Advances in Systems Science and Applications (2011), Vol.11, No.

3-4 315-321

Simulation of the Flow Field of Cement Mixer Based on

Numerical Methods
Li Shuiping, Li Xiaotian, Shi Lugang
School of Mechanical and Electronical Engineering, Wuhan University of Technology, Wuhan 430070,
P.R.China

Abstract Different kinds of research methods of the flow field in the cement mixer were introduced and
compared, theory bases of the numerical simulation method of the flow field in the cement mixer were described,
whole of thoughts of realization this method were analyzed. And then, the geometrical model and mathematical
mode of the cement mixer were created by Fluent, which was based on CFD theory; and the characteristic of
flow field in the cement mixer was simulated by Mixture model, SIMPLE algorithm and standard k-ε turbulence
model. The calculation results showed that the distribution of flow field in cement mixer could be simulated
preferably by the numerical technology, which supplied the valuable theoretical guidance for designing of the
cement mixer.
Keywords Cement mixer Numerical simulation Flow field

1.Introduction

Cement slurry is frequently needed to grout the cracks of the buildings when reinforcing the
constructions. Currently, most cement slurry is produced by the common mechanical mixer
which makes grout by vane, scraper or spiral device. And the mixer is composed by gear, mixing
shaft and impeller. It will easily produce the dead angle and precipitation at the bottom of mixing
tank in the mixing process due to the blade is fixed on the shaft. Apart from that, it has the
problems which are big energy waste, low efficiency and can not prepare for different
water-cement ratio cement slurry fast. Therefore, the development of a new type of
high-performance cement slurry mixing equipment has become rather important.
Currently,cement mixer commonly can be divided into three categories:colloidal mills
mixer, blades mixer, water-jet mixer. Although these devices are generally able to achieve the
desired technical effect, the mixing efficiency and cost is another matter. colloidal mills mixer is
an advanced equipment, mixing, high efficiency, high concentration slurry can be stirred, but the
structure is more complex and costly. the structure of blades mixer is simple, the cost is
inexpensive, and widely used in engineering, but the efficiency is low. water-jet mixer is a
low-cost equipment, have the characteristics of high speed stirring, but because the mixing
uniformity is not high, to some extent these limit its wider use. So how to provide advanced
cement mixing equipment is an important issue that engineering and technical personnel in the
country have to faced.
For the design of cement mixer, traditional theory and experimental measurements usually
used in analysis of the flow field. Study found that in the mixer there are various of slurry flow
forms at the same time, combined with the interaction of cement powder and water, make the
two-phase flow field very complex and difficult to carry out a detailed theoretical analysis. Also
the complex flow characteristics and opacity of the slurry limit the application of experimental
methods. In process of today's research about fluid mechanical design, the CFD method to
simulate the flow fluid inside the machine has become an important technical means when the
software of CFD have emerged. In engineering, numerical simulation has been more widely
ISSN 1078-6236 International Institute for General Systems Studies, Inc
316 Li: Simulation of the Flow Field of Cement Mixer Based on Numerical Methods

used, it can be replaced experiment to cut down the cost in some extent, In addition, numerical
simulation can provide a lot of information about flow field, it can give reliable basis for the
optimal design of fluid machinery for designers.
Comprehending of the cement slurry flow characteristics and the mechanism of mixing is a
basic prerequisite for the structural design [1]. The main purpose of this paper is to simulate and
analyze the flow field of the equipment based on the numerical simulation, and then study the
mixed performance.

