Study of Solid-Liquid Mixing in Agitated Tanks Through Computational Fluid Dynamics Modeling
Study of Solid-Liquid Mixing in Agitated Tanks Through Computational Fluid Dynamics Modeling
Solid-liquid mixing is one of the most important mixing operations due to its vast applications in many unit
operations such as crystallization, adsorption, solid-catalyzed reaction, suspension polymerization, and activated
sludge processes. In this study, a computational fluid dynamics (CFD) model was developed for solid-liquid
mixing in a cylindrical tank equipped with a top-entering impeller to investigate the effect of impeller type
(Lightnin A100, A200, and A310), impeller off-bottom clearance (T/6-T/2, where T is tank diameter), impeller
speed (150-800 rpm), particle size (100-900 µm), and particle specific gravity (1.4-6) on the mixing quality.
An Eulerian-Eulerian (EE) approach, standard k-ε model, and multiple reference frames (MRF) techniques
were employed to simulate the two-phase flow, turbulent flow, and impeller rotation, respectively. The impeller
torque, cloud height, and just suspended impeller speed (Njs) computed by the CFD model agreed well with
the experimental data. The validated CFD model was then employed to calculate the solid concentration
profiles by which the degree of homogeneity was quantified as a function of operating conditions and design
parameters.
Experimental Setup
The schematic diagram of the experimental setup used in this
study is shown in Figure 1a. The mixing vessel was a transparent
flat-bottomed cylindrical tank of 40 cm inner diameter (T) and
60 cm height and was filled with solid particles and water to a
height (H) equivalent of the tank diameter (40 cm) occupying
a total volume of about 0.050 m3. To avert vortex formation
around the agitator, the tank was fitted with four equally spaced
baffles having width of 3.4 cm (T/12). To prevent the accumula-
tion of the solid particles behind the baffles, a 0.8 cm (T/50)
gap between each baffle and the tank wall was considered. The
tank was equipped with a top-entering impeller assembly and
impeller speed was set to the desired revolutions per minute
using a variable frequency drive (VFD). Three axial flow
impellers, each with a 17.8 cm diameter (D), were used: Lightnin
A100, A200, and A310 impellers (see Figure 1b). The pumping
direction was downward for all impellers tested in this study.
These axial-flow impellers produce high flow and are more cost-
effective in flow controlled operation like solid suspension.40
Cooke and Heggs41 reported that the hollow blade turbine is
an efficient impeller for the solid-liquid mixing operations
under gassed conditions. The impeller off-bottom clearance (C)
was selected upon the test conditions between T/6 and T/2. A
rotary-torque transducer (Staiger Mohilo, Germany) was em-
ployed to measure impeller torque and speed. In this study, glass
beads and water were used as the solid phase and liquid phase,
respectively. Electrical resistance tomography (Industrial To-
mography Systems, Manchester, UK) was used to measure the
concentration profiles within the mixing vessel as described by
Hosseini et al.42 Eight tomographic sensor planes were posi-
tioned around the circumference of the mixing tank with the
lowest one, numbered plane one, at a height of 4.25 cm from
Figure 1. Experimental setup: (a) mixing tank and (b) impellers used in
base, and the same interval was maintained vertically for this study.
subsequent planes, numbered accordingly in the upward direc-
tion. Each plane had 16 stainless steel sensors which were
situated at equal intervals on the tank periphery. The height, conservation laws of mass and momentum for two-phase flow,
width, and the thickness of the electrodes were 20, 30, and 1 the resulting conservation equations for phase k can be written
mm, respectively. The reference for study was taken based on as follows:
the single ground electrode located between plane four and five.
More details regarding the experimental setup and procedure Continuity equation43
have been reported by Hosseini et al.42 ∂(akFk)
+ ∇ · (akFkuk) ) 0
f
(1)
∂t
Mathematical Model
In the present study, the mathematical model is formulated where F is fluid density, b
u is the velocity vector, R is the volume
based on the Eulerian-Eulerian multifluid model. Applying the fraction, and subscript k symbolizes phase k.
