This article was downloaded by: [University of Saskatchewan Library]

On: 08 January 2015, At: 11:14

Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House,
37-41 Mortimer Street, London W1T 3JH, UK

Drying Technology: An International Journal

Publication details, including instructions for authors and subscription information:
http://www.tandfonline.com/loi/ldrt20

Innovative Hydrocyclone Inlet Designs to Reduce

Erosion-Induced Wear in Mineral Dewatering Processes
a b a a b
Peng Xu , Z. Wu , A. S. Mujumdar & B. Yu
a
Department of Mechanical Engineering , National University of Singapore , Singapore
b
Department of Physics , Huazhong University of Science and Technology , Hubei, P.R. China
Published online: 09 Feb 2009.

To cite this article: Peng Xu , Z. Wu , A. S. Mujumdar & B. Yu (2009) Innovative Hydrocyclone Inlet Designs to Reduce Erosion-
Induced Wear in Mineral Dewatering Processes, Drying Technology: An International Journal, 27:2, 201-211

To link to this article: http://dx.doi.org/10.1080/07373930802603433

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained
in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the
Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and
should be independently verified with primary sources of information. Taylor and Francis shall not be liable for
any losses, actions, claims, proceedings, demands, costs, expenses, damages, and other liabilities whatsoever
or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of
the Content.

This article may be used for research, teaching, and private study purposes. Any substantial or systematic
reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any
form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://
www.tandfonline.com/page/terms-and-conditions
 Drying Technology, 27: 201–211, 2009
Copyright # 2009 Taylor & Francis Group, LLC
ISSN: 0737-3937 print/1532-2300 online
DOI: 10.1080/07373930802603433

Innovative Hydrocyclone Inlet Designs to Reduce

Erosion-Induced Wear in Mineral Dewatering Processes
Peng Xu,1,2 Z. Wu,1 A. S. Mujumdar,1 and B. Yu2
1
Department of Mechanical Engineering, National University of Singapore, Singapore
2
Department of Physics, Huazhong University of Science and Technology, Hubei, P.R. China
Downloaded by [University of Saskatchewan Library] at 11:14 08 January 2015

escalating energy costs. In this work, we focus on the

The hydrocyclone is a mechanical separation device that is used application of the hydrocyclone in mineral processing with
widely in mineral processing. The solid particles in mineral slurries a CFD model specially constructed to predict erosion rates.
are separated according to their density, size, and shape by the We expect such a model to assist as a design and analysis
centrifugal force generated by an induced vortex motion in a cylin- tool to arrive at better designs that suffer lower erosion
der-on-cone vessel. The larger and denser particles move closer to
the wall region due to their greater inertia and descend by gravity; rates. Although clearly very important from an economic
a higher concentration suspension is thus collected at the bottom of standpoint, this area has received little attention in the
the hydrocyclone. The cleaned liquid and hydrodynamically smaller literature on hydrocyclones.
particles exit through an overflow outlet at the top of the hydrocy- Although many efforts have been made to study exper-
clone. Higher velocities, within limit, generally yield higher collec- imentally the flow in hydrocyclones,[1–3] little detailed flow
tion efficiency. However, higher stream velocities cause severe
erosion of the internal wall of hydrocyclones as mineral slurries information was generated until the availability of laser
generally are abrasive. The objective of this study is to use the com- Doppler anemometry (LDA) necessary to measure local
putational fluid dynamic (CFD) technique to model the turbulent fluid velocities within the hydrocyclone.[4–6] It is well
swirling flow and predict regions of significant wear and how they known that the LDA technique is expensive, largely limited
are influenced by design of the inlet ducting. New inlet designs are to the dispersed liquid phase, and at present suitable only
proposed and investigated numerically for their erosion characteris-
tics, pumping power requirements, and collection efficiency. for laboratory-scale studies rather than industrial design
Successful innovation in hydrocyclone design will lead to and application. Hence, empirical equations=models are
reduced maintenance and lower operating energy costs in mineral often used by designers of hydrocyclones. However,
processing. empirical equations=models suffer from their inherent
deficiency as they can only be used within the limits of
Keywords Air core; Computational fluid dynamics; Erosion the experimental data from which the empirical parameters
rate; Hydrocyclone; Turbulence model were determined. No innovative changes can be predicted
with empirical modeling. In view of these shortcomings,
mathematical models based on the basic fluid mechanics
INTRODUCTION
are highly desirable to intensify innovation.
The hydrocyclone is an important and popular indus- The CFD technique is gaining popularity in process
trial apparatus used to separate by centrifugal action dis- design and optimization as it provides a good means of
persed solid particles from a liquid suspension fed to it. predicting equipment performance of the hydrocyclone
It is widely used in industry, particularly in the mineral under a wide range of geometric and operating conditions;
and chemical processing industries because of its simplicity it also offers an effective way to design and optimize the
in design and operation, high capacity, low maintenance hydrocyclones. The first CFD model utilizing the standard
and operating costs, as well as its small physical size.[1,2] k
e turbulence model for a hydrocyclone was developed
Recently, the application of a hydrocyclone in the wet by Boysan et al.[7] The k
e turbulence model intrinsically
process of beneficiation has attracted more interest with makes the assumption that the turbulence is isotropic
the current boom in advanced mineral processing driven because only one scalar velocity fluctuation is modeled.
by the expanding mineral commodities markets and Furthermore, the Bousinessq approximation on which the
eddy viscosity intrinsically relies implies equilibrium
Correspondence: Peng Xu, Department of Mechanical between stress and strain. Therefore, the k
e turbulent
Engineering, National University of Singapore, 9 Engineering
Drive 1, Singapore 117576; E-mail: [email protected] model is not suitable to simulate a turbulent flow with high