2.Research Methods of the Mixer Flow

2.1 Introduction of Research Methods of the Flow Field

2.1.1 Theoretical analysis method
On the basis of the relevant experimental data and the law of conservation of energy and so
on, it will establish the velocity field distribution of fluid flow, the stress field distribution of
fluid flow , mass conservation law and energy conversion law by the mathematical analysis, thus
construct the mathematical model of fluid flow[2] .As a result, it can forecast the intensity and
characteristics of the fluid flow of mixer as well as its function and influence on mass transfer
and energy exchange.
2.1.2 Experimental Methods
(1) Visualization of fluid method
Visualization of fluid method which can monitor the trajectory of fluid is a basic method of
studying flow field. There are anisotropic thin (particulate) method, laser-induced fluorescence
method and dye tracer method [3].
Anisotropic thin (particulate) act refers to adding the anisotropic particles (such as titanium
dioxide, mica, aluminum stearate, etc.) which have the reflective capacity to the fluid so that the
anisotropic particles flow with the fluid, and the strong reflective capacity of particulate makes
the flow trajectory of fluid can be visible. In addition, we are able to study the movement law of
two-phase fluid and the migration law of material according to the selective dispersion of some
particles in fluid (for example, water phase is chosen for mica, while hydrophobic organic phase
are chosen for aluminum and mica).
(2) Video measurement
Using advanced camera equipments (such as magnetic resonance fluid imager, Laser
Doppler veloeimetry, particle image vefocimetry etc.) to film the track of fluid flow and then
obtain it’s boundary conditions as well as the related parameters [3]. PIV technique is non-contact
measurement technology which has many advantages such as high measurement accuracy, wide
range of speed measurement speed, having simple principle and small influence from
outside .Consequently, it is able to depict the whole velocity vector of the total velocity field, so
the flow field information of one profile can be analyzed at once.
2.1.3 Numerical Simulation Method
Numerical simulation is that adopting numerical method and logical way to reflect changes
prototype process and the rules of movement in the computer. Mathematical model, different
from the physical model, is often made from a single equation, a group of equation or groups
formulas. We can get the various locations of the basic physical quantities (such as speed,
pressure, temperature, concentration, etc.) and their changes in the very complex flow by
numerical simulation. In addition, it can optimize the design of structural by combining with
CAD.
2.2Comparison of the Research Methods of the Mixer’s Internal Flow Field
Advances in Systems Science and Applications (2011), Vol.11, No.3-4 317

Table 1 Comparison of the Research Methods

Fheoretical Nnalysis
Experimental Method Numerical Simulation
Method
The results whose all It is easy to choose
kinds of influencing The experimental results different physical
factors are clearly visible which are true and parameters to make
Main Advantage

have universality. It is believable,are the basis of the various effective

the theoretical basis theoretical analysis and and sensitive tests
which is used to guide the numerical method. because you are not
experimental research and restricted by physical
verify the new numerical model and
calculation method. mathematical model.
Experiments are often
Main Disadvantages

We are often required to

limited by model size, flow
It strongly rely on the
abstract and simplify the
disturbances and measuring
mathematical
object of calculation. For
accuracy, in addition ,they
model,and numerical
the non-linear situation,
will also encountered lots of
processing method
only the analytic results of
problem such as the
will lead to false
simple flows can be
enormous cost of funding
results.
given. and manpower,material
resources and the long cycle.
Numerical simulation method combined with traditional theoretical analysis method and
experimental measurement method form a perfect system of studying the problems of the mixing
equipment’s internal fluid flow.
From table 1, we can we see clearly that these methods have their own advantages and
disadvantages. Both traditional theoretical analysis method and experimental test method can not
obtain the characteristics of flow field mainly due to the following two points: first, the
visualization methods such as staining PIV and tracer method can not be used to measure the
flow field which has high concentration and poor visibility; second, at present, all of the contact
velocimeter can not test the velocity vector of cementing slurry flow field accurately. Therefore,
the slurry flow field can only be studied by numerical simulation.

3. Numerical Simulation Method of the Internal Flow Field of Mixer

3.1 Theoretical Basis of Numerical Simulation

The theoretical basis of numerical simulation is computational fluid dynamics. Fluid flow is
dominated by physical conservation laws which relates to the principal basic laws including
mass conservation law, momentum conservation law, energy conversion and conservation law,
components conversion and balance law，and so on[2,5]. These laws of the basic equations can be
expressed as follows:
∂ ( ρφ ) K (1)
+ div( ρ uφ ) = div(Γφ gradφ ) + S
∂t
Expand the form of:
∂ ( ρφ ) ∂ ( ρ uφ ) ∂ ( ρ vφ ) ∂ ( ρ wφ )
+ + +
∂t ∂x ∂y ∂z
∂ ∂φ ∂ ∂φ ∂ ∂φ (2)
= (Γ )+ (Γ )+ (Γ )+S
∂x ∂x ∂y ∂y ∂z ∂z
Where φ is the general dependent variable, and so can represent the solution of variables
of u、v、w、T ; Γ is the generalized diffusion coefficient; S is the generalized source term.
318 Li: Simulation of the Flow Field of Cement Mixer Based on Numerical Methods

