XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 59
using a viscosity double that of water, evident from a sen-
sitivity analysis not shown here. It is also noted that in the
study by Rodrigues et al (2023) the low flow rate produced
a high Fe recovery of 87% but a low product grade of only
45% Fe. The values calculated in Table 2 are given as the
blue circular points. These two sets of results show good
agreement. The low separation densities arise from the rela-
tively low flow velocity, insufficient to carry the denser par-
ticles to the overflow.
It is further noted that for a suspension density equiva-
lent to that of water, given by the grey curve, the separation
densities are much lower. In fact, the hematite separates at
a size finer than 10 microns while the silicates separate at
a diameter of 20 mm. Clearly the potential to recover the
hematite and reject the silicates across the much broader size
range of 0–300 mm is not possible under dilute conditions.
Analysis of Run 13 – High Volumetric Feed Rate
In the study by Rodrigues et al (2023), Run 13 pro-
duced the best overall separation with an Fe recovery of
73% and product grade of 65.6% Fe (pure hematite is
69.9% Fe). The improved performance was attributed
to the higher volumetric feed rate of 8 L/min, and lower
suspension viscosity achieved by reducing the feed pulp
density to 17 wt% solids. In this case the viscosity was set
equal to that for water. The overall component balance
is given below in Table 3 showing the overflow rate was
7.73 L/min corresponding to a superficial channel velocity
of 0.0179 m/s.
Table 4 shows the results obtained for Run 13 by
combining the laminar flow and hindered settling models.
For example, the terminal velocity of a particle of diam-
eter, d=10 mm, and density, rp=9477 kgm–3 in water is
Table 2. Summary of parameter values for Run 1 based on n=5, U′=0.0042 m/s,
rsusp=1500 kgm–3, m=0.002 Nsm–2
d,
mm
D
50, kgm–3
u
t, m/s
f
-
Ul ,
m/s
10 5495 0.000118 0.624 0.0042
30 3356 0.000575 0.385 0.0042
50 2889 0.001263 0.292 0.0042
70 2671 0.002144 0.241 0.0042
90 2541 0.003189 0.208 0.0042
110 2455 0.004378 0.186 0.0042
130 2392 0.005689 0.169 0.0042
150 2345 0.007108 0.156 0.0042
0
1000
2000
3000
4000
5000
6000
0 20 40 60 80 100 120 140
Particle Diameter (m)
Figure 3. Variation in the separation density, D
50 ,with the particle diameter for Run 1
of Rodrigues et al (2023), denoted by orange triangles. The data calculated in this study
is given by the blue circles based on an assumed suspension density of 1500 kgm–3. The
lower result given by the grey circles is based on a suspension density of 1000 kgm–3
Separation
Density
(kgm-3)
using a viscosity double that of water, evident from a sen-
sitivity analysis not shown here. It is also noted that in the
study by Rodrigues et al (2023) the low flow rate produced
a high Fe recovery of 87% but a low product grade of only
45% Fe. The values calculated in Table 2 are given as the
blue circular points. These two sets of results show good
agreement. The low separation densities arise from the rela-
tively low flow velocity, insufficient to carry the denser par-
ticles to the overflow.
It is further noted that for a suspension density equiva-
lent to that of water, given by the grey curve, the separation
densities are much lower. In fact, the hematite separates at
a size finer than 10 microns while the silicates separate at
a diameter of 20 mm. Clearly the potential to recover the
hematite and reject the silicates across the much broader size
range of 0–300 mm is not possible under dilute conditions.
Analysis of Run 13 – High Volumetric Feed Rate
In the study by Rodrigues et al (2023), Run 13 pro-
duced the best overall separation with an Fe recovery of
73% and product grade of 65.6% Fe (pure hematite is
69.9% Fe). The improved performance was attributed
to the higher volumetric feed rate of 8 L/min, and lower
suspension viscosity achieved by reducing the feed pulp
density to 17 wt% solids. In this case the viscosity was set
equal to that for water. The overall component balance
is given below in Table 3 showing the overflow rate was
7.73 L/min corresponding to a superficial channel velocity
of 0.0179 m/s.
Table 4 shows the results obtained for Run 13 by
combining the laminar flow and hindered settling models.
For example, the terminal velocity of a particle of diam-
eter, d=10 mm, and density, rp=9477 kgm–3 in water is
Table 2. Summary of parameter values for Run 1 based on n=5, U′=0.0042 m/s,
rsusp=1500 kgm–3, m=0.002 Nsm–2
d,
mm
D
50, kgm–3
u
t, m/s
f
-
Ul ,
m/s
10 5495 0.000118 0.624 0.0042
30 3356 0.000575 0.385 0.0042
50 2889 0.001263 0.292 0.0042
70 2671 0.002144 0.241 0.0042
90 2541 0.003189 0.208 0.0042
110 2455 0.004378 0.186 0.0042
130 2392 0.005689 0.169 0.0042
150 2345 0.007108 0.156 0.0042
0
1000
2000
3000
4000
5000
6000
0 20 40 60 80 100 120 140
Particle Diameter (m)
Figure 3. Variation in the separation density, D
50 ,with the particle diameter for Run 1
of Rodrigues et al (2023), denoted by orange triangles. The data calculated in this study
is given by the blue circles based on an assumed suspension density of 1500 kgm–3. The
lower result given by the grey circles is based on a suspension density of 1000 kgm–3
Separation
Density
(kgm-3)