201
 202 XU ET AL.

swirl, flow reversal, and=or flow separation.[8] Many The latter includes heat and mass transfer, which is not
authors have adopted the renormalization group (RNG) present in classification in a hydrocyclone and is a purely
k
e model with a swirl correction to enhance the precision fluid dynamic problem. Interested readers are referred to
of simulations because it includes additional terms for the work of Huang et al., who provide some useful insight
the dissipation rate, e, development, which significantly into the relative performance of various turbulence models
improves the accuracy for rapidly strained flows.[9–12] applied to swirling two-phase flow.[27–29]
However, Suasnabar found that the swirl constant in the Erosion of parts of the internal wall of the hydrocyclone
RNG k
e model needs to be increased to improve predic- is a critical issue in mineral dewatering both from safety
tions; this in turn can cause numerical instability.[13] There- and economic considerations. The injected solids particles,
fore, the application of the RNG k
e model is also limited such as sand and ore particles, impinge the inside surfaces
for modeling a hydrocyclone. The Reynolds stress model of the components of the hydrocyclone, causing mechan-
(RSM) solves the transport equation for each individual ical wear and eventual failure of the devices. Therefore,
Reynolds stress, which enables RSM to model anisotropic the ultimate goal in the design of a hydrocyclone is not
turbulence and strained flows where the Bousinessq only to provide better performance but also resistance
approximation is known to be invalid. Some recent studies to wear. As testing for erosion of industrial devices
Downloaded by [University of Saskatchewan Library] at 11:14 08 January 2015

indicate that RSM can improve accuracy of the numerical generally requires special equipment and methodology,
solution.[14–19] However, the predictions are not what they CFD modeling has been widely adopted as an effective tool
could be and there is debate about appropriate modeling to predict the wear response at low price and without costly
options. Recent advances in computational power have experimentation.[30–34] Although CFD modeling has been
begun to make large eddy simulation (LES) practical for tested to estimate the erosion rate in a hydrocyclone
engineering problems. LES is intrinsically a dynamic simu- with one novel inlet design, the results found in the litera-
lation and can capture time-dependent vortex oscillations ture are very few and largely preliminary; they do not
and nonequilibrium turbulence, which suggest that it indicate in detail the effect of flow and geometry on the
should be appropriate for modeling hydrocyclone. Com- erosion rate.[35] Hence, further modeling efforts are
pared with other turbulence models, LES provides better needed for advancing our capability in predicting wear of
predictions including velocity, air core, and separation hydrocyclones.
efficiency.[16,17,20,21] Certainly, LES simulations of large This work presents a CFD model of a hydrocyclone
industrial dense medium hydrocyclones will be computa- based on Fluent version 6.3. First, results using different
tionally impractical except for a test case.[17] turbulence models, viz. k
e, RSM, and LES, are com-
A striking feature of the flow field is the presence of an pared with published experimental results for a 75-mm
air core in the hydrocyclone. At the central axis of the standard hydrocyclone.[36] The air core geometry is pre-
hydrocyclone, the low pressure developed supports forma- dicted using a VOF multiphase model. Then, the erosion
tion of a rotating air column called the air core. The rate for four designs of a 75-mm hydrocyclone fitted with
geometry and the stability of the air core have been empiri- different inlets is calculated. This article is organized into
cally found to have a strong influence on the performance six sections. The following describes the CFD model
of the hydrocyclone.[2,17] In the mineral processing field, including the turbulence models, the multiphase model,
the air core dimension is a critical variable, since a large the particle tracking model, and the erosion model. The
air core diameter leads to a condition known as ‘‘roping.’’ next section explains the numerical methodology and initial
As the interface between liquid and air phases is difficult to and boundary conditions, followed by a comparison of
determine, the nature of the air core has often been simulated results with experimental data. This is followed
neglected in many previous CFD modeling efforts.[20,22] by an illustration of the predicted erosion rates and a dis-
While some others made a simplified assumption about cussion of the influence of inlet geometry on the erosion
its formation and behavior.[23,24] The volume-of-fluid rate in the modeled hydrocyclone. The last section includes
(VOF) model has been proven to be an effective method discussion with conclusions and future prospects.
for modeling the air core and predicting the formation
and shape of air core.[14–18,21,25] For solids movement, a MODEL DESCRIPTION
Eulerian-Lagrangian model has been successfully applied
to dilute flow phenomena in a hydrocyclone.[8] In the last Turbulence Model
several decades, this field had advanced so much that Slack The free surface, air core, and presence of solid particles
et al. proposed an automated tool for novice analysts to make the swirling turbulent flow highly anisotropic; this
carry out simulations with standard CFD solvers.[26] adds to the difficulty for modeling hydrocyclones using
There is analogy between CFD modeling of liquid- CFD. Three kinds of turbulence models, k
e model,
particle swirling flow in a hydrocyclone and gas-droplet= RSM, and LES, are often adopted for modeling the
particle flow that occurs in a cylinder-on-cone spray dryer. turbulent flow in hydrocyclones. The 75-mm standard
 INNOVATIVE HYDROCYCLONE INLET DESIGNS 203

