Study On Sand Particles Drying in A Fluidized Bed Dryer Using CFD
Study On Sand Particles Drying in A Fluidized Bed Dryer Using CFD
Abstract
INTRODUCTION
Fluidized bed dryers are common for drying of wet sand particles, pellets, powders
and other chemicals in various industries such as mineral separation plant, petroleum,
agricultural and pharmaceutical industries. The process involved in drying using
130 Jibin Antony and M.B. Shyamkumar
fluidized bed dryer is convection where the heated air in direct contact with the wet
feed, dries the material suspended in a fluidizing bed. In the drying process, the air
from a blower is heated by a natural gas fueled air heater and passes upward into the
dryer through a perforated air distributor plate. This fluidizes the wet feed material.
As the heated air passes through the wet material, the air carries away the moisture
content and is exhausted through an outlet. Fig.1 shows the drying process of sand
particles. The advantages of fluidized bed dryers over other drying methods are (1)
high drying rate because of good interaction between the particles and air, (2) better
temperature control and operation throughout the process and (3) less maintenance
cost due to absence of moving components inside the dryer.
flow patterns that are difficult, expensive or impossible to study using experimental
techniques.
Although the tools for applying single-phase CFD are widely available, application of
multiphase CFD is however, still complicated from both the physical and numerical
point of view. Furthermore, multiphase CFD simulations are time consuming and
therefore average flow predictions in large scale equipment is not obtained readily.
Currently there are two approaches for multiphase modelling; the Euler-Lagrange
approach and Euler-Euler approach. For gas-solid flow modelling usually Eulerian-
Eulerian models also called Granular Flow Models (GFM) are employed. They are
more suitable for simulating larger and complex industrial fluidized-bed dryers
containing billions of solid particles.
Research work on fluidized bed dryers are improving now a days, because of its
applications in large scale industries and its advantages over other drying processes.
Chejne and Hernandez,[1] developed a one dimensional steady state mathematical
model to simulate the coal gasification process in fluidized bed dryer. The model
developed, incorporates two phases namely solid and liquid. The model predicts the
temperature and particle size distribution for solid phase and that for the gaseous
phase. Izadifar and Mowla,[2] developed a mathematical model to simulate the drying
of moist paddy in a fluidized bed dryer. The model proposed was capable of studying
the variation of moisture content of paddy through the dryer. The predicted model
showed good agreement with the experimental results.
Gareth et. al,[3] presented a paper on the application of chaos analysis to pressure
fluctuations in a conical fluidized bed of dry and wet pharmaceutical granules, as it
progresses to the dry state. Changes in bed hydrodynamics arising from particle size
distribution have been found to be most easily resolvable at low gas velocities, mostly
less than 2 m/s.
Ozbey and Soymelez,[4] experimentally investigated the batch drying of wheat grains
in a fluidized bed dryer having a swirling flow field in its drying medium. The effect
of the air temperature and mass flow rate on the drying process was also studied.
Liang et. al,[5] developed a new numerical model based on two fluid model including
the kinetic theory of granular flow, to simulate the coal gasification in a bubbling
fluidized bed gasifier. They predicted the flow behaviour of gas and solid phases in
the bed and freeboard, which are difficult to measure through the experiments. The
relationship between gas composition profiles with the height of gasifier and the
contours of temperature, velocity and solid volume fraction were discussed for both
phases. Assari et. al,[6] introduced a mathematical model for batch drying based on
Eulerian two fluid models. The effects of parameters such as inlet gas temperature and
velocity on the moisture content and the temperature of solid and gas at the outlet are
discussed.
132 Jibin Antony and M.B. Shyamkumar
Rozainee et. al,[7] utilized CFD modelling to determine the trajectories and residence
time of burning rice husk particles in the fluidized bed combustor at different
secondary air flow rates. From their study, the secondary air flow rate which is 80%
of the primary air flow rate, gave the highest average bed temperature and lowest
residual carbon content in the ash. Mahmood et. al,[8] studied the hydrodynamics of a
vertically vibrating fluidized bed using an Eulerian-Eulerian two fluid model (TFM)
incorporating the kinetic theory of granular flow and including the frictional stress
effects. They studied the influence of frictional stresses, vibration amplitudes and
frequency, on behavior of the particles. The ability of the two-fluid model for
predicting the hydrodynamics of vibrating fluidized beds was also discussed.