4428 Ind. Eng. Chem. Res., Vol. 49, No. 9, 2010
Where µk, µkt, and kk are the shear viscosity, turbulent viscosity,
and turbulent kinetic energy, respectively. Drag force is
∂
∂t
(Fk) +
∂
∂xi
bi) )
(Fku
∂
∂xj ((
µ+
µt ∂k
σk ∂xj
+) )
expressed by Gk + Gb - Fε - YM(13)
∑ K (u
(( ) )
b
FD ) lk
f
l - uk)
f
(4) ∂ ∂ ∂ µt ∂ε
(Fε) + bi) )
(Fεu µ+ +
∂t ∂xi ∂xj σε ∂xj
where Klk is the exchange coefficient between liquid and the
ε ε2
kth phase. The solid-liquid exchange coefficient (Kpl) is C1ε (Gk + C3εGb) - C2εFε - YM (14)
calculated using the following equation: k k
( )
3RpRlFl Rer f f where Gk, Gb, and YM are the generation of turbulence kinetic
Kpl ) CD |u - ul | (5) energy due to the mean velocity gradients, buoyancy, and the
4utp2dp utp p contribution of the fluctuating dilatation in compressible tur-
bulence to the overall dissipation rate. C1ε, C2ε, and C3ε are
where utp, Rer, CD, and dp are terminal velocity of solid particle, constants (see Table 1).13 In this table, V is the component of
relative Reynolds number (Rer ) Fldp|u bp - bul|/µ), drag coef- the flow velocity parallel to the gravitational vector and u is
ficient, and particle diameter, respectively. The subscripts l and the component of the flow velocity perpendicular to the
p stand for the liquid and solid particles. The following empirical gravitational vector. The terms σk and σε indicate the k and ε
equations were used to calculate CD:44 turbulent Prandtl numbers. The turbulent (or eddy) viscosity
(µt) is defined as43
24
CD ) [1 + 0.15(RlRer)0.687] if Rer e 1000 (6)
RlRer k2
µt ) FCµ (15)
ε
CD ) 0.44 if Rer > 1000 (7)
where Cµ is constant. Gk is defined as
The combination of gravity force and buoyancy force is given
by ∂uj
Gk ) -Fui′uj′ (16)
∂xi
f+Ff ) π d 3(F - F )g
F l b (8)
g B
6 p p where ui′ and uj′ are the fluctuating velocity component due to
turbulence in the directions i and j, respectively, and ui′uj′ are
where dp, Fp, and Fl are particle diameter, particle density, and the Reynolds stresses.
liquid density, respectively. Virtual mass force can be derived CFD Simulation. In this study, a commercial CFD software
from the following:19 package (Fluent V6.3) was used to simulate the mixing of solid
suspension in an agitated vessel by solving the conservation of
f ) - 1 πd 3F d (uf - uf ) mass and momentum equations. The first step is grid generation,
Fvm (9)
12 p l dt p l
which divides the calculation domain into the discrete control
where b up and bul are the particle and liquid velocities, respec- volumes. Gambit 2.4 (Fluent Inc.) was used to discretize the
tively. The lift force or Saffman force, or lift due to shear and flow domain with tetrahedral cells (unstructured grid). The
rotation of the fluid, is given by18,45 advantage of using an unstructured grid is that a complex
geometry (e.g., impeller) can be meshed easily. Mesh refinement
f ) π d 3 Fl C ((uf - uf ) × ω) near the impeller was accomplished using the mesh growth
Fl (10) factor function. This factor controls mesh density by allowing
4 p 2 s l p
the mesh elements to grow slowly as a function of the distance
where ω is the angular velocity and Cs ) 0.1524 for Rep g 40, from the impeller blade to the vessel walls. The grid generated
Cs can be found as follows if Rep < 40:18,45 had skewness smaller than 0.6, indicating a very good mesh
formation. The finer the mesh, the better the result will be
[( ReSR
( )) ( )]
ReSR
0.5 0.5
4.1126 achieved to capture the flow detail, but it should not be so fine,
Cs ) 1.0 - 0.234 e-0.1Rep + 0.234
ReSR0.5 Rep Rep because the computational time increases with reductions in the
(11) mesh size.13 The optimum grid size was obtained by decreasing
the size to a final value below which the changes in the velocity
where Rep is the particle Reynolds number (Rep ) (Flutpdp)/µ) and kinetic energy profiles were less than 3%. The number of
and ReSR is the rotational Reynolds number which can be cells used for A100, A200, and A310 impellers were 311 876,
calculated as follows: 286 870, and 345 046, respectively. The standard wall functions
Ind. Eng. Chem. Res., Vol. 49, No. 9, 2010 4429
13
Table 1. Standard k-ε Constants
C1ε C2ε C3ε Cµ σk σε
1.44 1.92 tanh|V/u| 0.09 1.0 1.3
in Fluent were used for the modeling of the near wall regions.