hydrocyclone is adopted to validate the current CFD unknown and require modeling. The simplest subgrid-scale
models since relevant experimental data are available in model is the Smagorinsky-Lilly model (SLM). This model
the literature. proposes that the subgrid-scale eddy viscosity is related
In mineral processing, the fluid suspensions are gener- to the local average grid spacing and the mean strain rate.
ally dilute (<10%); thus, the incompressible Navier-Stokes The Smagorinsky constant is not a universal constant,
equations supplemented by a suitable turbulence model which is the most serious shortcoming of this simple model.
are appropriate for modeling the flow in hydrocyclones. Brennan[16] and Narasimha et al.[17] adopted the SLM sub-
The k
e model is a semi-empirical model based on model grid-scale model and took the Smagorinsky constant as the
transport equations for turbulence kinetic energy k and its default value 0.1 in their large eddy simulation of hydro-
dissipation rate e with the assumption that the flow is fully cyclone, and their numerical results give good agreement
turbulent and the effects of molecular viscosity are negli- with experimental results. The renormalization group
gible. Compared with standard k
e model, the RNG (RNG) subgrid-scale model is very effective to model the
k
e model includes additional terms for dissipation rate low-Reynolds-number effects encountered in transitional
e development, which can significantly improve the accu- flows and near-wall regions where the molecular viscosity
racy for rapidly strained flows. The RNG k
e model also has more significance. Delgadillo et al. have used the
Downloaded by [University of Saskatchewan Library] at 11:14 08 January 2015

provides an option to account for the effects of swirl or RNG subgrid-scale model in the large eddy simulation of
rotation by modifying the turbulent viscosity appropri- hydrocyclone,[21,25,37] and they stated that the RNG LES
ately. Overall, the RNG k
e model is more responsive model captures the dynamics of the flow in hydrocyclone
to the effects of rapid strain and streamline curvature than and they have also proven that CRNG ¼ 0.157 is the most
the standard k
e model is; this explains the superior per- acceptable value.[37] It should be pointed out that the
formance of the RNG k
e model for the highly swirling LES model requires highly accurate spatial and temporal
flow in a hydrocyclone. Furthermore, the adoption of discretization, finer mesh than a comparable RANS simu-
effective viscosity allows the RNG k
e model to better lation, and more computer resources. And there are still
handle low Reynolds number and near wall flows. unanswered questions about appropriate subgrid-scale
The Reynolds stress model (RSM) closes the Reynolds- models for the multiphase flows encountered in mineral
averaged Navier-Stokes equations (RANS) by solving processing applications.
transport equations for the individual Reynolds stresses Therefore, four CFD models, RNG k
e, QPS RSM,
without isotropic eddy-viscosity hypothesis and together SLM, and RNG LES will be performed in a 75 mm stan-
with an equation for the dissipation rate. Since the RSM dard hydrocyclone. The numerical results will be compared
accounts for the effects of streamline curvature, swirl, with each other and that of experiment.
rotation, and rapid changes in strain rate in a more
rigorous manner, it has greater potential to give accurate Multiphase Model
predictions for complex flows in hydrocyclone. The RSM In the hydrocyclone, the centrifugal force generated by
has been proven to be an appropriate turbulence model the tangential acceleration pushes the fluid to the wall
for a hydrocyclone.[14,15,17–19] In the simulation process of and creates a low pressure in the central axis, which gives
a hydrocyclone, the standard linear pressure strain (LPS) the right conditions to suck air into the device and form
model and quadratic pressure strain (QPS) model are often an air core. In the current simulation, the formation pro-
adopted. Narasimha et al. stated that the constants in the cess, location, and shape of the air core will be taken into
LPS correlation need to be adjusted to match the velocity account instead of neglecting it or making simple assump-
predictions with data.[17] Comparing with the default tions. The volume of fluid (VOF) free surface model is a
LPS model, the QPS model has been demonstrated to give simplification of the mixture model and has been success-
superior performance in a range of basic shear flow, includ- fully applied in predicting and modeling the air core in a
ing plane strain, rotating plane shear, and axisymmetric hydrocyclone.[14–18,21,25] The VOF model can simulate
expansion=contraction. This improved accuracy should two or more immiscible fluid phases, in which the position
be beneficial for the complex flows in hydrocyclone.[14] of the interface between the fluids is of interest. In the VOF
Therefore, the QPS RSM is chosen in the current method, the variable density equations of motion are
simulation. solved for the mixture, and an additional transport equa-
Large eddy simulation (LES) provides an alternative tion for the volume fraction of each phase is solved, which
approach in which large eddies are explicitly resolved in a can track the interface between the air core and the liquid
time-dependent simulation using the filtered Navier-Stokes in a hydrocyclone. The single momentum equation is
equations. Velocity profiles of LES are resolved by a filter- solved throughout the domain, and the resulting velocity
ing operation of the velocity field, and the smaller scales or field is shared among the phases.
residuals are modeled in a particular manner. The subgrid- The simple VOF model has been adopted by most
scale stresses resulting from the filtering operation are engineers for modeling the air core in a hydrocyclone.
 204 XU ET AL.

And the comparison of a Eulerian multiphase model and however, the default values are not updated to reflect the
VOF by Brennan indicates that the velocity predictions material being used. Therefore, these parameters have to
using the two models are essentially the same.[16] Therefore, be updated for different materials. It is known that one
the air core will be resolved with a VOF model in the cur- of the main parameters that influences the erosion rate is
rent simulation. But for the dense slurry, the more sophis- the particles impingement angle. The impingement angle
ticated Eulerian multiphase model will be more suitable. function can be used as the following model and defined
by a piece-linear profile:[31,39]
Particle Tracking
In most mineral processing operations, the solid phase is f ðaÞ ¼ sinð2aÞ
3 sin2 ðaÞ for a  18:43o ð2aÞ
sufficiently dilute (<10%). Hence, we can employ the dis- 2
f ðaÞ ¼ cos ðaÞ=3 for a > 18:43 o
ð2bÞ
crete phase model (DPM), the fundamental assumption
of which is that the dispersed second phase occupies a To calculate the erosion rate from Eq. (1), the diameter
low volume fraction. The Lagrangian DPM follows the function and velocity exponent function are adopted as
Euler-Lagrange approach. The fluid phase is treated as a 1.8E-09 and 1.73.[38,40,41] The CFD model records the num-
ber, velocity, mass, and impact angle of the various parti-
Downloaded by [University of Saskatchewan Library] at 11:14 08 January 2015

