3748 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
track these interactions in real-time using acoustic sensing
techniques could potentially be useful for a more proac-
tive control strategy, which would improve the mills’ overall
performance.
The rock-water mixture (equivalent to AG mill condi-
tion) showed similar and lower acoustic emissions at dif-
ferent mill operating conditions and with the varying mass
of the rock sample. In the case of the steel ball-water mix-
ture, the mill noise emission was significantly higher and
equivalent to that obtained for only steel balls. This implies
the presence of water in the ball-water mixture had little
effect on the steel balls and liner/lifter interactions dur-
ing grinding. Furthermore, it can also be deduced from
the result that when mill slurry is extremely less viscous,
frequent collisions of steel balls and liner/lifter may occur.
When the rock sample is combined with steel balls in the
mill, it interferes with the steel balls and liner/lifter interac-
tions by acting as a cushion which dampens the mill noise
significantly, as reported in the literature (Watson, 1985).
During the process, the rock sample absorbed most of the
impact and noise emission energies caused by the steel balls
at the toe angle, resulting in significant breakage. As the
mill speed increases, more kinetic energy is transferred to
the cataracting steel balls, and the fraction of steel balls
that are overthrown above the mill’s toe angle increases the
acoustic emission. Based on these observations, the acoustic
emission can support the optimum operating conditions
of the AG/SAG mill along with particle size distribution
(PSD) and power draw. On the other hand, there was a
steady drop in the mill acoustic emission with an increase
in the rock mass (500 g, 1000 g, and 1500 g), while main-
taining the percentage of the steel ball loading. Higher rock
mass enhanced the restrictions of the ball-to-ball and ball-
to-liner collision and dampened the mill noise accordingly.
Understanding the relationship between mill rock loading
and acoustics can help determine whether the mill is under-
loading or overloading and set the required acoustic range
for optimal mill filling level.
The mill acoustic response was further reduced when
water was added to the rock-steel balls mixture (SAG mill
conditions) relative to the observation made for the rock-
steel balls mixture without water. This is because water addi-
tion created a slurry environment, which could affect the
rheological flow (viscosity) and charge dynamics. Also, the
slurry environment may form a dense coating around the
steel balls and liner/lifter, providing additional cushioning.
These factors may have a significant impact on noise reduc-
tion in the wet grinding environment involving a rock-ball-
water mixture. When the mill speed was changed from 50
to 60 rpm, the average noise intensity increased. A further
increase of the mill speed to 70 rpm led to a decrease in the
noise intensity. This observation explains that mill speed is
crucial and affects the charge trajectory inside the mill. The
drop in acoustic at 70 rpm (close to the mill critical speed
of 77.5) could suggest that some of the charge tends to cen-
trifuge, limiting their contribution to the noise emission by
impact. This reflected very well in the PSD by demonstrat-
ing poor rock size reduction. In addition, it can also be
stated that acoustic emission intensity is less prevalent in
wet grinding conditions (rock-ball-water mixture) than in
dry grinding (rock-ball mixture) where the slurry or pulp
environment is deficient.
To the lifter heights and configurations in Figure 4A,
the resulting mill acoustic emission increased with increas-
ing the lifter heights from 2 cm to 3.5 cm. Under the same
operating conditions, the height of the lifters inside the AG/
SAG mill could impact the trajectory of the mill. The low
lifter height of 2 cm projected the charge around the suit-
able toe angle, corresponding to an enhanced breakage rate
(PSD). When the lifter heights were changed to 3 cm and
3.5 cm, the mill acoustic intensity was increased further
accordingly, which is believed to project the charge con-
tinuously above the mill toe angle. The subsequent effect
was observed in reduced PSD in Figure 4B. The PSD can
be improved with reduced noise emission by reducing or
adjusting the mill speed, causing the charge inside the mill
to fall at the suitable toe position.
EFFECT OF FEED ORE HETEROGENEITY
IN BATCH MODE AND ACOUSTICS
The heterogeneity of the feed ore characterised by hardness
and size distribution has been identified as sensitive and
produces different acoustics in the grinding mill (Owusu
et al., 2022b Owusu et al., 2023a). In a preliminary study,
monitoring the acoustic response (from RMS and PSDE
feature extractions) of model quartz and iron ore in a lab-
oratory ball mill under the same conditions revealed that
model quartz (the hard mineral in this study) produced
grinding characteristics with higher acoustic intensity
than iron ore (the soft mineral) (Owusu et al., 2021d).
As demonstrated in Figure 5 (Owusu et al., 2021b), the
acoustic intensities of the two different rock hardness tend
to decrease gradually from 5 minutes to 20 minutes as the
production of fines increases with grinding time, which is
similar to the work done by Watson and Morrison (1985).
In the ball mill wet grinding study, the reduction in mill
acoustic emission with time can be attributed to the pro-
duction of finer particle size distribution, the formation of
slurry coating and varying viscosity that tend to retard the
direct and frequent collisions of the steel balls themselves
track these interactions in real-time using acoustic sensing
techniques could potentially be useful for a more proac-
tive control strategy, which would improve the mills’ overall
performance.