In this technique, the logarithmic law for mean velocity is known
to be valid for 30 < y* < 300. The y* values calculated in this
study were within these limits. The multiple reference frame
(MRF) technique was employed to model the rotation of the
Figure 2. Impeller torque as a function of impeller speed (A310 impeller,
impeller in the mixing vessel.47 A rotating frame was used for C ) T/3, X ) 10 wt %, dp ) 210 µm, and SG ) 2.5).
the region containing the impeller while a stationary frame was
used for regions that are stationary containing the tank walls
and baffles. The moving reference frame was a cylinder of 24
cm diameter and 9 cm height, and it was centered on the
impeller. This frame did not include the shaft. The governing
equations of the flow domain inside the rotating frame were
solved in the frame of the enclosed impeller while those outside
the rotating frame were solved in the stationary frame. A steady
transfer of information was made at the MRF interface as the
solution progressed. This method facilitates incorporation of the
impeller motion even with a complex geometry. No-slip and
no-penetration conditions were imposed on the transport equa-
tions on the tank walls. Symmetric boundary conditions, hence
zero normal velocity and zero normal gradients for all variables,
were assumed on the liquid surface. Since the shear stress is
zero at a symmetry boundary, it is also called the slip wall
boundary condition. A no-slip boundary condition was applied
to the shaft. The local solid concentrations were monitored
during the simulations. The steady state was achieved when the
fluctuations of the solid concentrations in monitoring points were
not noticeable. The simulation was considered converged when
the scaled residuals for all transport equations were below 10-3.
A single simulation run required 6000-8000 iterations to
converge. The simulations were performed on a 3.0 GHz, 2
GB RAM, Pentium IV which took about 4-5 days for Figure 3. Cloud height (A310 impeller, C ) T/3, X ) 10 wt %, dp ) 210
convergence. The simulations were carried out at the operating µm, and SG ) 2.5): (a) experimental result at N ) 320 rpm, (b) CFD result
at N ) 320 rpm, (c) experimental result at N ) 600 rpm, and (d) CFD
conditions summarized in Table 2. result at N ) 600 rpm.
the fluid at a given impeller speed. The height of this interface
Results and Discussion
from the bottom of the vessel is called cloud height and above
The effects of different parameters (Table 2) on the quality this interface there is only an occasional visit by a few solid
of the solid-liquid mixing were investigated in this work. To particles.48 The CFD model developed in this study was used
validate the model, CFD results for impeller torque, cloud to estimate the cloud height. Solid concentration contours
height, and just suspended impeller speed (Njs) were compared computed on a vertical plane was employed to estimate the cloud
to experimental data. Figure 2 illustrates the impeller torque as height. Figure 3 depicts the cloud height visualized using the
a function of the impeller speed for the A310 impeller. These CFD model and the digital photography at N ) 320 and 600
results show very good agreement between the impeller torque rpm. Both CFD and digital photography techniques were used
calculated by the CFD model and the experimentally determined to obtain the normalized cloud height as a function of the
value. The torque did not change significantly as the solids were impeller speed for A310 impeller (Figure 4). It can be seen that
suspended within the mixing tank for the solid concentration the cloud height was reasonably predicted by the CFD simula-
employed in this study. The power numbers computed using tion. The fluctuation of the cloud height was not significant at
the CFD model for A100, A200, and A310 impellers were 0.39, the steady-state. Figure 4 shows that the deviation of the CFD
1.48, and 0.29, respectively. These results are in good agreement result from the experimental value is more pronounced at 200
with those reported in the literature. rpm. The turbulent and fluid kinetic energy at the lower impeller
Some researchers have derived models to predict the homo- speed lifted a small percentage of the solid particles from the
geneity as a solid cloud height. There is a distinct level (clear bottom of the tank. However, the amount of energy imposed
interface) to which most of the solid particles are lifted within by the impeller was not sufficient to maintain the suspension.