continuum by solving the time-averaged Navier-Stokes

equations, whereas the dispersed phase is solved by track- cles for each of the grids that form the internal geometry
ing a large number of particles through the calculated flow of the hydrocyclone. Then, the erosion rate of the hydro-
field. The dispersed phase can exchange momentum, mass, cyclone walls is determined using Eqs. (1) and (2). There
and energy with the fluid phase. are many parameters affecting the erosion rate, such as flow
The dispersion of particles can be accounted for with a rate, design of the inlet, geometry, and dimensions of the
stochastic tracking model, in which the turbulent disper- hydrocyclone and slurry properties, etc., among which the
sion of particles is predicted by integrating the trajectory inlet has a very important effect on the wear characteristics
equations for individual particles and using the instan- of hydrocyclone. Thus, as a preliminary work, we will
taneous fluid velocity. Also, unsteady tracking is used, calculate erosion rate for a hydrocyclone with four different
where at the end of each time step the trajectory is updated inlets and discuss the influence of the design of inlet ducting
with the instantaneous velocity. As for the slurry feed con- on wear characteristics of hydrocyclone.
centrations in excess of 10% by volume, the DPM is not
suitable and the Eulerian multiphase model is more
appropriate for tracking particles in a hydrocyclone. SIMULATION RESULTS
The simulations are performed using Fluent CFD soft-
Erosion Model ware package (version 6.3.26). As noted earlier, published
The impingement of solid particles with hydrocyclone experimental data on a 75 mm standard hydrocyclone are
walls can cause considerable wear, which is an issue of great used to validate the CFD model. Figure 1 shows the dimen-
industrial concern, both from safety and economic consid- sions and computational mesh for the test hydrocyclone;
erations. The damage induced by the erosion can cause Table 1 lists its geometrical parameters. The geometry used
equipment failure. Hence, estimation of potential erosion in the simulation is the same as that of the experiment;[36]
of the hydrocyclone wall is important to predict the lifetime also the same area of quadrate inlet is used. Four turbu-
of the equipment; it is useful to know how it is affected by lence models, RNG k
e, QPS RSM, SLM, and RNG
geometry and different operating conditions. Because of LES were tested. In the simulation, the velocity inlet
experimental difficulties, CFD analysis is an effective tool boundary condition and pressure outlet boundary con-
to investigate the erosion rate of hydrocyclone. ditions are applied for the vortex finder and the spigot.
Particle erosion and accretion rates can be computed at The inlet flow rate is 1.12 kg=s and the pressure at the
wall boundaries based on the following model equations. two outlets is 1 atm. The physical constants for the liquid
The erosion rate is defined as:[38] phase were set as those of water. The solid particle density
is 2700 kg=m3 and its mass fraction is 4.8% at the inlet. The
X
N
_ p Cðdp Þf ðaÞvbðvÞ
m unsteady solver with time steps 10
4 to 10
6 s was applied
Rerosion ¼ ð1Þ
A in the simulations. The second-order upwinding and
p¼1
SIMPLE pressure-velocity coupling algorithm were
where Cðdp Þ is a function of particle diameter, a is the adopted. The flow field reaches steady state after about
impact angle of the particle path with the wall face, f (a) 2 s; therefore, the simulated results presented here are those
is a function of impact angle, v is the relative velocity, b for 3 s after simulation starts; i.e., when the flow is steady.
(n) is a function of relative particle velocity, and A is the The flow problem is simulated with three-dimensional
area of the cell face at the wall. The three functions C, f, unstructured mesh of hexahedral cells. Trial numerical
and b can be defined as boundary conditions at the wall; results indicated that the solution is independent of the
 INNOVATIVE HYDROCYCLONE INLET DESIGNS 205

Fluid Flow Field

Figure 1a shows several representative streamlines in a
75 mm standard hydrocyclone. It clearly indicates a swirl-
ing flow pattern and splitting overflow and underflow in
hydrocyclone. In order to explore the inner flow field in
a hydrocyclone, three different horizontal planes situated
60, 120, and 170 mm from the top wall of a 75 mm standard
hydrocyclone are selected to give a general description of
velocity field. On each plane, the axial and tangential velo-
city profiles are compared with those of the experimental
results.
Figures 2a and 2b show the axial and tangential velocity
profiles at 60 mm from the top wall of the hydrocyclone,
where four numerical results using RNG k
e, QPS
Downloaded by [University of Saskatchewan Library] at 11:14 08 January 2015