Jungkee and Hamid,[9] carried out the CFD simulation of a pharmaceutical bubbling
bed drying process at three different scales. They studied the gas–solid flow patterns,
mixing and drying of pharmaceutical particles for three different scales of bubbling
fluidized bed dryers. Their results were compared with the experimental data obtained
at the laboratory scale to validate.
Yuping et.al,[10] done their simulations for biomass drying using a novel fluidized
bed dryer. Biomass drying is performed mainly in rotary dryers, which occupy a large
floor area. To explore the efficient drying of biomass, a fluidized bed dryer was
proposed. A high drying rate could be achieved because of the fast mass and heat
transfer rate in the fluidized bed dryer.
Askarishahi et.al,[11] in their paper done CFD study on solids flow pattern and solids
mixing characteristics in bubbling fluidized bed to find the effect of fluidization
velocity and bed aspect ratio on drying process. In their work, using the two-fluid
model based on the kinetic theory of granular flow, a set of governing equations was
solved by applying finite volume method in 2-D to save the time for simulations.
Adam et.al,[12] developed a CFD model of coal gasification in a circulating fluidized
bed reactor. Eulerian–Lagrangian approach was used to simulate the fluidized bed
hydrodynamics. Gasification with air and air/steam mixture was considered for their
study. Results of the simulations coincide well with the measured syngas
composition.
In the present work, CFD simulations are carried out to study the sand particles drying
in a fluidized bed dryer. Here the wet sand particles are fed to the dryer through a
conveyor. The hot air from the bottom of the dryer passes through the nozzles
embedded in the dryer bed and rises up. This upwind do not allow the wet sand
particles to settle down. Instead, the sand particles are in a fluidized state inside the
dryer. During the fluidized state movement of sand particles from inlet to outlet, the
particle drying will occur. This process is simulated with the aid of ANSYS FLUENT
14.5 software.
Study on Sand Particles Drying in a Fluidized Bed Dryer using CFD 133
MATHEMATICAL MODELLING
Governing Equations
The governing equations for Eulerian multiphase model include conservation of mass,
momentum and energy equations. Equations of solid and gas phases are developed
based on Eulerian-Eulerian model, using the averaging approach. The system of
governing equations is summarized below:
Conservation of momentum:
Gas phase:
Solid phase:
where ρ is the density, ε is the volume fraction, v is the velocity vector, p is the
pressure, is the shear stress tensor, g is the gravitational acceleration vector, h is the
134 Jibin Antony and M.B. Shyamkumar
heat transfer coefficient, q is the heat flux, Q is the heat transfer rate, H is the latent
heat and K is the drag coefficient.
FLUENT needs the domain in which the flow takes place to evaluate the solution.
The flow domain as well as the grid generation throughout the domain have been
created in ICEM CFD software. The three dimensional geometry created using
SOLIDWORKS is imported into ICEM CFD for grid generation. Tetrahedral meshing
is used for the current geometry (Fig.3). Three different mesh sizes are generated to
study the grid independency of the meshed model. After meshing the necessary initial
and boundary conditions are to be imposed in FLUENT 14.5 for starting the
simulation.
Study on Sand Particles Drying in a Fluidized Bed Dryer using CFD 135
NUMERICAL METHEDOLOGY
Boundary and initial conditions
In order to obtain a result with minimal error, appropriate boundary conditions for the
computational domain have to be implemented. Here the inlet boundary condition is
imposed with velocity inlet and the outlet boundary condition is the pressure outlet,
which is set as default pressure. Wall boundary conditions are no-slip condition for
both solid and gas phase. The higher viscous effect and higher velocity gradient near
the wall have been dealt with the standard wall function. For patching sand volume
fraction, the sand in the part of the column up to which the sand particles were
initially fed has been used. Initially a sand volume fraction of 0.8 of the static bed
height of column has been used and the volume fraction of air at the inlet is set as
default value. Table.1 shows the boundary and initial conditions for the CFD analysis
of drying sand particles.