4430 Ind. Eng. Chem. Res., Vol. 49, No. 9, 2010
n
speed was not used for the experimental measurements of the
cloud heights. In fact, we stop agitating the tank after each test ∑ (X V
j V)2
-X
1
and the impeller speed was increased with the same rate for all homogeneity ) 1 - (18)
n
experiments.
A suspension is considered as complete (off-bottom) if no where n, XV and X j V are the number of planes, solid volume
solid particle remains stationary on the bottom of the tank for concentration, and the average solid volume concentration within
more than 1 or 2 s and the impeller speed at this condition is the vessel, respectively. Figure 9 shows the degree of homo-
known as the just suspended impeller speed:49 geneity as a function of impeller power for A310 impeller. The
( )
CFD results are in good agreement with those obtained from
( )
g∆F 0.45 X0.13dp0.2υ0.1 the tomography measurements. As expected, the homogeneity
Njs ) S (17)
FL D0.85 of the system increased with an increase in impeller power/
speed. Once the homogeneity reached the maximum, any further
where S is constant for a given system geometry, ∆F ) Fs - Fl increase in impeller power/speed was not beneficial but detri-
and g, Fs, Fl, υ, dp, X, and D are gravity, particle density, liquid mental. Other researchers also reported the similar phenomenon
density, kinematic viscosity, particle diameter, solid weight in solid-liquid mixing.4,6,52 In order to elucidate the relationship
fraction, and impeller diameter, respectively. In order to calculate between the homogeneity and the impeller speed, the CFD
Njs using the CFD model developed in this study, the average model was used to generate the solid concentration contours as
solid concentration for a horizontal plane located 1 mm above a function of impeller speed (Figure 10). It can be seen that the
the bottom of the tank was measured as a function of the homogeneity within the tank improved with an increase in
impeller speed (Figure 5). The tangents to the curve were drawn impeller speed and the maximum homogeneity was achieved
at the points having maximum and minimum slopes. Njs was at N ) 500 rpm. However, due to the centrifugal force inside
the impeller speed corresponding to the point at which the two the circulation loops at higher impeller speed, the formation of
tangents crossed each other. The similar method was employed the regions with low solid concentrations was observed which
by Mak4 to predict Njs through experimental data. Njs estimated eventually decreased the homogeneity of the system.52 There-
by this technique was 342 rpm which is in good agreement with fore, the optimal impeller speed has a significant effect on the
that calculated using eq 17 (333 rpm). degree of homogeneity and should always be between two
Ind. Eng. Chem. Res., Vol. 49, No. 9, 2010 4431
Figure 8. Axial solid concentration profiles computed using CFD for the
A310 impeller (C ) T/3, X ) 10 wt %, dp ) 210 µm, and SG ) 2.5).
crucial impeller speeds, Njs and the impeller speed for the
maximum homogeneity, preferably closer to the latter.
The type of impeller has a significant effect on the extent of
Figure 6. (a) Solid concentration contours computed using CFD. (b) 3D homogeneity. Axial impellers are preferred for solid-liquid
image of solid concentration generated from the 2D tomography images. mixing because they produce high flow and are more cost-
(c) Velocity vectors (A200 impeller, N ) 250 rpm, C ) T/2, X ) 10 wt %, effective in flow controlled operations like solid suspension.40
dp ) 210 µm, and SG ) 2.5).