In the macro category of macro continuous mechanics, two-phase flow analysis can be
divided into two types: (1) local homogeneous flow analysis; (2) two-phase slip flow analysis.
Cementing slurry flow in stirred tank belongs to two-phase slip flow which mainly has three
types of analysis model: Euler-Lagrange model, Eulerian-Eulerian model that is a mixture of the
two-fluid model and Mixture model. Euler-Lagrange model centralizes the advantages of
macroscopic pseudo-fluid model and micro-dynamics model, that is, take into account not only
the solid phase and liquid-phase interactions, and can describe the collision between particles,
but it needs to record the spatial location and movement of every particle in the flow field ,
therefore a larger computer memory is necessary. Currently, there are certain difficulties in the
simulation of high concentration of solid-liquid two-phase flow [4, 5]. At present, multi-fluid
model is given a wide range of applications. In a multi-fluid model, the continuous (liquid)
phase movements is described by Euler equations, assuming that the dispersed (particle) phase is
the mutual penetration and continuous quasi-fluid, its movement is also described by Euler
equation, and then solve the momentum, energy and quality equations of each phase. However,
the computational complexity of multi-fluid model results in poor stability, so it confronts many
difficulties in the practical application.
Mixture model is a simplified multiphase flow model, which is used to simulate each phase
of the multiphase flow at different speeds, however, the local equilibrium under short spatial
scale is assumed. For mixture model, there is a strong coupling between different phases. When
there exist a wide range of particle distribution, Mixture model can achieve good results as
perfect multiphase flow model. Due to space limitations, the control equations of two-phase flow
will not be listed here.
3.2 The Process of Numerical Simulation of the Internal Flow Field
At present, the numerical simulation of mixing process is on the basis of macro flow field.
First of all, it obtains the velocity field distribution law of macro-flow field and then joins tracer
when the flow field is stable. Through the calculation of tracer concentration changes with space
and time to simulate the mixing process. This method applies to the flow field of relatively low
concentration. In spite of that, there are great difficulties in the simulation of high concentration
slurry flow field. In this paper, cementing slurry two-phase flow field was simulated by mixture
model, and then the mixing effects were assessed according to the results of simulation. The aim
of numerical simulation is to obtain the useful data used to instruct the optimal design of mixer,
such as mixing time and mixing power number. The basic flow of numerical simulation is shown
in Figure1.
In the process of numerical simulation, we can create the geometric model according to the
structure of mixer. Generally, we should deal with the physical mode reasonably during the
establishment of geometric model since because it is usually rather complex [6, 8]. In order to
reduce the unnecessary computation time, the physical model need to be simplified as much as
possible, such as the cover of the mixer which has little impact on the effect of stirring, therefore,
it ought to be neglected during the geometric model establishment; for those structures which
have greater impact on the effect of stirring, not only to retain but also to establish them
accurately. The establishment and dispersion of the mathematical model which directly affects
the accuracy and precision of simulation are the core parts of numerical analysis. The key of
numerical simulation is to establish the mathematical model which can reflect the essence of
engineering problems. For different fluids, we can choose different suitable mathematical
models, such as RNG k-ε two-equation turbulence model or two-Realizable k-ε model may be
the best choice in a flow field with a cyclone or strong bending wall [9]. We discretize the
regional of calculation before numerical calculation, that is, divide the space of calculation into
many sub-regions, and to identify nodes in each region in order to generate grid. And then, the
control equations are discretize in the discrete grids, as a result, partial differential equations turn
into the algebraic equations of each node for numerical solution. Algebraic equations have a
Advances in Systems Science and Applications (2011), Vol.11, No.3-4 319

variety of discrete formats including QUICK format which can provide higher calculation
accuracy when we study swirling flow by structured grid. So the QUICK format may be the best
choice during the numerical simulation of the internal flow field in mixer.
the physical
model of mixer

establishing
geometric model

analyze，
compare,optim
ize
mesh generation

establishing
mathematical
model

visual data
mining algorithm
optimization

numerical
simulation

judgment of no
convergence

yes

Figure 1 The basic flow of numerical simulation

Apart from a few simple questions, the discrete equations can not be directly solved,
because certain adjustments of the dispersion equations are necessary, and we ought to deal with
the order and way of unknown variables (speed, pressure, temperature, etc.) especially in order
to solve them easily. SIMPLE algorithm is widely used in engineering. Before carrying out the
computer simulation, setting the model boundary conditions is required because the boundary
conditions which include the inlet velocity, boundary conditions, outlet pressure, the
concentration of slurry, are necessary for solving the discretized equations. Setting appropriate
initial value and sub-relaxation factor appropriate before the iterative solution in favor of
iterative calculations converge faster and more easily. Once convergence of iterative solution
achieves prescriptive convergence precision, the data we need can be obtained through the user
interface.

4. Process of Numerical Simulation of Flow Field of Mixer by Fluent

4.1 Tools of Numerical Simulation of Flow Field
On the basis of the theory and method above, we can achieve the numerical simulation of
flow field by Gambit and Fluent soft wares [10]. Fluent is specific CFD software which is used to
simulate and analyze a complex geometry in the field of fluid flow and heat exchange while
GAMBIT is professional pre-processing software for Computational Fluid Dynamics, its main
features include three aspects: geometric modeling, mesh generation and assignment of
boundaries. Among them, mesh generation is the most important function. It generates grid file
containing boundary information. Fluent can realize numerical simulation of flow field of mixer
effectively in connection with its pre-processing software Gambit.
320 Li: Simulation of the Flow Field of Cement Mixer Based on Numerical Methods

4.2 The Establishment of Geometric Model

Mesh generation is the necessary condition to establish discrete control equations, because
the quality of the mesh has a direct impact on the accuracy of simulation results. Through
analyzing the physical model of mixer, this paper used Gambit to establish the mesh model on
the basis of geometric model of mixer, as shown in Figure 2.