Table.2 shows the numerical schemes and turbulence model employed in CFD.
1 46,18,423 85 92 8.23
2 65,35,919 85 90 5.88
3 87,96,534 85 89 4.7
Here the outlet air velocity is considered as the parameter for grid independence
study. Figures 4-6 demonstrates the convergence history of outlet air velocity for
three different grid sizes. The actual experimental result at KMML for outlet air
velocity is 85 m/s. For a grid size of 4.6 million, the simulation result obtained was 92
m/s. For 6.5 million and 8.7 million mesh, the outlet air velocity obtained are 90 m/s
and 89 m/s respectively. This implies that 6.5 million mesh and 8.7 million mesh
results provide less error percentage than 4.6 million mesh. As the error percentage
between 6.5 million mesh and 8.7 million mesh is minimum and considering the
computational time, 6.5 million mesh is considered for all simulations.
Study on Sand Particles Drying in a Fluidized Bed Dryer using CFD 137
Figure.4: Convergence history of Outlet air velocity for 4.6 million mesh
Figure.5: Convergence history of Outlet air velocity for 6.5 million mesh
138 Jibin Antony and M.B. Shyamkumar
Figure.6: Convergence history of Outlet air velocity for 8.7 million mesh
Validation study
Table.4 shows the validation results. The outlet air temperature in the fluidized bed
dryer is monitored experimentally with the PLC circuits and is constant at 353K.
Through numerical simulation it is found to be 359K, with a percentage error of
1.69%. Similarly, the percentage error in outlet air velocity is found to be 5.88 %.
The fluidization process is nicely captured through this simulation. Initially the
highest concentration of sand is at the bottom of the dryer. With the increase in
velocity of the upward hot air, the sand particles will start to rise up in the dryer and
the sand particles will be in a fluidized state thereafter.
142 Jibin Antony and M.B. Shyamkumar
Figures 13 and 14 shows the contours of air and sand temperature distribution inside
the dryer. The temperature of air is maximum at the nozzle inlet. Through the nozzles
the hot air rises up. This upwind suspends the wet sand particles in a fluidized state.
The hot air surrounding the wet sand causes the sand to dry. Thus the temperature of
air reduces as it flows out through the outlet of the dryer.
The temperature of sand particles over time inside the dryer is monitored at different
inlet air temperatures as shown in Table.5. It is observed from Fig.15 that with
increase in air temperature from 400 K to 440 K, there is increase in sand
temperature. In KMML, once the sand temperature is reached 373 K, it is taken out. It
is observed that with increase in inlet air temperature, the sand particles obtain 373K
much faster showing an increased drying rate of sand. At inlet air temperature, 430 K
the change is less compared with other inlet temperatures. So for optimization of
dryer the inlet air temperature can be selected as 430 K. Sand temperature at different
inlet air temperature is plotted below.
Figure.15: Variation of sand temperature with time for different inlet air temperature
From figure 15, it is observed that as the temperature of the inlet air is increased, the
sand particle temperature inside the dryer is also increased. For the inlet temperature
144 Jibin Antony and M.B. Shyamkumar
of 400K, the sand particles reach 373 K around 2.5 s inside the dryer. With increase in
temperature to 410 K, it is reduced to approximately 2.3 s. With increase in inlet air
temperature, the sand particles attain 373 K inside the dryer with less time . i.e., the
drying rate is increased with increase in temperature. But from 430 K to 440 K the
change is less compared with other inlet temperatures. So for the current study, the
optimum temperature for inlet air is observed at 430 K.
CONCLUSIONS
CFD study of sand particles drying in a fluidized bed dryer has been carried out using
ANSYS FLUENT 14.5. Eulerian multiphase model has been used to simulate the
flow of particles inside the dryer. The simulation results are aggregable with the
experimental results obtained, for outlet air temperature and outlet air velocity. With
the increase in inlet air temperature, the sand particle temperature inside the dryer is
also increased. As inlet air temperature is increased, sand particles drying rate is also
increased.
REFERENCES