In this study, the performances of three axial-flow impellers
4432 Ind. Eng. Chem. Res., Vol. 49, No. 9, 2010
( 4gcdp(Fs - Fl)
)
1/2
Vt ) (19)
3CDFl
Fpdp2
tp ) (20)
18µ
Conclusions
The computational fluid dynamics (CFD) modeling was
successfully employed to explore the effect of operating
conditions and design parameters on the mixing quality for the
solid-liquid mixing operations. This technique provided in-
depth details about the distribution of the solid phase within
the mixing vessel. The impeller torque and cloud height
calculated by the CFD model were in good agreement with the
experimentally determined values. The average solid concentra-
tion at the bottom of the tank as a function of the impeller speed
was determined through CFD to estimate the just suspended
impeller speed (Njs), which was close to that predicted by the
Zwietering’s correlation. The validated CFD model was then
utilized to obtain the axial solid concentration profiles by which
the degree of homogeneity was quantified as a function of the
impeller power/speed, impeller type, impeller off-bottom clear-
ance, particle size, and specific gravity of solid particles. As
expected, the homogeneity of the system increased with an
increase in impeller power/speed. Once the homogeneity reached
Figure 13. (a) Effect of particle size on the degree of homogeneity for the maximum, any further increase in impeller power/speed was
A310 impeller at N ) 400 rpm, X ) 10 wt %, C ) T/3, and SG ) 2.5. (b)
Effect of the solid specific gravity on the degree of homogeneity for A310 not beneficial but detrimental due to the formation of the regions
impeller at N ) 400 rpm, X ) 10 wt %, C ) T/3, and dp ) 210 µm. with low solid concentrations inside the circulation loops at
higher impeller speed. The CFD results for the degree of
homogeneity were in good agreement with those measured by
the electrical resistance tomography. It was found that the A100
impeller was more efficient in terms of homogeneity than the
A310 and A200 impellers and that the optimum impeller
clearance was T/3 for the solid suspension systems. The CFD
results also showed that the physical properties of the solid
particles such as the particle size and the specific gravity
significantly affect the degree of homogeneity in solid-liquid
mixing operations.
Acknowledgment
The financial support of the Natural Sciences and Engineering
Research Council of Canada (NSERC) and Ryerson University
is gratefully acknowledged.
Nomenclature
Figure 14. Particle response time as a function of the particle size (at SG
C ) impeller off-bottom clearance (m)
) 2.5) and the solid specific gravity (at dp ) 210 µm).
CD ) drag coefficient
resistance to flow resulting in lower degree of homogeneity. C1ε, C2ε, C3ε ) constants
The validated CFD model was used to compute the degree of Cs ) Saffman force constant
homogeneity for four different specific gravity values (SG ) Cµ ) turbulent (or eddy) viscosity constant
1.4, 2.5, 4.0, and 6.0) at a fixed impeller speed (N ) 400 rpm) D ) impeller diameter (m)
for the A310 impeller (Figure 13b). These results show that dp ) particle diameter (m)
the specific gravity of the solid phase considerably affects the b
FB ) buoyancy force (N)
4434 Ind. Eng. Chem. Res., Vol. 49, No. 9, 2010
b
FD ) drag force (N) (4) Mak, A. T. C. Solid-Liquid Mixing in Mechanically Agitated Vessels.
b
Fg ) gravity force (N) Ph.D. Dissertation, University of London, London, 1992.
b
Fl ) lift force (N)
(5) Baldi, G.; Conti, R.; Alaria, E. Complete Suspension of Particles in
Mechanically Agitated Vessels. Chem. Eng. Sci. 1978, 33, 21.
b
Fvm ) virtual mass force (N) (6) Einenkel, W. Influence of physical Properties and Equipment Design
g ) gravitational acceleration (m/s2) on the Homogeneity of Suspensions in Agitated Vessels. Ger. Chem. Eng.
Gk ) generation of turbulence kinetic energy due to the mean 1980, 3, 118.
velocity gradients (kg/m · s3) (7) Shamlou, P. A.; Koutsakos, E. Solids Suspension and Distribution
in Liquids under Turbulent Agitation. Chem. Eng. Sci. 1989, 44, 529.
Gb ) generation of turbulence kinetic energy due to buoyancy (kg/ (8) Hicks, M. T.; Myers, K. J.; Bakker, A. Cloud Height in Solids
m · s 3) Suspension Agitation. Chem. Eng. Commun. 1997, 160, 137.
H ) height of solid suspension in the tank (m) (9) Armenante, P. M.; Nagamine, E. U. Effect of Low Off-Bottom
k ) turbulent kinetic energy (m2/s2) Impeller Clearance on the Minimum Agitation Speed for Complete
Suspension of Solids in Stirred Tanks. Chem. Eng. Sci. 1998, 53, 1757.
