XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 1529
The results have an R2 of 0.96, showing a good cor-
relation, and with an excellent dispersion throughout the
parity. The high energy results are expected to generally per-
form better during the Geopyörä breakage test. Therefore,
the main GPI result is calculated with the high energy test
data, which is the chosen method to calculate the Geopyörä
Index.
A Student’s t-Test was employed to investigate the
relationship between GPI values derived from low energy
(GPI_LE) and high energy (GPI_HE) Geopyörä tests, con-
firming the parity between the two. Despite the influence
of energy levels on fragmentation, the test showed that the
mean GPI value at low energy is not significantly less than
its high energy counterpart, with a p-value of 0.105, defy-
ing initial predictions. This result, corroborated by the high
R2 value of 0.96 in Figure 5, indicates a substantial correla-
tion between the two energy levels’ GPI values. Moreover,
the analysis provides a statistical assurance by depicting a
95% upper bound for the mean difference, ensuring with
95% confidence that the true disparity is less than 0.13904.
Such an upper bound suggests that the difference between
the GPI values at different energy inputs is minor.
CONCLUSIONS
The Geopyörä Index represents a new breakage resistance
parameter with the potential to be used in modelling of com-
minution circuits as a direct parameter in power-models,
be integrated into both future and existing power models
to assist in comminution circuit design and optimization.
DATA ANALYSIS
As part of the analysis of the Geopyörä Index, it is possible
to demonstrate the correlation between the SMC DWI and
the new parameter, since the SMC test results are available
for all samples used in this work. Figure 4 shows the plot of
the SMC Test DWI as a function of the GPI.
A good correlation is observed, with a coefficient of
determination (R2) of 0.97. The linear trendline presents
a slope of 0.65, demonstrating that the GPI is a parameter
with a value about 35% higher than the DWI, being pos-
sible not only to easily convert one index to the other by
using this linear correlation, but also to directly use the GPI
for the same applications as the DWI, such as com-
minution modelling.
Since the GPI is a function of the Ecs and the t10, it is
also possible to calculate two GPI values for each sample by
using the two Geopyörä energy levels that are derived from
the standard test procedure. Since fragmentation increases
proportionally with energy, given that a higher breakage
energy generates more fines, it is expected that there will
be a good parity between the high energy GPI and the low
energy GPI. Figure 5 presents the parity chart for the two
GPI values calculated for each of the two energy samples in
the database.
y =0.918x
=0.9632
0
5
10
15
20
25
0 5 10 15 20 25
Low energy GPI (kWh/m³)
Figure 5. High and low energy GPI parity plot
High
energy
GPI
(kWh/m³)
Previous Page Next Page

Extracted Text (may have errors)

XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 1529
The results have an R2 of 0.96, showing a good cor-
relation, and with an excellent dispersion throughout the
parity. The high energy results are expected to generally per-
form better during the Geopyörä breakage test. Therefore,
the main GPI result is calculated with the high energy test
data, which is the chosen method to calculate the Geopyörä
Index.
A Student’s t-Test was employed to investigate the
relationship between GPI values derived from low energy
(GPI_LE) and high energy (GPI_HE) Geopyörä tests, con-
firming the parity between the two. Despite the influence
of energy levels on fragmentation, the test showed that the
mean GPI value at low energy is not significantly less than
its high energy counterpart, with a p-value of 0.105, defy-
ing initial predictions. This result, corroborated by the high
R2 value of 0.96 in Figure 5, indicates a substantial correla-
tion between the two energy levels’ GPI values. Moreover,
the analysis provides a statistical assurance by depicting a
95% upper bound for the mean difference, ensuring with
95% confidence that the true disparity is less than 0.13904.
Such an upper bound suggests that the difference between
the GPI values at different energy inputs is minor.
CONCLUSIONS
The Geopyörä Index represents a new breakage resistance
parameter with the potential to be used in modelling of com-
minution circuits as a direct parameter in power-models,
be integrated into both future and existing power models
to assist in comminution circuit design and optimization.
DATA ANALYSIS
As part of the analysis of the Geopyörä Index, it is possible
to demonstrate the correlation between the SMC DWI and
the new parameter, since the SMC test results are available
for all samples used in this work. Figure 4 shows the plot of
the SMC Test DWI as a function of the GPI.
A good correlation is observed, with a coefficient of
determination (R2) of 0.97. The linear trendline presents
a slope of 0.65, demonstrating that the GPI is a parameter
with a value about 35% higher than the DWI, being pos-
sible not only to easily convert one index to the other by
using this linear correlation, but also to directly use the GPI
for the same applications as the DWI, such as com-
minution modelling.
Since the GPI is a function of the Ecs and the t10, it is
also possible to calculate two GPI values for each sample by
using the two Geopyörä energy levels that are derived from
the standard test procedure. Since fragmentation increases
proportionally with energy, given that a higher breakage
energy generates more fines, it is expected that there will
be a good parity between the high energy GPI and the low
energy GPI. Figure 5 presents the parity chart for the two
GPI values calculated for each of the two energy samples in
the database.
y =0.918x
=0.9632
0
5
10
15
20
25
0 5 10 15 20 25
Low energy GPI (kWh/m³)
Figure 5. High and low energy GPI parity plot
High
energy
GPI
(kWh/m³)

Help

loading