XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 3773
suggest that in small mills, small balls may not be effective
for milling a feed of wide range of particle size distribution
(PSD), which the same size balls may be able to mill in
larger mills. It can also be argued that, for a feed of a narrow
PSD, appropriately sized balls in small mills may appear
more efficient than large balls just by the virtue of a narrow
collision energy distribution.
Figure 7 shows the spatial distribution of power
between shear and impact components of collisions in a
1.8 m Ø mill. It is evident in Figure 7 that in the region
of high power dissipation (marked by the red, yellow, and
green hues), the proportion of power that is taken by shear
component of collisions increases as the ball top-up size is
varied from large to small, as demonstrated by a larger area
covered by the red hues. The opposite occurs with the pro-
portion of power that is taken by the impact component of
collisions. Similar maps were obtained for all the other mill
sizes simulated.
The results from the DEM simulations presented here
suggest that the changes in the mill collision environment
because of changes in ball sizes are significant, and should
be considered in mechanistic models predicting ball size-
mill performance relationships. It appears that reducing
the ball size does not only increase the number of contact
sites, but increases the probability of attrition by increasing
the proportion of power taken by the shear components of
low-energy collisions. If the energies dissipated exceed the
threshold energy to effect breakage, the mill collision envi-
ronment response to varying the ball top-up size appears to
be in favour of the mill performance response one would be
looking for when switching to smaller ball sizes or to larger
ball sizes.
The difference on the extent of the effect of ball size
on collision environment from the different mills simu-
lated raises a question of what is the appropriate mill size
for batch testing the effect of ball size on grinding perfor-
mance. It is acknowledged that the lifter configuration has
an influence on the extent of the outcome from each mill
simulated. However, the lab-scale and pilot mills have little
(or no) room for changing the lifter configuration batch
tests are normally conducted without consideration of lifter
configuration.
The results presented in Figures 3 to 7 also suggest that
a model predicting the wear of balls in a mill should be a
weighted average of the abrasion driven wear model and the
impact driven wear model. This is because of the power dis-
sipated, the proportion that is taken by shear components
and impact components of collisions are both significant
for all the ball sizes simulated.
Effect of Mill Diameter on Collision Environment
Figure 3 and 4 also shows the effect of mill diameter on
collision environment. It is evident from the data presented
that moving from small lab-scale mill to pilot mills changed
the collision environment from being impact dominated to
shear dominated. For an example, when using a top-up size
of 30 mm balls, moving from a 0.3 m diameter mill to a 1.8
m diameter mill, the proportion of shear power increased
from 29% to 60%, a similar observation as that reported
in Yahyaei et al (2015). The authors acknowledge that the
mills compared here have different lifter configurations
(height &number), and the effect of lifter configuration is
not quantified. However, it is postulated that the effect is
Figure 6. Effect of ball size on energy spectra from the 0.6 m Ø mill
suggest that in small mills, small balls may not be effective
for milling a feed of wide range of particle size distribution
(PSD), which the same size balls may be able to mill in
larger mills. It can also be argued that, for a feed of a narrow
PSD, appropriately sized balls in small mills may appear
more efficient than large balls just by the virtue of a narrow
collision energy distribution.
Figure 7 shows the spatial distribution of power
between shear and impact components of collisions in a
1.8 m Ø mill. It is evident in Figure 7 that in the region
of high power dissipation (marked by the red, yellow, and
green hues), the proportion of power that is taken by shear
component of collisions increases as the ball top-up size is
varied from large to small, as demonstrated by a larger area
covered by the red hues. The opposite occurs with the pro-
portion of power that is taken by the impact component of
collisions. Similar maps were obtained for all the other mill
sizes simulated.
The results from the DEM simulations presented here
suggest that the changes in the mill collision environment
because of changes in ball sizes are significant, and should
be considered in mechanistic models predicting ball size-
mill performance relationships. It appears that reducing
the ball size does not only increase the number of contact
sites, but increases the probability of attrition by increasing
the proportion of power taken by the shear components of
low-energy collisions. If the energies dissipated exceed the
threshold energy to effect breakage, the mill collision envi-
ronment response to varying the ball top-up size appears to
be in favour of the mill performance response one would be
looking for when switching to smaller ball sizes or to larger
ball sizes.
The difference on the extent of the effect of ball size
on collision environment from the different mills simu-
lated raises a question of what is the appropriate mill size
for batch testing the effect of ball size on grinding perfor-
mance. It is acknowledged that the lifter configuration has
an influence on the extent of the outcome from each mill
simulated. However, the lab-scale and pilot mills have little
(or no) room for changing the lifter configuration batch
tests are normally conducted without consideration of lifter
configuration.
The results presented in Figures 3 to 7 also suggest that
a model predicting the wear of balls in a mill should be a
weighted average of the abrasion driven wear model and the
impact driven wear model. This is because of the power dis-
sipated, the proportion that is taken by shear components
and impact components of collisions are both significant
for all the ball sizes simulated.
Effect of Mill Diameter on Collision Environment
Figure 3 and 4 also shows the effect of mill diameter on
collision environment. It is evident from the data presented
that moving from small lab-scale mill to pilot mills changed
the collision environment from being impact dominated to
shear dominated. For an example, when using a top-up size
of 30 mm balls, moving from a 0.3 m diameter mill to a 1.8
m diameter mill, the proportion of shear power increased
from 29% to 60%, a similar observation as that reported
in Yahyaei et al (2015). The authors acknowledge that the
mills compared here have different lifter configurations
(height &number), and the effect of lifter configuration is
not quantified. However, it is postulated that the effect is
Figure 6. Effect of ball size on energy spectra from the 0.6 m Ø mill