2298 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
comparison of the average content under different experi-
mental conditions.
It was observed that the iron content in the concen-
trate increases with the addition of flotation time, reach-
ing 62.58 ± 1.07% at 4 minutes, but experiences a slight
decrease at 5 minutes. Regarding the statistical analysis,
the datasets produced in the three flotation stages met the
Anderson-Darling criterion for the response variables, thus
being treated as normally distributed. The Bartlett tests,
with a 95% confidence level, yielded a p-value greater than
0.05 for all experiments, indicating that there is no sig-
nificant difference between the variances. Therefore, in the
ANOVA test, it was assumed that the variances are equal.
The analysis of variance indicated that the variation in the
time factor had a greater contribution to the response vari-
able than the random error, resulting in a p-value less than
0.05. Consequently, it is concluded that the flotation time
in the studied range is a statistically significant variable in
terms of Fe content in the concentrate.
Through the Tukey test analysis, which simultaneously
compares all experiments, when the green dashed line con-
taining the zero-value represented in Figure 1b does not
coincide with the range of the blue bar, it means that the
compared means are different. Thus, a significant differ-
ence only occurs when comparing the experiments using 1
minute with the others, and between the experiments using
2 and 4 minutes. From 3 minutes onwards, there is no sig-
nificant difference between the longer flotation times.
Regarding the recovery of Fe, it was observed that there
is an increase with the increase in flotation time up to 3
minutes, reaching 82,64 ± 2.09%, but a slight decrease
occurs at 4 minutes. In this case, the analysis of variance
indicated that the random error had a greater contribu-
tion to the response variable than the flotation time factor,
resulting in a p-value greater than 0.05. Therefore, the flo-
tation time in the analyzed range is not a statistically signifi-
cant variable in terms of Fe recovery in the concentrate, and
thus, the Tukey test is not necessary. For this reason, the
definition of the flotation time was based on the analysis
of the Fe content, theoretically selecting the time of 3 min-
utes since there is no significant variation in content from
this time onwards. However, it was found that in most tests
there was still a reasonable amount of foam when complet-
ing 3 minutes. Therefore, a time of 4 minutes was selected
for subsequent stages.
In the second stage of flotation tests, using the experi-
mental standards from Table 1, experiments were con-
ducted using the standard depressor corn gritz and the
alternative depressor sorghum flour. The results comparing
the two sets of tests are presented in Figure 2.
Figure 2. Influence of the depressant type on flotation performance
comparison of the average content under different experi-
mental conditions.
It was observed that the iron content in the concen-
trate increases with the addition of flotation time, reach-
ing 62.58 ± 1.07% at 4 minutes, but experiences a slight
decrease at 5 minutes. Regarding the statistical analysis,
the datasets produced in the three flotation stages met the
Anderson-Darling criterion for the response variables, thus
being treated as normally distributed. The Bartlett tests,
with a 95% confidence level, yielded a p-value greater than
0.05 for all experiments, indicating that there is no sig-
nificant difference between the variances. Therefore, in the
ANOVA test, it was assumed that the variances are equal.
The analysis of variance indicated that the variation in the
time factor had a greater contribution to the response vari-
able than the random error, resulting in a p-value less than
0.05. Consequently, it is concluded that the flotation time
in the studied range is a statistically significant variable in
terms of Fe content in the concentrate.
Through the Tukey test analysis, which simultaneously
compares all experiments, when the green dashed line con-
taining the zero-value represented in Figure 1b does not
coincide with the range of the blue bar, it means that the
compared means are different. Thus, a significant differ-
ence only occurs when comparing the experiments using 1
minute with the others, and between the experiments using
2 and 4 minutes. From 3 minutes onwards, there is no sig-
nificant difference between the longer flotation times.
Regarding the recovery of Fe, it was observed that there
is an increase with the increase in flotation time up to 3
minutes, reaching 82,64 ± 2.09%, but a slight decrease
occurs at 4 minutes. In this case, the analysis of variance
indicated that the random error had a greater contribu-
tion to the response variable than the flotation time factor,
resulting in a p-value greater than 0.05. Therefore, the flo-
tation time in the analyzed range is not a statistically signifi-
cant variable in terms of Fe recovery in the concentrate, and
thus, the Tukey test is not necessary. For this reason, the
definition of the flotation time was based on the analysis
of the Fe content, theoretically selecting the time of 3 min-
utes since there is no significant variation in content from
this time onwards. However, it was found that in most tests
there was still a reasonable amount of foam when complet-
ing 3 minutes. Therefore, a time of 4 minutes was selected
for subsequent stages.
In the second stage of flotation tests, using the experi-
mental standards from Table 1, experiments were con-
ducted using the standard depressor corn gritz and the
alternative depressor sorghum flour. The results comparing
the two sets of tests are presented in Figure 2.
Figure 2. Influence of the depressant type on flotation performance