RSM, SLM LES, and RNG LES turbulence models are

compared with the experimental results. The four results
are all close to the experimental ones, but the QPS RSM,
SLM LES, and RNG LES computations track the axial
and tangential velocities more closely than RNG k
e. It
can be seen clearly that the predicted axial and tangential
FIG. 1. (a) Schematic dimensions of the standard hydrocyclone with velocities near the wall are near zero because the no-slip
stream lines; (b) grid representation used in simulation. boundary condition is used for the walls. For the axial
velocity near the wall, there is little difference between
the numerical results and the experimental ones, whereas
characteristics of the mesh size. Combined with the VOF for tangential velocity, the experimental value is nearly
model, the air core can be predicted with each one of the zero, which is the same as the numerical result. Also,
four turbulence models for comparison. The erosion rate because of the air core, the axial velocity near the central
of the hydrocyclone with four different inlets was also core of the hydrocyclone is quite high, whereas the tangen-
predicted with the RNG LES model. tial velocity approaches 0, which is consistent with the
experimental results and the physical conditions.
Figures 2c and 2d describe the axial and tangential velo-
city profiles in the 120 mm plane. The difference between
MODEL VALIDATION the predicted velocities with RNG k
e model and those
In this section, the simulated flow field, air core, and of experiment is so large that it may be concluded that the
particle classification results are compared with experiment RNG k
e model is not suitable for modeling the turbulent
to validate the model. Moreover, a comparison between flow in a hydrocyclone. The predicted results of QPS RSM,
the experimental and the numerical results is also made SLM LES, and RNG LES models are all similar to each
for other variables of the hydrocyclone; e.g., pressure drop other and rather close to the experimental values. The
and volume split ratio. predicted behavior near the wall and at the center of hydro-
cyclone is the same as that along the 60 mm plane.
Figures 2e and 2f display the axial and tangential velo-
TABLE 1 city profiles in the 170-mm plane. The predicted axial velo-
Geometry of the 75-mm standard hydrocyclone modeled city with RNG k
e model is contrary to the experimental
Parameter Dimension results (see Fig. 2e). This can be explained from the fact
that the flow near the central zone of the hydrocyclone is
Diameter of hydrocyclone 75 mm drained off due to the underflow because of the disappear-
Diameter of the inlet 25 mm ance of the air core (see Fig. 3); hence, the predicted axial
Diameter of the vortex finder 25 mm velocity approaches negative values. The close perfor-
Diameter of the spigot 12.5 mm mances of QPS RSM, SML LES, and RNG LES models
Length of the cylindrical section 75 mm can be found in Figs. 2e and 2f, which can better predict
Length of the conical section 186 mm the turbulent flow in the hydrocyclone.
Length of vortex finder 50 mm From the above comparisons, it is seen that the predicted
Cone angle 20 axial and tangential velocities of the RNG k
e turbulence
 206 XU ET AL.
Downloaded by [University of Saskatchewan Library] at 11:14 08 January 2015

FIG. 2. Axial and tangential velocity profile—comparison with experimental results at (a) and (b) 60 mm, (c) and (d) 120 mm, and (e) and (f) 170 mm
from the top wall of 75-mm standard hydrocyclone.

model are far from the experimental results, whereas the RSM and SML LES models perform better near the center,
performances of QPS RSM, SML LES, and RNG LES the RNG LES model can track the turbulent velocities near
models are close to each other and the experimental results. the wall better. Furthermore, the absolute error is small
Therefore, we conclude that the RNG k
e model is not for the axial velocity and nearly zero for tangential velocity
suitable for modeling hydrocyclones, whereas QPS RSM, near wall. Although the no-slip boundary condition is
SML LES, and RNG LES models can capture the velocity adopted by most of current CFD models, a more accurate
profiles at different locations of the flow and can be used to wall boundary condition with suitable boundary layer mesh
model the hydrocyclone. Comparison between the latter can be applied to improve the simulation results considering
three turbulence models indicates that although the QPS the error in axial velocity. Another point should be noted is
 INNOVATIVE HYDROCYCLONE INLET DESIGNS 207
Downloaded by [University of Saskatchewan Library] at 11:14 08 January 2015

FIG. 3. The predicted air volume fraction and air core formation with FIG. 4. The predicted air volume fraction and air core formation with
RNG k-e turbulence model from time 0–3 s. QPS RSM turbulence model from time 0–3 s.

that the QRS RSM turbulence model combined with a VOF disappears; the predicted air core diameter with RNG
multiphase model can lead to numerical stability, whereas k
e model is 0.2. The numerically computed air core
the LES model consumes significantly more computing diameter of Delgadillo et al. with RNG k
e model is
resources and times. 0,[21,37] which is much closer to the current results. Thus,
the RNG k
e model can not predict the air core in a
hydrocyclone, which is its major weakness.
Formation of the Air Core Figure 4 exhibits the air core formation with the QPS
The ability to predict well the development of the air core RSM turbulence model. The air core is formed in the first
in the hydrocyclone is a test of the CFD model. Figures 3–6 second and remains steady after 2 s. A nearly parabolic
show the predicted air core formation with RNG k
e, shape of the air core is well predicted by QPS RSM turbu-
QPS RSM, SLM LES, and RNG LES turbulence models, lence model, which agrees well with the experimental data.
respectively. Figure 3 indicates the evolution of the air The predicted air core diameter is about 10.6 mm, which is
volume fraction with the RNG k
e model at real times very close to the experimental value of 10 mm. Figures 5
from 0.02 to 3.0 s. The variable of air volume fraction in and 6 show the air core development with SLM and
the first second is distinct and then the air fraction changes RNG LES turbulence models. The similar formation pro-
are very minor. After the first second, a small change in the cess as QPS RSM model can be found in Figs. 5 and 6.
air volume fraction can only be found at the top end of Compared with the air core with RSM model, the shape
hydrocyclone in the following one second. Between 2 to of air core with LES is more regular and closer to the
3 s, there is no variability of the air volume fraction. experimental shape. The predicted diameters are 11.5 and
Further simulations indicate that the flow varies little after 10.45 mm for SLM LES and RNG LES, respectively.
1 s and reaches steady state after 2 s. It can be seen from The error of SLM LES is larger than 10%, and that of
Fig. 3 that the air core is formed at about 0.6 s and then RNG LES is less than 5%.
 208 XU ET AL.
Downloaded by [University of Saskatchewan Library] at 11:14 08 January 2015

FIG. 5. The predicted air volume fraction and air core formation with FIG. 6. The predicted air volume fraction and air core formation with
SLM LES turbulence model from time 0–3 s. RNG LES turbulence model from time 0–3 s.