Klk ) exchange coefficient between liquid and the phase kth (10) Bujalski, W.; Takenaka, K.; Paolini, S.; Jahoda, M.; Paglianti, A.;
Kpl ) solid-liquid exchange coefficient Takahashi, K. Suspension and Liquid Homogenization in High Solids
N ) impeller speed (1/s) Concentration Stirred Chemical Reactors. Chem. Eng. Res. Des. 1999, 77,
n ) number of planes 241.
Njs ) just suspended impeller speed (1/s) (11) Biswas, P. K.; Dev, S. C.; Godiwalla, K. M.; Sivaramakrishnan,
C. S. Effect of Some Design Parameters on the Suspension Characteristics
P ) impeller power (W) of a Mechanically Agitated Sand-Water Slurry System. Mater. Des. 1999,
p ) pressure (Pa) 20, 253.
Rep ) particle Reynolds number (12) Wu, J.; Zhu, Y.; Pullum, L. Impeller Geometry Effect on Velocity
Rer ) relative Reynolds number and Solids Suspension. Chem. Eng. Res. Des. 2001, 79, 989.
(13) Versteeg, H. K.; Malalasekera, W. An Introduction to Computa-
S ) Zwietering constant
tional Fluid Dynamics: The Finite Vol. Method, 2nd ed.; Pearson Education:
SG ) specific gravity of the solid particles New York, 2007.
T ) tank diameter (m) (14) Peker, S. M.; Helvaci, S. S. Solid-Liquid Two Phase Flow; Elsevier:
tp ) particle response time (s) Amsterdam, 2008.
u ) component of the flow velocity perpendicular to the gravita- (15) Sommerfeld, M.; Decker, S. State of the Art and Future Trends in
CFD Simulation of Stirred Vessel Hydrodynamics. Chem. Eng. Technol.
tional vector (m/s) 2004, 27, 215.
b
uk ) phase k velocity (m/s) (16) Decker, S.; Sommerfeld, M. Calculation of Particle Suspension in
b
ul ) liquid velocity (m/s) Agitated Vessels with the Euler-Lagrange Approach. Inst. Chem. Eng. Symp.
b
up ) particle velocity (m/s) Ser. 1996, 140, 71.
(17) Chen, X. Q.; Pereira, J. C. F. Computational Modeling of a Dilute
utp ) terminal velocity of solid particle (m/s)
Turbulent Liquid-Solid Flow Using a Eulerian-Lagrangian Approach. Int.
ui′, uj′ ) fluctuating velocity component due to turbulence in the J. Num. Methods Heat Fluid Flow 2000, 10, 409.
direction i and j, respectively (m/s) (18) Derksen, J. J. Numerical Simulation of Solids Suspension in a
V ) component of the flow velocity parallel to the gravitational Stirred Tank. AIChE J. 2003, 49, 2700.
vector (m/s) (19) Zhang, X.; Ahmadi, G. Eulerian-Lagrangian Simulations of Liquid-
Gas-Solid Flows in Three-Phase Slurry Reactors. Chem. Eng. Sci. 2005,
Vt ) particle-free settling velocity (m/s) 60, 5089.
XV ) volume fraction of solids in suspension (20) Tyagi, M.; Roy, S.; Harvey, A. D., III; Acharya, S. Simulation of
j V ) average volume fraction of solids in suspension
X Laminar and Turbulent Impeller Stirred Tanks Using Immersed Boundary
xi ) particle coordinates (m) Method and Large Eddy Simulation Technique in Multi-Block Curvilinear
Geometries. Chem. Eng. Sci. 2007, 62, 1351.
X ) weight percent of the solid particles
(21) Gidaspow, D. Multiphase Flow and Fluidization: Continuum and
YM ) fluctuating dilatation in compressible turbulence to the overall Kinetic Theory Descriptions; Academic Press: Boston, MA, 1994.
dissipation rate (kg/m · s3) (22) Micale, G.; Montante, G.; Grisafi, F.; Brucato, A.; Godfrey, J. CFD
Simulation of Particle Distribution in Stirred Vessels. Chem. Eng. Res. Des.