Figure 2 The mesh model of mixer

To reduce quantity of mesh cells and to improve precision and convergence, we used
structured mesh in the nozzles. For the internal structure with complex geometric shape, the
flow field changes rapidly, it is inconvenient to adopt structured mesh, so a
structured-unstructured mixed type mesh was employed in this computation.
4.3 Solve the Discrete Mathematical Model
Mesh model file will be optimized after being imported in FLUENT. We discreted control
equations of mesh nodes by numerical discretization of QUICK format in this paper. Numerical
algorithm is one of the methods for solving the discrete control equations of grid model, and
SIMPLE algorithm is quite applicable.
4.4 Preliminary Analysis of Simulation Results
To a large extent, the effect of mixing depends on the internal structure of flow field, so the
existence of turbulence which has complex flow field is an important prerequisite for enhancing
the effect of mixing.

Figure 3 Velocity vectors colored by radial velocity (m/s)

After analyzing the velocity vector distribution chart, it can be seen that the mixer produced
a strong vortex under the action of the spiral blade combined the nozzles. The radial velocity is
Advances in Systems Science and Applications (2011), Vol.11, No.3-4 321

shown in figure 3.

5. Future Research

At present, the literatures about the numerical simulation of cement mixer are rare.
Although the numerical simulation method has its unique advantages, there are many aspects
need to be further improved, which mainly include the numerical simulation accuracy and
convergence rate.
Numerical simulation accuracy of multiphase flow mainly depends on the discrete
mathematical model, therefore how to establish mathematical model in accordance with the
mechanical structure of mixer is one of the difficulties of numerical simulation. Low–order
dispersion and high-order dispersion has its own excellent shortcomings, so the way to solve the
discrete control equations is a very key technology. The purpose of numerical simulation of flow
field is obtaining the useful data to optimize the mechanical structure, thus making further
analysis of the simulation results is one of the future research directions.

Acknowledgements

This research is funded by the National Key Technology R&D Program of China (No.
2006BAJ03A09-08).

References
[1] Zhao Jianjun，Yuan Shouqi，Liu Houlin, Huang Zhongfu and Tan Mingao. Simulation of
solid-liquid two-phase turbulent flow in double-channel pump based on Mixture model.
Journal of Transactions of the CSAE, 24 (1)( 2008) 7-10.
[2] Wang Fujun. Analysis of Computational Fluid Dynamics—Principle and Application of
CFD Software. Beijing: Tsinghua University Press , 2006.
[3] Zhan Hanhui, Chenghao, Liu Jianwen and Zhan Xuehui. Principle of Secondary Flow.
Changsha:Central South University Press, 2005.
[4] Anders Darelius, Anders Rasmuson, Berend vanWachem, Ingela Niklasson and Staffan
Folestad. CFD simulation of the high shear mixing process using kinetic theory of granular
flowand frictional stress models. Journal of Chemical Engineering Science, 63(8)( 2008)
2188-2197.
[5] Zhou Lixing. Dynamics of Multiphase Turbulent Reacting Fluid Flows, Beijing: National
Defense Industry Press, 2002.
[6] Wu Bo,Yan Hongzhi,Duan Yiqun. Study on 3-D Turbulence Numerical Simulation and
Wear Characteristics of Slurry Pump. Journal of China Mechanical Engineering.
20(6)( 2009) 719-780.
[7] Huang Si, Wang Guo-yu. A 3D Numerical Simulation of Solid-liquid Turbulent Flow with
High Solid-concentration in a Centrifugal Pump. Journal of Coal Mining Machinery.
Vol.11(2005) 53-57.
[8] K.H. Javed, T. Mahmud and J.M. Zhu. Numerical simulation of turbulent batch mixing in a
vessel agitated by a Rushton turbine. Journal of Chemical Engineering and Processing,
45(2)( 2006) 99–112.
[9] Benjamin Coesnon, Mourad Heniche, Christophe Devals, Franc¸ois Bertrand and Philippe
A. Tanguy. A fast and robust fictitious domain method for modeling viscous flows in
complex mixers: The example of Propellant make-down. International Journal for
Numerical Methods In Fluids, 58(4)( 2008) 427–449.
[10] Fluent Inc. FLUENT User Guide.2003.