Proceedings of the ASME 2016 International Mechanical Engineering Congress and Exposition

IMECE2016
November 11-17, 2016, Phoenix, Arizona, USA

IMECE2016-66012

EFFECTS OF STATOR CONFIGURATIONS ON THE FLOTATION

PERFORMANCE OF INDUCED GAS FLOTATION MACHINE
Ji-Gu Lee. Ji-Yun Kang.
Graduate School of Mechanical Engineering, Graduate School of Mechanical Engineering,
Sungkyunkwan University Sungkyunkwan University
Suwon 16419, Republic of Korea Suwon 16419, Republic of Korea

Youn-Jea Kim
School of Mechanical Engineering,
Sungkyunkwan University
Suwon 16419, Republic of Korea

ABSTRACT
Induced Gas Flotation (IGF) vessel is used for water [4]. The Euler-Euler equation is more suitable than the
treatment of plant industries such as oil sand and chemical Reynolds-averaged Navier Stokes (RANS) equation in two-
plants. An understanding of the interaction between the stator phase flow which includes dispersed multi-phase media, such
and rotor is essential for the design of IGF with consideration as bubbles. Many numerical studies have been performed using
of geometric blade configuration is essential for the design of the Euler-Euler equation. Salem-Said et al. [5] experimentally
IGF. In this study, the effect of the number of stator blades on investigated the effects of a bubble size on the performance of
flotation performance was numerically investigated using the Dorr-Oliver cell for single and two phase flows, turbulent
commercial code, ANSYS CFX ver. 16.1. The two-phase kinetic energy and air volume fraction according to various
(water and air) flow characteristics in the forced-air bubble size, and derived experimental data in single- and two-
mechanically stirred Dorr-Oliver flotation cell were considered. phase flows using the Euler–Euler equation. Shi et al. [6]
The flotation performance was evaluated on the basis of the evaluated the performance of various stator configurations the
correlations among the number of stator blades (8, 12, 16, 20, particle image velocimetry (PIV) technique and CFD. Xia et al.
24), power number and void fraction. By comparing the result [7] verified the flow field transition using a turbulence model
of each case, the newly designed model with 12 stator blades and analyzed the effect of rotation speed on the flotation
which had the highest flotation performance was derived. performance. Liu et al. [8] investigated flow characteristics and
the effect of collisions between the bubbles and particles in
1. INTRODUCTION water. While many studies of rotors have been performed, few
Induced gas flotation (IGF) vessels are used for industrial have assessed the effect of the stator on performance. In this
water treatment such as in oil sand and chemical plants, since study, the effect of stator configuration on flotation
they can separate a particles (oil, solid) of water from waste performance was numerically considered. In particular, the
water using microbubble flotation characteristics [1,2]. In effect of the number of stator blades on flotation performance
particular, IGF vessels can remove a particle from water, by was evaluated by varying the flow parameters, such as air
injecting a bubble into the rotor inlet of a vessel. Thus, study of superficial velocity, void fraction, and power number
the geometrical configurations of the stator and rotor, which are (dimensionless number).
crucial components of IGF vessels, has been performed to
improve flotation performance. Since the accuracy of 2. NUMERICAL ANALYSIS
computational fluid dynamics (CFD) has been improved by
computer technology numerical CFD analysis is a powerful 2.1 REFERENCE MODEL
method [3]. Compared to single– and two-phase flows, single- Figure 1 shows the schematic of The IGF vessel. To predict
phase flow at the initial design stage is useful for determining the complicated flow fields, the Dorr-Oliver cell (6 𝑚3 ), which
the flow characteristics of IGF while two-phase flow is more creates bubbles and floating particles using mechanical
suitable for definge the void fraction and turbulence dispersion agitation, was selected as a reference model. The IGF vessel

1 Copyright © 2016 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/conferences/asmep/90992/ on 06/28/2017 Terms of Use: http://www.asme.org/abo

 Table 1. Parameters considered in this study
Cases The number of stator blades, N
Case 1 8
Case 2 12
Case 3 16(ref) *
Case 4 20
Case 5 24
* denotes reference value
Table 2. Boundary conditions
Rotating velocity [RPM] 100, 150, 200, 250*, 300
Working fluid Water at 25℃

Turbulence model Shear stress transport

Fig. 1. Schematic of IGF
Rotational domain Moving reference frame
The Number of Rotor 6
blades Stator N
Interface condition Frozen rotor
Convergence criteria 1e-4
* denotes reference value