The general mass balances are also calculated and com- Considering numerical stability of RSM and relatively high
pared with experiments as listed in Table 2. The experi- accuracy near the wall of RNG LES turbulence model, the
mental split ratio of the 75-mm standard hydrocyclone is RNG LES will be adopted in calculating the erosion rate of
95.1%, whereas the predicted split ratios with the RNG a hydrocyclone.
k
e, QPS RSM, SLM LES, and RNG LES turbulence
models are 78.75, 95.7, 95.6, and 92%, respectively. For EROSION RATE
pressure drop, the experimental result is 46.7 kPa, and the The centrifugal effect is one of major affection for
numerical results are 38.3, 41.1, 40.2, and 38.4 kPa, respect- separation efficiency of particles,[42] while the geometry of
ively. The experimental and numerical air core diameters inlet has important influences on centrifugal effect. The
are also listed in Table 2. In all, QPS RSM, SLM LES, geometry of the feed inlet can also affect energy consump-
and RNG LES can be used for modeling a hydrocyclone. tion.[43] Furthermore, the geometry of the feed inlet is an

TABLE 2
General mass balance for four different turbulent models
Experiment RNG k
e QPS RSM SLM LES RNG LES
Feed flow rate (kg=s) 1.117 1.12 1.12 1.12 1.12
Overflow flow rate (kg=s) 1.062 0.882 1.072 1.071 1.03
Underflow flow rate (kg=s) 0.055 0.238 0.058 0.053 0.09
Split ratio (%) 95.1 78.75 95.7 95.6 92.0
Pressure drop (kPa) 46.7 38.3 41.13 40.2 38.4
Air core diameter (mm) 10.0 0.2 10.6 11.5 10.45
 INNOVATIVE HYDROCYCLONE INLET DESIGNS 209

important condition affecting the erosion rate of the hydro- the wear phenomenon of the vortex finder is more distinct
cyclone. Therefore, we will present four hydrocyclones with with increasingly finer ores escape from it. It can be seen
different feed inlets to simulate and calculate the erosion from Fig. 8 that there is hardly any wear in the inlet duct-
rate. Although the standard inlet as shown in Fig. 7a is ing. The erosion rate variation in the first 0.2 s is very sharp
common in mineral processing, it does not perform so well but the development later on is not obvious.
in wear resistance despite its popularity. Therefore, three Figure 9 shows the erosion rate of the inner wall of the
hydrocyclones with different inlets were tested for the ero- simulated hydrocyclones fitted with different inlets. Table 3
sion rate they are subject to with all parameters being fixed. lists the maximum and average erosion rates and computed
Figure 7 shows the detailed inlet geometry and dimension pressure drop for each case. Although the standard hydro-
for the four cases. In order to compare the effect of the cyclone with tangential inlet (Fig. 9a) has been widely used
inlet geometry on the erosion rate, the same fluid and in mineral processes, the erosion rate is the highest com-
particle velocity of 2.25 m=s is adopted for each case, the pared with the other three designs. Also, an obvious wear
flow rate of solid particles is set as 0.05 kg=s, and particle hot spot can be found at the bottom of the cone section,
diameter is 11.5 mm. In calculation of the erosion rate of where the erosion rate is very high. The maximum and inte-
the hydrocyclone, the interactions of the solid particles gral erosion rates are 3.72E-4 and 1.87E-6 kg=(m2 s),
Downloaded by [University of Saskatchewan Library] at 11:14 08 January 2015

and the continuous phase need to be taken into account. respectively. However, the pressure drop is the lowest,
Figure 8 displays the variations of erosion rate of a stan- 32.8 kPa. For the modified tangential inlet (Fig. 9b), there
dard hydrocyclone (Fig. 7a) with time from 0.1 to 0.9 s. At is no obvious wear hot spot, but the erosion rate is still high
the beginning, the erosion rate on the wall of the hydrocy- compared with the involute inlet. The maximum and aver-
clone is not very obvious because the solid particles have age erosion rates are 7.61E-7 and 4.72E-8 kg=(m2 s),
not yet reached the inner body of the device (Fig. 8a). With respectively, and the pressure drop is very high (81.7 kPa).
the particles injected into the device, more and more parti- For the involute inlet, which can provide a smooth
cles collide the wall inducing distinct wear (Fig. 8b). Two transition from pressure energy to rotational momentum,
wear ‘‘hot spots’’ can be found at the intersection of the the distribution of erosion rate is relatively uniform and
cylindrical and conical sections and the middle part of the value is low. For the circular involute inlet, the
conical section. At the hot spot in the middle part of coni- maximum computed erosion rate is only 4.32E-7 kg=(m2 s)
cal section, the erosion rate is more than 1E-4 kg=(m2 s), and the average value is 2.91E-8 kg=(m2 s), whereas for the
which is a very high value. Thus, this hot spot will cause elliptical involute inlet, the maximum and integral erosion
device failure. After 0.1 s, wear occurs at the vortex finder rates are 4.37E-7 and 3.90E-8 kg=(m2 s), respectively.
and spigot as well, and the latter is more obvious than the Moreover, the pressure drop of circular involute inlet
former. Note that the erosion rates at the bottom of conical (45.7 kPa) is much smaller than that of elliptical involute
and spigot are very high. From 0.2 s, the bottom of the inlet (72.3 kPa). It can be seen from Fig. 9 that the erosion
vortex finder begins to appear wear zone, and from 0.3 s rate at the inlet is nearly zero, whereas the erosion rate for
conical section and spigot is much higher that that of
cylindrical section and vortex finder.