The rock-water mixture (equivalent to AG mill condi-
tion) showed similar and lower acoustic emissions at dif-
ferent mill operating conditions and with the varying mass
of the rock sample. In the case of the steel ball-water mix-
ture, the mill noise emission was significantly higher and
equivalent to that obtained for only steel balls. This implies
the presence of water in the ball-water mixture had little
effect on the steel balls and liner/lifter interactions dur-
ing grinding. Furthermore, it can also be deduced from
the result that when mill slurry is extremely less viscous,
frequent collisions of steel balls and liner/lifter may occur.
When the rock sample is combined with steel balls in the
mill, it interferes with the steel balls and liner/lifter interac-
tions by acting as a cushion which dampens the mill noise
significantly, as reported in the literature (Watson, 1985).
During the process, the rock sample absorbed most of the
impact and noise emission energies caused by the steel balls
at the toe angle, resulting in significant breakage. As the
mill speed increases, more kinetic energy is transferred to
the cataracting steel balls, and the fraction of steel balls
that are overthrown above the mill’s toe angle increases the
acoustic emission. Based on these observations, the acoustic
emission can support the optimum operating conditions
of the AG/SAG mill along with particle size distribution
(PSD) and power draw. On the other hand, there was a
steady drop in the mill acoustic emission with an increase
in the rock mass (500 g, 1000 g, and 1500 g), while main-
taining the percentage of the steel ball loading. Higher rock
mass enhanced the restrictions of the ball-to-ball and ball-
to-liner collision and dampened the mill noise accordingly.
Understanding the relationship between mill rock loading
and acoustics can help determine whether the mill is under-
loading or overloading and set the required acoustic range
for optimal mill filling level.
The mill acoustic response was further reduced when
water was added to the rock-steel balls mixture (SAG mill
conditions) relative to the observation made for the rock-
steel balls mixture without water. This is because water addi-
tion created a slurry environment, which could affect the
rheological flow (viscosity) and charge dynamics. Also, the
slurry environment may form a dense coating around the
steel balls and liner/lifter, providing additional cushioning.
These factors may have a significant impact on noise reduc-
tion in the wet grinding environment involving a rock-ball-
water mixture. When the mill speed was changed from 50
to 60 rpm, the average noise intensity increased. A further
increase of the mill speed to 70 rpm led to a decrease in the
noise intensity. This observation explains that mill speed is
crucial and affects the charge trajectory inside the mill. The
drop in acoustic at 70 rpm (close to the mill critical speed
of 77.5) could suggest that some of the charge tends to cen-
trifuge, limiting their contribution to the noise emission by
impact. This reflected very well in the PSD by demonstrat-
ing poor rock size reduction. In addition, it can also be
stated that acoustic emission intensity is less prevalent in
wet grinding conditions (rock-ball-water mixture) than in
dry grinding (rock-ball mixture) where the slurry or pulp
environment is deficient.
To the lifter heights and configurations in Figure 4A,
the resulting mill acoustic emission increased with increas-
ing the lifter heights from 2 cm to 3.5 cm. Under the same
operating conditions, the height of the lifters inside the AG/
SAG mill could impact the trajectory of the mill. The low
lifter height of 2 cm projected the charge around the suit-
able toe angle, corresponding to an enhanced breakage rate
(PSD). When the lifter heights were changed to 3 cm and
3.5 cm, the mill acoustic intensity was increased further
accordingly, which is believed to project the charge con-
tinuously above the mill toe angle. The subsequent effect
was observed in reduced PSD in Figure 4B. The PSD can
be improved with reduced noise emission by reducing or
adjusting the mill speed, causing the charge inside the mill
to fall at the suitable toe position.
EFFECT OF FEED ORE HETEROGENEITY
IN BATCH MODE AND ACOUSTICS
The heterogeneity of the feed ore characterised by hardness
and size distribution has been identified as sensitive and
produces different acoustics in the grinding mill (Owusu
et al., 2022b Owusu et al., 2023a). In a preliminary study,
monitoring the acoustic response (from RMS and PSDE
feature extractions) of model quartz and iron ore in a lab-
oratory ball mill under the same conditions revealed that
model quartz (the hard mineral in this study) produced
grinding characteristics with higher acoustic intensity
than iron ore (the soft mineral) (Owusu et al., 2021d).
As demonstrated in Figure 5 (Owusu et al., 2021b), the
acoustic intensities of the two different rock hardness tend
to decrease gradually from 5 minutes to 20 minutes as the
production of fines increases with grinding time, which is
similar to the work done by Watson and Morrison (1985).
In the ball mill wet grinding study, the reduction in mill
acoustic emission with time can be attributed to the pro-
duction of finer particle size distribution, the formation of
slurry coating and varying viscosity that tend to retard the
direct and frequent collisions of the steel balls themselves