Fig. 2. Design parameters of IGF

consists of a shaft with an air inlet, a rotor with 6 blades, a

stator with 16 blades and a vessel. The Two-phase flow of
water and air as the working fluid was used for numerical
analysis. The diameter of the vessel (D) is 2430 mm, the height
(H) is 1632 mm and the clearance (C) between the stator and
the bottom of the vessel is 237 mm. The thickness of blade is Fig. 3. Boundary conditions
20 mm. As shown in Table 1, flotation performance was
evaluated on the basis of the correlations among the number of accuracy of interpretation over the wall, shear stress transport
stator blades (8, 12, 16, 20and 24). Figure 2 shows the IGF (SST) was selected to consider the recirculation and turbulence
design parameters. In this study, the rotor speeds 100, 150, 200, flow effects.
250 and 300 rpm were considered.
2.3 NUMERICAL METHOD
2.2 BOUNDARY CONDITIONS Two-phase flow analyses are generally divided into Euler-
No-slip condition was applied to the interior wall of the IGF Lagrange and Euler-Euler approaches. The Euler-Lagrange
and vessel. The flow region was classified into rotational and approach is a continuum based on the Navier-Stokes equation.
stationary domains to assess the rotational effect of the rotor. In When the void fraction of a dispersed secondary phase is low,
particular, frozen rotor model by MRF method was applied to the Euler-Lagrange approach can be used. Since the approach
interface. Coriolis force and centrifugal force were applied in considers both phases as a continuum mixture, the Euler-Euler
this study. The relevant boundary conditions are given in Table approach is applied in this study. Thus, the volume rate can be
2 and Fig. 3. The numerical calculation was conducted using used in the Euler-Euler approach [9]. In this study, due to the
the commercial code, ANSYS CFX 16.1. To increase the presence of a large amount of bubbles in IGF, Euler-Euler

2 Copyright © 2016 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/conferences/asmep/90992/ on 06/28/2017 Terms of Use: http://www.asme.org/abo

 approach is used. The continuity and momentum equations are
as follows:

(i pi )
 (i pi V i )  0 (1)
t

( i pi V i )
 ( i pi V i V i )
t
  i p  ( i i (V i  (V i )T )  Si  M i
(2)

where i denotes the phase,  is the volume rate,  is the

mass density,  is the dynamic viscosity, p indicates the
pressure considering effect of gravity, V i is the velocity
vector, S is the outside momentum and M is the interface
adhesion.
Fig. 5. Grid dependency
2.4 GRID SYSTEMS
The inflation grid condition was applied to the rotor and the
surrounding vessel to improve the accuracy of the analysis. In
addition, the grid systems are densely applied in the vicinity of
the rotor jet to increase the accuracy of the floating analysis
(Fig. 4). The number of grids was considered to evaluate the

(a) Vessel Fig. 6. Comparison results of the water radial velocity

between the present study and existing experimental data [5]

grid dependency. Power number and void fraction ratio in the

reference model sensed the change from 500,000 to 2500,000
elements (Fig. 5). As a result, the volume fraction rate and
power number were constant at over 2,000,000 elements, which
led to application of 2,000,000 elements to IGF.

2.5 FLOTATION PERFORMANCE

The flotation performance was determined from the flow
pattern in the IGF. The void fraction ratio (VFR) on the water
surface and the power number ( 𝑃𝑜 ) were used to evaluate the
(b) Stator and rotor effect of the stator blade configuration. VFR is defined as the
Fig. 4. Grid systems volume of a constituent V𝑠𝑢𝑟𝑓𝑎𝑐𝑒 divided by the volume of all
constituents as follows:

3 Copyright © 2016 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/conferences/asmep/90992/ on 06/28/2017 Terms of Use: http://www.asme.org/abo

 (a) Longitudinal section of the reference model (c) Air superficial axial velocity

(b) Air superficial radial velocity (d) Air volume fraction

Fig. 7. Results of the reference model

𝑉 curve at r=0.49 m shows a similar tendency with the present

VFR = (3)
𝑉 study.

The power number is defined as the residence force exerted 3. RESULTS AND DISCUSSION
by the working fluid of rotating stator. The residence force can
be changed by altering the stator configuration and properties
3.1 FLOW FIELDS
of the working fluid. The power number, 𝑃𝑜 is a dimensionless Figure 7 shows air superficial velocity profiles in the axial
coefficient that is expressed as follows: and radial directions and the air volume fraction. Velocity and
𝑃 volume fraction were measured at r = 0.52m and 0.8m, where r
𝑃0 = (4) is the radius of the vessel. The air superficial axial velocity,
𝜌𝑁 𝐷
which is the flotation velocity, in the vicinity of the rotor
where P is the power consumption, ρ is the density of the
showed highest while the velocity became gradually constant
working fluid, N is the rotational speed and D is the diameter of far away from stator. In addition, the air superficial radial
the rotor. velocity, in the vicinity of the rotor was increased. Together
with the fact that the air volume fraction increased near the
2.6 VALIDATION surface (refer to Fig. 7(d), this indicates that the most of the
In order to verify the reliability of the numerical results, the spouted bubbles rose to the surface. As the radius was far away
water radial velocity was compared with existing experimental from stator, the flotation velocity increased, since the flow
data for Dorr-Oliver flotation cell [5] was compared and direction is opposite that of flotation. Thus, the flotation
depicted in Fig. 6. The water velocity at Y=0.3 m height of velocity also depended on the configuration of the stator blades.
reference model was about 0.6 m/s. It is noted that the velocity

4 Copyright © 2016 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/conferences/asmep/90992/ on 06/28/2017 Terms of Use: http://www.asme.org/abo

 (a) Void fraction ratio
(a) Reference model (b) Case 2
Fig. 9. Velocity vector distributions