Greek Symbols 2000, 78, 435.
Rk ) phase k volume fraction (23) Barrue, H.; Bertrand, J.; Cristol, B.; Xuereb, C. Eulerian Simulation
RL ) liquid phase volume fraction of Dense Solid-Liquid Suspension in Multi-Stage Stirred Vessel. J. Chem.
Eng. Jpn. 2001, 34, 585.
Rp ) particle volume fraction (24) Ranade, V. V. Computational Flow Modeling for Chemical Reactor
ε ) turbulent dissipation rate (m2/s3) Engineering; Academic Press: New York, 2002.
µ ) viscosity (Pa · s) (25) Oshinowo, L. M.; Bakker, A. CFD Modeling of Solids Suspensions
µk ) bulk viscosity (Pa · s) in Stirred Tanks. Presented at the TMS Annual Meeting, Seattle, WA, 2002;
p 215.
µt ) turbulent (or eddy) viscosity (Pa · s) (26) Micale, G.; Grisafi, F.; Rizzuti, L.; Brucato, A. CFD Simulation
ν ) kinematic viscosity of the liquid (m2/s) of Particle Suspension Height in Stirred Vessels. Chem. Eng. Res. Des.
F ) density (kg/m3) 2004, 82, 1204.
Fk ) phase k density (kg/m3) (27) Ochieng, A.; Lewis, A. E. CFD Simulation of Solids Off-Bottom
Fl ) liquid density (kg/m3) Suspension and Cloud Height. Hydrometallurgy 2006, 82, 1.
(28) Cokljat, D.; Slack, M.; Vasquez, S. A.; Bakker, A.; Montante, G.
Fp ) particle density (kg/m3) Reynolds-Stress Model for Eulerian Multiphase. Progress Comput. Fluid
σk, σε ) k and ε turbulent Prandtl numbers Dyn. 2006, 6, 168.
ω ) angular velocity (rad/s) (29) Wang, F.; Mao, Z.; Wang, Y.; Yang, C. Measurement of Phase
Holdups in Liquid-Liquid-Solid Three-Phase Stirred Tanks and CFD
Literature Cited Simulation. Chem. Eng. Sci. 2006, 61, 7535.
(30) Launder, B. E.; Spalding, D. B. The Numerical Computation of
(1) Tatterson, G. B. Fluid Mixing and Gas Dispersion in Agitated Tanks; Turbulent Flows. Comput. Meth. Appl. Mech. Eng. 1974, 3, 269.
McGraw-Hill: New York, 1991. (31) Altway, A.; Setyawan, H.; Margono, M.; Winardi, S. Effect of
(2) Paul, E. L.; Atiemo-Obeng, V. A.; Kresta, S. M. Handbook of Particle Size on Simulation of Three-Dimensional Solid Dispersion in Stirred
Industrial Mixing: Science and Practice; Wiley-Interscience: New York, Tank. Chem. Eng. Res. Des. 2001, 79, 1011.
2004. (32) Montante, G.; Micale, G.; Magelli, F.; Brucato, A. Experiments
(3) Zlokarnik, M. Stirring: Theory and Practice; Wiley-VCH: New York, and CFD Predictions of Solid Particle Distribution in a Vessel Agitated
2001. with Four Pitched Blade Turbines. Chem. Eng. Res. Des. 2001, 79, 1005.
Ind. Eng. Chem. Res., Vol. 49, No. 9, 2010 4435
(33) Ljungqvist, M.; Rasmuson, A. Numerical Simulation of the Two- (45) Saffman, P. G. The Lift on a Small Sphere in a Slow Shear Flow.
Phase Flow in an Axially Stirred Vessel. Chem. Eng. Res. Des. 2001, 79, J. Fluid Mech. 1965, 22, 385.
533. (46) Launder, B. E.; Spalding, D. B. Lectures in Mathematical Models
(34) Sha, Z. L.; Palosaari, S.; Oinas, P.; Ogawa, K. CFD Simulation of of Turbulence; Academic Press: London, 1972.
