with Direct Nu- outlet configurations is shown in Fig. 13 whereas the time-averaged
merical Simulations (DNS) using a Fourier series approximation in the total pressure drop and the pressure losses in the cyclone body

4000 4000 4000

3500 3500 3500

3000 3000 3000

pipe scroll bend


2500 2500 2500
1 1.1 1.2 1 1.1 1.2 1 1.1 1.2

Fig. 13. Instantaneous total pressure drop across the cyclone with different gas outlet configurations.

Please cite this article as: D. Misiulia, S. Antonyuk, A.G. Andersson, et al., High-efficiency industrial cyclone separator: A CFD study, Powder
Technol., https://doi.org/10.1016/j.powtec.2019.10.064
10 D. Misiulia et al. / Powder Technology xxx (2019) xxx

3500 1
pipe pipe
3000 scroll
scroll
bend 0.8
bend
Pressure losses [Pa]

2500

2000
0.6
1500

0.4
1000

500
0.2
0 cut-size
Total pressure drop Upstream VF inlet Downstream VF inlet
0
Fig. 14. Pressure losses in the cyclone with different gas outlet configurations. 10-7 10-6 10-5

(upstream the vortex finder inlet) and in the vortex finder and gas out-
let (downstream the vortex finder inlet) are plotted in Fig. 14. Fig. 15. Collection efficiency of a cyclone with different gas outlet configuration.
Surprisingly, both the outlet scroll and bend outlet increased the
total pressure drop across a cyclone instead of the opposite. Moreover,
they lead to dramatical increase in oscillations of the instantaneous expected, since the outlet scroll and radial bend were downstream from
pressure drop. The amplitude of these oscillations (i.e. the difference be- the vortex finder and did not significantly affect the flow field in the cy-
tween the maximum and minimum value of the instantaneous pressure clone body where the particles are separated from the gas flow. The pre-
drop) was very small in the cyclone separator with the conventional dicted cut-size of all cyclones was practically the same and about 1.1 μm.
outlet pipe and was equal to 0.65% of the average total pressure drop. Cyclones with such a small value of the cut-size (i.e. high separation ca-
However installing the outlet scroll and radial bend downstream from pability) can be used as a final separation stage in industry.
the vortex finder increased the amplitude of oscillations to 16.2% and
33.96%, respectively (Fig. 13).
5. Conclusions
Fig. 14 shows that the pressure losses in the cyclone body upstream
from the vortex finder, i.e. the pressure losses from the cyclone inlet to
An industrial high-efficiency cyclone separator with three different
the vortex finder inlet, were nearly 2000 Pa and were almost indepen-
gas outlet configurations (conventional outlet pipe, outlet scroll and ra-
dent of the gas outlet configuration downstream from the vortex finder.
dial bend) has been computationally investigated using Reynolds stress
This can be explained by the fact that the gas outlet configuration has
model simulations and Large eddy simulations and the effects of the
minor effects on the velocity distribution in the cyclone body. Placing
outlet scroll and radial bend placed downstream the vortex finder on
the outlet scroll and radial bend downstream from the vortex finder,
the flow pattern and cyclone performance have been revealed. The fol-
however, increased the pressure losses in the vortex finder and down-
lowing conclusions can be drawn:
stream from it, i.e. from the vortex finder inlet to the cyclone gas outlet.
This can be explained by higher velocities in the outlet scroll and radial • a unique “vortex-breakdown with bifurcated double helix” phe-
bend (Fig. 12). nomenon occurs in the high-efficiency cyclone separator with an
Time-averaged pressure drop coefficients (based on the inlet veloc- ordinary outlet pipe. The observation is that the inner vortex split
ity and mean axial velocity in a cyclone body) of a cyclone with different into two vortices. This is a phenomenon that has never been
gas outlet configurations are listed in Table 4. It shows that the installa- found in cyclone separators before. This phenomenon does not ap-
tion of an outlet scroll and radial bend downstream from the vortex pear if the outlet scroll or radial bend is placed downstream the
finder increased the dimensionless pressure losses, the Euler number, vortex finder. Additional studies are needed to reveal the reason
by 5.1% and 6.4% respectively. A negative effect in pressure drop reduc- and nature for the double helix;
tion with the scroll outlet was also reported by Lazarev [4] and Idel'chik • an outlet scroll and radial bend have insignificant effects on the ve-
[23]. The investigated radial bend (R = 1.5D) was probably too abrupt locity field in a cyclone body except regarding the axial velocity
and located too close to the vortex finder. These negative effects can distribution in the inner vortex as described above;
be overcome by optimization of the gas outlet design. • installation of an outlet scroll and radial bend downstream from
The effects of the outlet scroll and radial bend on the cyclone collec- the vortex finder increases the dimensionless pressure losses by
tion efficiency are presented in Fig. 15. The investigated gas outlet con- 5.1% and 6.4% respectively. Moreover it significantly destabilizes
figurations had almost no effect on the cyclone grade efficiency curve, as the pressure losses by increasing the amplitude of instantaneous
pressure drop oscillations from 0.65% to 16.2% and 33.96% respec-
tively;
Table 4 • the outlet scroll and radial bend have almost no effect on the cy-
Pressure drop coefficients (time-averaged Euler numbers) of a cyclone with different gas clone separation capability;
outlet configurations.
• the geometry of the investigated scroll outlet is not properly de-
Pressure drop coefficients (Euler numbers) Gas outlet configuration signed and needs to be optimized;
Pipe Scroll Bend • placing the radial bend (R = 1.5D) right downstream from the vor-
tex finder is not recommended since it leads to high pressure
Euin 8.41 8.84 8.95
losses. This negative effect can be probably overcome by increasing
Eu 1404 1475 1494
the bend radius and placing the radial bend further downstream
from the vortex finder.

