1988 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
model is applied, a well-established model widely used in
various industrial applications in CFD and fluid dynamics
(Launder &Spalding, 1983).
Particles and Flow Model
In this study, CFD simulations are performed initially in
batch mode mono-size fluidised bed hydrodynamics fol-
lowed by continuous mode mono-size particles with the
effect of inclined channels. A fixed particle density of ρs =
2650 kg/m3 is used for both batch and continuous mode
simulations. The model parameters are discussed in the fol-
lowing sections.
Batch Model for Fluidised Bed
This investigation employs a batch mode fluidisation model
to validate the CFD model. The operating parameters and
material properties are derived exclusively from experi-
ments conducted for the specific purpose of validation. It is
important to highlight that identical laboratory equipment
and materials are utilized for both the fluidised bed experi-
ments and the subsequent continuous flow experiments.
For the batch mode fluidised bed simulations, four
distinct particle sizes – 700, 600, 450, and 350 µm – are
selected to investigate the influence of particle size on bed
hydrodynamics. The bed is initialized with a height of
0.142 m and a solid volume fraction of 0.56. Across oth-
erwise identical experimental conditions, eight different
fluidisation rates are explored for each particle size: 0.4,
0.8, 1.2, 1.4, 1.6, 1.8, 2.0, and 2.4 L/min. Rather than
conducting eight separate simulations, the fluidisation rate
is incrementally increased in 2-second intervals (refer to
Figure 5c). Data from the simulations are then extracted
and analysed for every two-second interval to assess valida-
tion parameters at different fluidisation rates.
Validation parameters, namely bed height and bed dif-
ferential pressure, are assessed at the point of steady-state
conditions and subsequently compared with correspond-
ing experimental data. The bed height is determined by
measuring the maximum height attained by particles under
steady-state conditions. Meanwhile, the bed differential
pressure is computed as the mass flow-averaged total pres-
sure difference between two boundary surfaces, symbol-
izing the high- and low-pressure boundaries within the
domain. These pressure boundaries, referred to as observa-
tion planes, are positioned at the same heights as pressure
transducers in the experiments, as depicted in Figure 3.
Continuous Flow RC Model
A continuous flow model is constructed to match the
experimental conditions for Run 6 of Starrett and Galvin
(2023) The particle sizes for the continuous flow simula-
tions are chosen from the spectrum of particles employed in
the experiment, ranging from 30 to 710 µm. The simula-
tions initiate with an incoming slurry fed through the feed
inlet, and fluidisation water enters the domain through the
lower inlet. Additionally, a side-water inflow is introduced
through the side-water inlet (refer to Figure 1b). Operating
conditions for the continuous flow model are adopted from
the experiment and detailed in Table 1.
For simplicity, the underflow is omitted in the model,
and instead, a plane below the side-water inlet is con-
structed and utilized to monitor the underflow as particles
reach and pass through it. In the subsequent phase of the
study, the underflow will be incorporated into the CFD
model, along with multi-size particles. The partition of par-
ticles to underflow and overflow is obtained by monitoring
the mass flow rates at the outlet and across the underflow
plane. Partition numbers for each particle size are then fit-
ted to a partition curve as shown in Eq. (1).
expdln^3h
P
E
Dh 100
1
1
p
50
#=
+
-^D n
(1)
where D represents the particle diameter and D50 is defined
as the separation size. The sharpness of the partition curve
is quantified by the Ecarte Probable (Ep) with smaller values
denoting a sharper separation.
Geometry and Boundary Conditions
The numerical simulations closely follow the experimen-
tal approach undertaken by Starrett and Galvin (2023).
A three-dimensional computational domain is established
mirroring the specifications of the equipment utilised in
the experiment. The computational domain includes
inclined channels to replicate the configuration of the actual
equipment. The schematic diagram of the RC and the 3D
CFD domain are illustrated in Figure 1. Boundary condi-
tions consist of a velocity inlet at the domain’s inlet and
a pressure outlet at the outlet. Additionally, a side-water
inlet and feed inlet are incorporated in the domain for the
Table 1. Operating conditions for continuous flow (Starrett
&Galvin, 2023)
Operating Parameters Units
Feed Rate 22.0 t/m2/h
Feed Pulp %Solids 51.2 (wt. %)
Side-Water Rate 3.0 (L/min)
Fluidisation Rate 1.0 (L/min)
model is applied, a well-established model widely used in
various industrial applications in CFD and fluid dynamics
(Launder &Spalding, 1983).