CONCLUDING REMARKS
Four turbulence models, RNG k
e, QPS RSM, SLM
LES, and RNG LES, were used to predict the aerodynamic
performance of a 75 mm standard hydrocyclone. The com-
parison of numerical and experimental results indicates that
the RNG k
e turbulence model is not suitable for modeling
highly swirling flows in hydrocyclones, whereas QPS RSM,
SML LES, and RNG LES models can capture well the velo-
city profiles and predict the formation of air core. With a
VOF multiphase model, the air core formation was analyzed
in detail and the diameter of a steady air core was success-
fully predicted. The effects of inlet on the erosion rate were
investigated with the RNG LES model. The involute inlet
can eliminate the wear hot spot and lower the level of con-
centrated wear. This is only a preliminary study of the design
FIG. 7. Four hydrocyclones with different inlet duct designs: (a) stan- and optimization process concerning erosion rate of a hydro-
dard tangential inlet, (b) modified tangential inlet, (c) circular involute cyclone. In our future study, other parameters and con-
inlet, and (d) elliptical involute inlet. ditions such as inlet flow rate, particle characteristics, etc.,
 210 XU ET AL.
Downloaded by [University of Saskatchewan Library] at 11:14 08 January 2015

FIG. 8. Surface plots of the computed local erosion rates in the modeled hydrocyclone with a standard tangential inlet at various times: (a) 0.1 s,
(b) 0.2 s, (c) 0.3 s, (d) 0.6 s, and (e) 0.9 s.

TABLE 3
Computed erosion rate for four inlet duct designs
Pressure Maximum Face average
drop erosion rate erosion rate
Inlet (kPa) (kg=(m2 s)) (kg=(m2 s))
Standard tangential 32.8 3.72E-4 1.84E-6
inlet
Modified tangential 81.7 7.62E-7 4.72E-8
inlet
Circular involute 45.7 4.32E-7 2.91E-8
inlet
Elliptical involute 72.3 4.37E-7 3.90E-8
inlet

that can affect erosion rate will be investigated as all of the

performance parameters should be taken into account for
good design and operation of the hydrocyclone and to
increase its service life.

ACKNOWLEDGMENTS
This work was jointly supported by the National Natu-
ral Science Foundation of China through grant number
10572052 and the Graduate Science & Technology Inno-
vation Foundation of Huazhong University of Science
and Technology Through project number HF-05-15-2007-
012, as well as the Foundation for Study Abroad of Edu-
cation of Ministry of China. Partial support of M3TC at
NUS is also acknowledged.

REFERENCES
FIG. 9. Computed local erosion rates of the inner wall of tested hydro- 1. Bradley, D. The Hydrocyclone; Pergamon Press: London, 1965.
cyclone fitted with different inlets: (a) standard tangential inlet, (b) modi- 2. Svarovsky, L. Hydrocyclones; Holt, Rinehart and Winston: London, 1984.
fied tangential inlet, (c) circular involute inlet, and (d) elliptical involute 3. Lilge, E.O. Hydrocyclone fundamentals. Transactions of the
inlet. Institution of Mining and Metallurgy 1962, 71, 285–337.
 INNOVATIVE HYDROCYCLONE INLET DESIGNS 211

4. Dabir, B.; Petty, C.A. Measurement of mean velocity profiles in a 24. Rajamani, R.K.; Devulapalli, B. Hydrodynamics modeling of swirling
hydrocyclone using laser Doppler anemometry. Chemical Engineering flow and particle classification in large-scale hydrocyclones. KONA
Communications 1986, 48, 377–388. Powder and Particle Journal 1994, 12, 95–103.
5. Hsieh, K.T.; Rajamani, R.K. Mathematical-model of the hydrocyclone 25. Delgadillo, J.A.; Rajamani, R.K. Exploration of hydrocyclone designs
based on physics of fluid-flow. AIChE Journal 1991, 37, 735–746. using computational fluid dynamics. International Journal of Mineral
6. Dai, G.Q.; Chen, W.M.; Li, J.M.; Chu, L.Y. Experimental study of Processing 2007, 84, 252–261.
solid-liquid two-phase flow in a hydrocyclone. Chemical Engineering 26. Slack, M.D.; Del Porte, S.; Engelman, M.S. Designing automated
Journal 1999, 74, 211–216. computational fluid dynamics modeling tools for hydrocyclone
7. Boysan, F.; Ayers, W.H.; Swithenbank, J. Fundamental mathematical- design. Minerals Engineering 2003, 17, 705–711.
modelling approach to cyclone design. Chemical Engineering Research 27. Huang, L.X.; Kumar, K.; Mujumdar, A.S. A parametric study of the
and Design 1982, 60, 222–230. gas flow patterns and drying performance of co-current spray dryer:
8. Ma, L.; Ingham, D.B.; Wen, X. Numerical modelling of the fluid and Results of a computational fluid dynamics study. Drying Technology
particle penetration through small sampling cyclones. Journal of 2003, 21, 957–978.
Aerosol Science 2000, 31, 1097–1119. 28. Huang, L.X.; Kumar, K.; Mujumdar, A.S. Simulation of a spray
9. Fraser, S.M.; Abdel-Razek, A.M.; Abdullah, M.Z. Computational dryer fitted with a rotary disk atomizer using a three-dimensional
and experimental investigations in a cyclone dust separator. Journal computational fluid dynamic model. Drying Technology 2004, 22,
of Process Mechanical Engineering 1997, 211, 247–257. 1489–1515.
Downloaded by [University of Saskatchewan Library] at 11:14 08 January 2015