Please cite this article as: D. Misiulia, S. Antonyuk, A.G. Andersson, et al., High-efficiency industrial cyclone separator: A CFD study, Powder
Technol., https://doi.org/10.1016/j.powtec.2019.10.064
D. Misiulia et al. / Powder Technology xxx (2019) xxx 11

Future extension of this work is firstly to investigate the reasons and Nomenclature and units
nature of the unique “double spiral vortex-breakdown” phenomenon,
Upper-case Roman
and secondly optimize the gas outlet configuration in terms of mini-
mum pressure losses. CD [–]: drag coefficient
Cr1−r5 [–]: RSM model coefficient
Cs [–]: isotropic diffusion coefficient
Acknowledgements Cs1,Cs2 [–]: RSM model coefficient
CμRS [–]: RSM model coefficient
This work was supported by the Alexander von Humboldt Founda- Cε1,Cε2 [–]: RSM model coefficient
Co [–]: Courant number
tion and partially by the Swedish Institute.
D [m]: cyclone body diameter
De [m]: vortex finder diameter
References Eu [–]: Euler number normalized by mean axial velocity in a cyclone body
Euin [–]: Euler number normalized by inlet velocity
[1] A. Alves, J. Paiva, R. Salcedo, Cyclone optimization including particle clustering, Pow- FD [N]: drag force
der Technol. 272 (2015) 14–22. FG [N]: net force due to gravity
[2] Advanced Cyclone Systems, Hurricane and ReCyclone systems, http://www. L [m]: characteristic length scale
acsystems.pt/ 2017 (accessed 7 March 2018). N [–]: number of mesh elements
[3] P.A. Funk, Reducing cyclone pressure drop with evasés, Powder Technol. 272 (2015) Rep [–]: particle Reynolds number
276–281. Sij [s−1]: mean strain rate tensor
[4] V.A. Lazarev, Cyclones and Vortex Dust Traps: A Handbook, 2nd ed. Ozon-NN, Q [m3 s−1]: volumetric gas flow rate
Nizhnii Novgorod, Russia, 2006.
[5] A.C. Hoffmann, L.E. Stein, Gas Cyclones and Swirl Tubes, 2nd ed. Springer Berlin Hei- Lower-case Roman
delberg, New York, 2008.
[6] D. Kahrimanovic, S. Puttinger, G. Aichinger, S. Pirker, Minimizing pressure drop in a [m]: cyclone inlet height
cyclone separators measurements and numerical simulations, 11th World Filtration aij [–]: anisotropy tensor
Congress, 16-20 April, 2012, Graz, Austria, 2012. b [m]: cyclone inlet width
[7] D. Misiulia, A.G. Andersson, T.S. Lundström, Computational investigation of an in- dp [m]: particle diameter
dustrial cyclone separator with helical-roof inlet, Chem. Eng. Technol. 38 (8) g [m2 s−1]: gravity
(2015) 1425–1434. k [m2 s−2]: turbulence kinetic energy
[8] C.G. Speziale, S. Sarkar, T.B. Gatski, Modelling the pressure-strain correlation of tur- l [m]: length scale
bulence: an invariant dynamical systems approach, J. Fluid Mech. 227 (1991) lDI [m]: length scale which splits the universal equilibrium range into inertial subrange and
245–272. dissipation range
[9] ANSYS CFX-Solver Theory Guide, Release 15.0. , ANSYS, Inc, Canonsburg, Pennsylva- lEI [m]: length scale that forms the demarcation between the energy containing range and
nia, 2013. the universal equilibrium range
[10] D. Misiulia, K. Elsayed, A.G. Andersson, Geometry optimization of a deswirler for cy- l0 [m]: integral scale