Particles and Flow Model
In this study, CFD simulations are performed initially in
batch mode mono-size fluidised bed hydrodynamics fol-
lowed by continuous mode mono-size particles with the
effect of inclined channels. A fixed particle density of ρs =
2650 kg/m3 is used for both batch and continuous mode
simulations. The model parameters are discussed in the fol-
lowing sections.
Batch Model for Fluidised Bed
This investigation employs a batch mode fluidisation model
to validate the CFD model. The operating parameters and
material properties are derived exclusively from experi-
ments conducted for the specific purpose of validation. It is
important to highlight that identical laboratory equipment
and materials are utilized for both the fluidised bed experi-
ments and the subsequent continuous flow experiments.
For the batch mode fluidised bed simulations, four
distinct particle sizes – 700, 600, 450, and 350 µm – are
selected to investigate the influence of particle size on bed
hydrodynamics. The bed is initialized with a height of
0.142 m and a solid volume fraction of 0.56. Across oth-
erwise identical experimental conditions, eight different
fluidisation rates are explored for each particle size: 0.4,
0.8, 1.2, 1.4, 1.6, 1.8, 2.0, and 2.4 L/min. Rather than
conducting eight separate simulations, the fluidisation rate
is incrementally increased in 2-second intervals (refer to
Figure 5c). Data from the simulations are then extracted
and analysed for every two-second interval to assess valida-
tion parameters at different fluidisation rates.
Validation parameters, namely bed height and bed dif-
ferential pressure, are assessed at the point of steady-state
conditions and subsequently compared with correspond-
ing experimental data. The bed height is determined by
measuring the maximum height attained by particles under
steady-state conditions. Meanwhile, the bed differential
pressure is computed as the mass flow-averaged total pres-
sure difference between two boundary surfaces, symbol-
izing the high- and low-pressure boundaries within the
domain. These pressure boundaries, referred to as observa-
tion planes, are positioned at the same heights as pressure
transducers in the experiments, as depicted in Figure 3.
Continuous Flow RC Model
A continuous flow model is constructed to match the
experimental conditions for Run 6 of Starrett and Galvin
(2023) The particle sizes for the continuous flow simula-
tions are chosen from the spectrum of particles employed in
the experiment, ranging from 30 to 710 µm. The simula-
tions initiate with an incoming slurry fed through the feed
inlet, and fluidisation water enters the domain through the
lower inlet. Additionally, a side-water inflow is introduced
through the side-water inlet (refer to Figure 1b). Operating
conditions for the continuous flow model are adopted from
the experiment and detailed in Table 1.
For simplicity, the underflow is omitted in the model,
and instead, a plane below the side-water inlet is con-
structed and utilized to monitor the underflow as particles
reach and pass through it. In the subsequent phase of the
study, the underflow will be incorporated into the CFD
model, along with multi-size particles. The partition of par-
ticles to underflow and overflow is obtained by monitoring
the mass flow rates at the outlet and across the underflow
plane. Partition numbers for each particle size are then fit-
ted to a partition curve as shown in Eq. (1).
expdln^3h
P
E
Dh 100
1
1
p
50
#=
+
-^D n
(1)
where D represents the particle diameter and D50 is defined
as the separation size. The sharpness of the partition curve
is quantified by the Ecarte Probable (Ep) with smaller values
denoting a sharper separation.
Geometry and Boundary Conditions
The numerical simulations closely follow the experimen-
tal approach undertaken by Starrett and Galvin (2023).
A three-dimensional computational domain is established
mirroring the specifications of the equipment utilised in
the experiment. The computational domain includes
inclined channels to replicate the configuration of the actual
equipment. The schematic diagram of the RC and the 3D
CFD domain are illustrated in Figure 1. Boundary condi-
tions consist of a velocity inlet at the domain’s inlet and
a pressure outlet at the outlet. Additionally, a side-water
inlet and feed inlet are incorporated in the domain for the
Table 1. Operating conditions for continuous flow (Starrett
&Galvin, 2023)
Operating Parameters Units
Feed Rate 22.0 t/m2/h
Feed Pulp %Solids 51.2 (wt. %)
Side-Water Rate 3.0 (L/min)
Fluidisation Rate 1.0 (L/min)