10. He, P.; Salcudean, M.; Gartshore, I.S. A numerical simulation of 29. Huang, L.X.; Mujumdar, A.S. Numerical study of two-stage horizon-
hydrocyclones. Chemical Engineering Research and Design 1999, tal spray dryers using computational fluid dynamics. Drying Tech-
77, 429–441. nology 2006, 24, 727–733.
11. Schuetz, S.; Mayer, G.; Bierdel, M.; Piesche, M. Investigations on the 30. Bozzini, B.; Ricotti, M.E.; Boniardi, M.; Mele, C. Evaluation of
flow and separation behaviour of hydrocyclones using computational erosion-corrosion in multiphase flow via CFD and experimental
fluid dynamics. International Journal of Mineral Processing 2004, 73, analysis. Wear 2003, 255, 237–245.
229–237. 31. Mazur, Z.; Campos-Amezcua, R.; Urquiza-Beltr
an, G.; Garcı́a-
12. Narasimha, M.; Sripriya, R.; Banerjee, P.K. CFD modelling of Gutiérrez, A. Numerical 3D simulation of the erosion due to solid par-
hydrocyclone—Prediction of cut-size. International Journal of ticle impact in the main stop valve of a stream turbine. Applied
Mineral Processing 2005, 71, 53–68. Thermal Engineering 2004, 24, 1877–1891.
13. Suasnabar, D.J. Dense Medium Cyclone Performance, Enhancements 32. Chen, X.; McLaury, B.S.; Shirazi, S.A. Application and experimental
via Computational Modeling of the Physical Process; Ph.D. Thesis, validation of a computational fluid dynamics (CFD)-based erosion
University of New South Wales, 2000. prediction model in elbows and plugged tees. Computers & Fluids
14. Cullivan, J.C.; Williams, R.A.; Cross, C.R. Understanding the 2004, 33, 1251–1272.
hydrocyclone separator through computational fluid dynamics. 33. Amezcua, A.C.; Mu~ noz, A.G.; Romero, C.A.; Czerwiec, Z.M.;
Chemical Engineering Research and Design 2003, 81, 455–465. Amezcua, R.C. Numerical investigation of the solid particle erosion
15. Cullivan, J.C.; Williams, R.A.; Dyakowski, T.; Cross, C.R. New rate in a steam turbine nozzle. Applied Thermal Engineering 2007, 27,
understanding of a hydrocyclone flow field and separation mechanism 2394–2403.
from computational fluid dynamics. Minerals Engineering 2004, 17, 34. Sciubba, E.; Zeoli, N. A study of sootblower erosion in waste-inciner-
651–660. ating heat boilers. Journal of Energy Resources Technology 2007, 129,
16. Brennan, M. CFD simulations of hydrocyclones with an air core: 50–53.
Comparison between large eddy simulations and a second moment 35. Olson, T.J.; Van Ommen, R. Optimizing hydrocyclone design using
closure. Chemical Engineering Research and Design 2006, 84, 495–505. advanced CFD model. Minerals Engineering 2004, 17, 713–720.
17. Narasimha, M.; Brennan, M.; Holtham, P.N. Large eddy simulation 36. Hsieh, K.T. Phenomenological Model of the Hydrocyclone; Ph.D.
of hydrocyclone—Prediction of air-core diameter and shape. Thesis, University of Utah, 1988.
International Journal of Mineral Processing 2006, 80, 1–14. 37. Delgadillo, J.A. Modeling of 75- and 250 mm Hydrocyclones and
18. Wang, B.; Chu, K.W.; Yu, A.B. Numerical study of particle—Fluid Exploration of Novel Designs Using Computational Fluid Dynamics;
flow in hydrocyclone. Industrial & Engineering Chemistry Research Ph.D. Thesis, University of Utah, Salt Lake City, Utah, 2006.
2007, 46, 4695–4705. 38. Fluent V6.3, User’s Guide; Fluent Inc.: Lebanon, NH, 2006.
19. Hsu, C.-Y.; Wu, R.-M. Hot zone in a hydrocyclone for particles 39. Finnie, I. Erosion of surfaces by solid particles. Wear 1960, 3, 87–103.
escape from overflow. Drying Technology 2008, 26, 1011–1017. 40. Edwards, J.K.; McLaury, B.S.; Shirazi, S.A. Evaluation of alternative
20. Souza, F.J.; Neto, A.S. Preliminary results of large eddy simulation of pipe bend fittings in erosive service. Proceedings of ASME
a hydrocyclone. Thermal Engineering 2004, 3, 168–173. FEDSM’00: ASME 2000 Fluids Engineering Division Summer Meet-
21. Delgadillo, J.A.; Rajamani, R.K. A comparative study of three ing, Boston, FEDSM 2000–11251, June 11–15, 2000.
turbulence-closure models for the hydrocyclone problem. Inter- 41. Edwards, J.K.; McLaury, B.S.; Shirazi, S.A. Modeling solid particle
national Journal of Mineral Processing 2005, 77, 217–230. erosion in elbows and plugged tees. Journal of Energy Resources
22. Nowakowski, A.F.; Cullivan, J.C.; Williams, R.A.; Dyakowski, T. Technology 2001, 123, 277–284.
Application of CFD to modeling of the flow in hydrocyclones. Is this 42. Hwang, K.-J.; Hsueh, W.-S.; Nagase, Y. Mechanism of particle separ-
a realizable option or still a research challenge? Minerals Engineering ation in small hydrocyclone. Drying Technology 2008, 26, 1002–1010.
2004, 17, 661–669. 43. Chu, L.-Y.; Qin, J.-J.; Chen, W.-M.; Lee, X.-Z. Energy consumption
23. Monredon, T.C.; Hsieh, K.T.; Rajamani, R.K. Fluid-flow model and its reduction in the hydocyclone separation process. III. Effect of
of the hydrocyclone—An investigation of device dimensions. Inter- the structure of flow field on energy consumption and energy saving
national Journal of Mineral Processing 1992, 35, 65–83. principles. Separation Science and Technology 2000, 35, 2679–2705.