(b) Power number

Fig. 8. Results of VFR and power number for each case
(a) Reference model (b) Case 2
Fig. 10. Void fraction distributions
3.2 FLOTATION PERFORMANCE
To evaluate flotation performance, the void fraction ratio
observed that the flow back to around the shaft of rotor and the
(VFR) and power number (Po) are shown in Fig. 8. In case
vortex field upward in the vessel were occurred. However, the
2with a rotation speed of 100 rpm, VFR was higher and the Po
vortex fields are around the rotor and in the upper vessel in
lower. However, IGF performance cannot be evaluated using
which the upward movement of bubbles was distributed was
only the VFR, because the waste-water capacity of the IGF is
decreased in Case 2 (Fig. 8(b)). Bubbles were interrupted by
limited. Thus, the power number is the proper index for
two vortex fields. Since the Case 2 showed less vortex field
evaluating IGF performance. With regard to the correlation of
upward, it exhibited higher flotation performance than the
the VFR and the number of stators, as the number of stator
reference model. Thus, the configuration of stator blades can be
blades decreased, the VFR increased as bubbles emerging from
designed to increase the higher flotation performance.
the rotor jet were interrupted by the stator blades.
3.4 VOID FRACTION RATIO
3.3 VELOCITY VECTOR DISTRIBUTION
Figure 10 shows the void fraction ratio in the axial direction
Flow fields generated by rotational rotor are shown in Fig. 9
to indicate the flotation performance of IGF. The void fraction
in terms of velocity vector fields and void fraction. Air is
ratio was influenced by different flow fields from the number of
injected into the rotor inlet, flows through the shaft, is
stator blades. The discharged bubbles floated to the water
discharged from the rotor jet and spreads throughout the vessel
surface. Bubble loss occurred due to collision with the vessel
as bubbles. By comparing the velocity vector profiles, it was

5 Copyright © 2016 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/conferences/asmep/90992/ on 06/28/2017 Terms of Use: http://www.asme.org/abo

 wall. Thus, the vessel configuration can exert a significant an [4] Malhorta, D., Taylor, P. R., Spiller, E. and LeVier M., 2009,
influence on the flotation performance. Recent advances in mineral processing plant design, Society for
Mining, Metallurgy, and Exploration (SME), New York, pp.
4. CONCLUSIONS 220~252.
In this study, the effects of the number of stator blades and [5] Salem-Said, A., Fayed, H. and Ragab, S., 2013, “Numerical
rotation speed on the flotation performance of an IGF vessel simulations of two-phase flow in a Dorr-oliver Flotation Cell
were investigated using the MRF method. As the rotating speed Model”, Advances in Mineral Analytical Techniques, Vol. 3,
increased, floating bubbles are dispersed, and the power No. 3, pp. 258~336.
number is increased. In particular, Case 2 (100rpm) exhibited [6] Shi, S., Zhang, Ming. and Chen D., 2015,”Experimental and
the best flotation performance and showed a few vortex fields computational analysis of the impeller angle in a flotation cell
with a relatively high VFR and low power number. by PIV and CFD”, International Journal of Mineral Processing,
Consequently, it resulted in 12 % of efficiency increment Vol. 142, pp. 2~9.
compared with the reference model [7] Xia. J., Rinne, A. and Gronstand, S., 2009, “Effect of
turbulence models on prediction of fluid flow in an Outotec
ACKNOWLEDGEMENT flotation cell”, Minerals Engineering, Vol. 22, No. 11, pp.
This research was supported by a grant (16IFIP-B089065- 880~885.
03) from the Plant R&D Program funded by Ministry of Land, [8] Liu, T. Y. and Schwarz, M. P., 2009, “CFD-based multiscale
Infrastructure and Transport of the Korean Government. modeling of bubble-particle collision efficiency in a turbulent
flotation cell”, Chemical Engineering Science, Vol. 64, No. 24,
REFERENCES pp. 5287~5301.
[1] Zhou, J. W., Song, T. and Shen, Z. C., 2010, “CFD [9] Lim, K. H. and Jo, Y. J., 2015, Review of boiling heat
simulation of gas-liquid flow in a large scale flotation cell”, J. transfer models in computational fluid dynamics codes.
of Computational Multiphase Flow, Vol. 2, No. 3, pp.143~150. KINS/RR-1369 (in Korean).
[2] Yiantos, J. B., 2007, “Fluid flow and kinetic modeling in [10] Chapple, D. and Kresta, S. M., 2002, “The effect of
flotation related process: columns and mechanically agitated impeller and tank geometry on power number for a pitched
cells”. Chemical Engineering Research and Design, Vol. 85, blade turbine”. Chemical Engineering Research and Design,
No. 12, pp. 1591~1603. Vol. 80, No. 4, pp. 364~372.
[3] Shi, S. X., Yu, Y., Yang, W. W. and Zhou, H. X., 2013,
“Flow filed test and analysis of KYF Flotation Cell by PIV”,
Applied Mechanics and Materials, Vol. 331, pp. 200~204.

6 Copyright © 2016 by ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/pdfaccess.ashx?url=/data/conferences/asmep/90992/ on 06/28/2017 Terms of Use: http://www.asme.org/abo