Solid Suspension in a Stirred Tank. J. Chem. Eng. Jpn. 2001, 34, 621. (47) Luo, J. Y.; Issa, R. I.; Gosman, A. D. Prediction of Impeller Induced
(35) Ranade, V. V.; Tayalia, Y.; Krishnan, H. CFD Predictions of Flow Flows in Mixing Vessels Using Multiple Frames of Reference. Inst. Chem.
near Impeller Blades in Baffled Stirred Vessels: Assessment of Computa- Eng. Symp. Ser. 1994, 136, 549.
tional Snapshot Approach. Chem. Eng. Commun. 2002, 189, 895. (48) Bittorf, K. J.; Kresta, S. M. Prediction of Cloud Height for Solid
(36) Kee, N. C. S.; Tan, R. B. H. CFD Simulation of Solids Suspension Suspensions in Stirred Tanks. Chem. Eng. Res. Des. 2003, 81, 568.
in Mixing Vessels. Can. J. Chem. Eng. 2002, 80, 721. (49) Zwietering, T. N. Suspending of Solid Particles in Liquid by
(37) Khopkar, A. R.; Kasat, G. R.; Pandit, A. B.; Ranade, V. V. Agitators. Chem. Eng. Sci. 1958, 8, 244.
Computational Fluid Dynamics Simulation of the Solid Suspension in a (50) Fradette, L.; Tanguy, P. A.; Bertrand, F.; Thibault, F.; Ritz, J.;
Stirred Slurry Reactor. Ind. Eng. Chem. Res. 2006, 45, 4416. Giraud, E. CFD Phenomenological Model of Solid-Liquid Mixing in Stirred
(38) Kasat, G. R.; Khopkar, A. R.; Ranade, V. V.; Pandit, A. B. CFD Vessels. Comput. Chem. Eng. 2007, 31, 334.
Simulation of Liquid-Phase Mixing in Solid-Liquid Stirred Reactor. Chem. (51) Kresta, S. M.; Wood, P. E. The Mean Flow Field Produced by a
Eng. Sci. 2008, 63, 3877. 45-Degree Pitched Blade Turbine: Changes in the Circulation Pattern Due
(39) Tamburini, A.; Cipollina, A.; Micale, G.; Ciofalo, M.; Brucato, A. to Off-Bottom Clearance. Can. J. Chem. Eng. 1993, 71, 42.
Dense Solid-Liquid Off-Bottom Suspension Dynamics: Simulation and (52) Bohnet, M.; Niesmak, G. Distribution of Solids in Stirred Suspen-
Experiment. Chem. Eng. Res. Des. 2009, 87, 587. sion. Ger. Chem. Eng. 1980, 3, 57.
(40) McDonough, R. J. Mixing for the Process Industries; Van Nostrand (53) Perry, R. H.; Green, D. Perry’s Chemical Engineers’ Handbook;
Reinhold: New York, 1992. McGraw-Hill: New York, 1984.
(41) Cooke, M.; Heggs, P. J. Advantages of the Hollow (Concave)
(54) Guiraud, P.; Costes, J.; Bertrand, J. Local Measurements of Fluid
Turbine for Multi-Phase Agitation under Intense Operating Conditions.
and Particle Velocities in a Stirred Suspension. Chem. Eng. J. 1997, 68,
Chem. Eng. Sci. 2005, 60, 5529.
75.
(42) Hosseini, S.; Patel, D.; Ein-Mozaffari, F.; Mehrvar, M. Study of
Solid-Liquid Mixing in Agitated Tanks through Electrical Resistance (55) Godfrey, J. C.; Zhu, Z. M. Measurement of Particle-Liquid Profiles
Tomography. Chem. Eng. Sci. 2010, 65, 1374. in Agitated Tanks. AIChE Symp. Ser. 1994, 299, 181.
(43) Bird, R. B.; Stewart, W. E.; Lightfoot, E. N. Transport Phenomena;
2nd ed.; John Wiley: New York, 2002. ReceiVed for reView July 13, 2009
(44) Gidaspow, D.; Bezburuah, R.; Ding, J. Hydrodynamics of Circulat- ReVised manuscript receiVed February 27, 2010
ing Fluidized Beds, Kinetic Theory Approach. Presented at the 7th Accepted March 17, 2010
Engineering Foundation Conference on Fluidization, Brisbane, Australia,
1992; p 75. IE901130Z