clone separator in terms of pressure drop using CFD and artificial neural network, lη [m]: Kolmogorov length scale
Sep. Purif. Technol. 185 (2017) 10–23. p [Pa]: static pressure
[11] D. Misiulia, S. Antonyuk, A.G. Andersson, T.S. Lundström, Effects of deswirler posi- Δp [Pa]: total pressure drop across a cyclone
tion and its centre body shape as well as vortex finder extension downstream on cy- r [m]: radius
clone performance, Powder Technol. 272 (2018) 14–22. s [m]: vortex finder length
[12] D. Misiulia, A.G. Andersson, T.S. Lundström, Large eddy simulation investigation of t [s]: time
an industrial cyclone separator fitted with a pressure recovery deswirler, Chem. Δt [s]: time step
Eng. Technol. 40 (4) (2017) 709–718. u [m s−1]: gas velocity
[13] D. Misiulia, A.G. Andersson, T.S. Lundström, Effects of the inlet angle on the collec- up [m s−1]: particle velocity
tion efficiency of a cyclone with helical-roof inlet, Powder Technol. 305 (2017) wax [m s−1]: axial component of gas velocity
48–55. win [m s−1]: inlet gas velocity
[14] C.T. Crowe, J.D. Schwarzkopf, M. Sommerfeld, Y. Tsuji, Multiphase Flows with Drop- wt [m s−1]: tangential component of gas velocity
lets and Particles, 2nd ed. CRC Press, Boca Raton, Florida, USA, 2012. xp [m]: particle position
[15] D.K. Lilly, A proposed modification of the Germano subgrid-scale closure method, y+ [−]: dimensionless parameter
Phys. Fluids A 4 (3) (1992), 633635. δij [–]: Kronecker delta
[16] S.B. Pope, Turbulent Flows, IOP Publishing, 2001. ε [m2 s−3]: turbulence eddy dissipation
[17] P. Meliga, F. Gallaire, J.-M. Chomaz, A weakly nonlinear mechanism for mode selec- ν [m2 s−1]: kinematic viscosity
tion in swirling jets, J. Fluid Mech. 699 (2012) 216–262. νt [m2 s−1]: turbulent kinematic viscosity
[18] G.E. Karniadakis, S.J. Sherwin, Spectral/hp Element Methods for Computational Fluid ρ [kg m−3]: gas density
Dynamics, Oxford University Press, 1999. ρp [kg m−3]: particle density
[19] J.J. Derksen, Separation performance predictions of a Stairmand high-efficiency cy- σμRS [–]: RSM model coefficient
clone, AICHE J. 49 (6) (2003) 1359–1371. θij [s−1]: vorticity tensor
[20] W. Peng, A.C. Hoffmann, P.J.A.J. Boot, A. Udding, H.W.A. Dries, A. Ekker, J. Kater, Flow Φij [m2 s−3]: pressure-strain correlation
pattern in reverse-flow centrifugal separators, Powder Technol. 127 (2002)
212–222.
[21] C. Cortés, A. Gil, Modeling the gas and particle flow inside cyclone separators, Prog. Other symbols
Energy Combust. Sci. 33 (2007) 409–452.
[22] L.S. Brar, K. Elsayed, Analysis and optimization of multi-inlet gas cyclones using large : time-averaged (mean)
eddy simulation and artificial neural network, Powder Technol. 311 (2017) ': fluctuating
465–483. 〈 〉: volume-averaged
[23] I.E. Idel'chik, Handbook of Hydraulic Resistance, 3rd ed. Begell House, 1996. ≪≫: area-averaged

Please cite this article as: D. Misiulia, S. Antonyuk, A.G. Andersson, et al., High-efficiency industrial cyclone separator: A CFD study, Powder
Technol., https://doi.org/10.1016/j.powtec.2019.10.064

S-ar putea să vă placă și