XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 3935
overlapping sphericity distribution of the pores after break-
age emphasizes an unchanged pore morphology despite
breakage, underscoring the brittle breakage nature of the
investigated slag.
Due to the fine, dendritic intergrowth within the
matrix and the lack of correlation with existing pores, ran-
dom breakage was observed. For more precise studies of
breakage characteristics, it is imperative to produce target
phase structures that are larger and preferably individu-
ally distributed, i.e., not extensively intergrown within the
matrix. The major advantage is that the partial volume
effect in XCT measurements, which is dominant in the
presence of numerous grain boundaries, can be avoided.
This artifact occurs when certain features within a volume
can only be described by a few voxels, potentially affecting
measurement accuracy. Fine pores, as well as subtle cracks
or pre-existing damage that are below the resolution limit
or described by too few voxels, can be combined as a mixed
gray values assigned to the target or matrix phase. Another
advantage is the avoidance of pseudo-homogeneity in the
sample. With a more pronounced and locally prevalent tar-
get phase, the influence of target phase properties on break-
age behavior can be more easily investigated compared to
when it is extensively intergrown throughout the sample
volume.
Another approach is to prepare the particles at a
smaller scale, which reduces the dominance of the den-
dritic structure in individual grains. This enables measure-
ments at higher resolutions, facilitating the investigation
of the brekage mechanism at the microscale. In the case
of slag that compose mainly minerals and behaves like a
glassy material and containing minimal to no pure metal
inclusions, a brittle breakage behavior can be assumed.
Capturing crack propagation using XCT is challenging
in such cases due to the high speed of crack propagation.
As a reference since quartz was a main component of the
studied slag, Martinelli’s works have demonstrated that
quartz particles can reach speeds of 650 m/s (Martinelli
et al. 2020). If a highly brittle breakage behavior is pre-
sumed, cracks may not initially form under stress until the
actual breakage occurs. In this case, it is not necessary to
conduct in-situ experiments at multiple force levels under
XCT. Instead, experiments can be limited to the unbroken
state and measurements taken following a force drop in the
force-distance curve.
Concerning the comminution process, this aprticular
slag system has been confirmed to show a random break-
age behavior (Võ et al. 2024). To energetically optimize the
comminution process and to improve the accessibility of
the EnAMs, a non-random breakage behavior is desired. In
the presence of such fine structures, multiple comminution
steps are required to achieve the smallest liberation prod-
uct size. As the comminution work generally decreases with
increasing product particle size, larger target phase grains
should be more energy-efficient to liberate than smaller
ones, allowing for better utilization of the comminution
work.
CONCLUSIONS
The comprehensive investigation into ex-situ and in-situ
single particle breakage experiments and subsequent opti-
cal examinations has provided valuable insights into the
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1 10 100 1000
sphere equivalent pore diameter /(μm)
F =5 N
F =100 N
broken
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 0.2 0.4 0.6 0.8 1
sphericity /(-)
F =5 N
F =100 N
broken
Figure 6. Sum distribution of sphere equivalent pore diameter and pore sphericities at different set force levels during in-situ
compression
sum
distributi
/
(-)
sum
distributi
/
(-)
overlapping sphericity distribution of the pores after break-
age emphasizes an unchanged pore morphology despite
breakage, underscoring the brittle breakage nature of the
investigated slag.
Due to the fine, dendritic intergrowth within the
matrix and the lack of correlation with existing pores, ran-
dom breakage was observed. For more precise studies of
breakage characteristics, it is imperative to produce target
phase structures that are larger and preferably individu-
ally distributed, i.e., not extensively intergrown within the
matrix. The major advantage is that the partial volume
effect in XCT measurements, which is dominant in the
presence of numerous grain boundaries, can be avoided.
This artifact occurs when certain features within a volume
can only be described by a few voxels, potentially affecting
measurement accuracy. Fine pores, as well as subtle cracks
or pre-existing damage that are below the resolution limit
or described by too few voxels, can be combined as a mixed
gray values assigned to the target or matrix phase. Another
advantage is the avoidance of pseudo-homogeneity in the
sample. With a more pronounced and locally prevalent tar-
get phase, the influence of target phase properties on break-
age behavior can be more easily investigated compared to
when it is extensively intergrown throughout the sample
volume.
Another approach is to prepare the particles at a
smaller scale, which reduces the dominance of the den-
dritic structure in individual grains. This enables measure-
ments at higher resolutions, facilitating the investigation
of the brekage mechanism at the microscale. In the case
of slag that compose mainly minerals and behaves like a
glassy material and containing minimal to no pure metal
inclusions, a brittle breakage behavior can be assumed.
Capturing crack propagation using XCT is challenging
in such cases due to the high speed of crack propagation.
As a reference since quartz was a main component of the
studied slag, Martinelli’s works have demonstrated that
quartz particles can reach speeds of 650 m/s (Martinelli
et al. 2020). If a highly brittle breakage behavior is pre-
sumed, cracks may not initially form under stress until the
actual breakage occurs. In this case, it is not necessary to
conduct in-situ experiments at multiple force levels under
XCT. Instead, experiments can be limited to the unbroken
state and measurements taken following a force drop in the
force-distance curve.
Concerning the comminution process, this aprticular
slag system has been confirmed to show a random break-
age behavior (Võ et al. 2024). To energetically optimize the
comminution process and to improve the accessibility of
the EnAMs, a non-random breakage behavior is desired. In
the presence of such fine structures, multiple comminution
steps are required to achieve the smallest liberation prod-
uct size. As the comminution work generally decreases with
increasing product particle size, larger target phase grains
should be more energy-efficient to liberate than smaller
ones, allowing for better utilization of the comminution
work.
CONCLUSIONS
The comprehensive investigation into ex-situ and in-situ
single particle breakage experiments and subsequent opti-
cal examinations has provided valuable insights into the
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
1 10 100 1000
sphere equivalent pore diameter /(μm)
F =5 N
F =100 N
broken
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
0 0.2 0.4 0.6 0.8 1
sphericity /(-)
F =5 N
F =100 N
broken
Figure 6. Sum distribution of sphere equivalent pore diameter and pore sphericities at different set force levels during in-situ
compression
sum
distributi
/
(-)
sum
distributi
/
(-)