XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 2845
with a window of 31 seconds to reduce the signal noise. An
example is shown in Figure 5, which presents the results of
online measurements of froth height and froth velocity for
the base case the air recovery was calculated and then plot-
ted against the superficial gas velocity. Similar experiments
were performed in triplicate for each funnel design, varying
Jg in randomised tests while keeping impeller speed and
froth depth constant (1200 rpm and approximately 3.5 cm,
respectively).
The summarised results obtained in the single-species
system are shown in Figure 6, presenting the weighted aver-
age air recovery obtained for each superficial gas velocity
and its variability, represented by the weighted standard
deviation. For each design, a quadratic curve was fitted to
all the air recovery data, considering every repeat. These
curves are arbitrary and are not intended as a representative
model for air recovery but are useful to visualise the results.
The results show that all the funnels generate an
improvement in froth stability, especially at higher airflows,
where this improvement is statistically significant for both
funnels. Funnel 2 resulted in the most stable froth, fol-
lowed by Funnel 1 and then the base case without a funnel.
The improved froth stability obtained with Funnel 2 could
be linked to changes in the air fluid dynamics as this design
redirects the air towards the centre of the tank, decreasing
the fast flow near the walls, while allowing bubbles to rise
Figure 5. Froth height (blue), froth velocity (red) and air recovery (green) measured for the base case scenario. The black lines
represent the filtered data for each case, while the mean and confidence interval are shown in each respective colour. The
greyscale represents the Jg used during the corresponding time. The white spaces between the greyscale bars represent the time
left for the system to stabilise after changing Jg values
with a window of 31 seconds to reduce the signal noise. An
example is shown in Figure 5, which presents the results of
online measurements of froth height and froth velocity for
the base case the air recovery was calculated and then plot-
ted against the superficial gas velocity. Similar experiments
were performed in triplicate for each funnel design, varying
Jg in randomised tests while keeping impeller speed and
froth depth constant (1200 rpm and approximately 3.5 cm,
respectively).
The summarised results obtained in the single-species
system are shown in Figure 6, presenting the weighted aver-
age air recovery obtained for each superficial gas velocity
and its variability, represented by the weighted standard
deviation. For each design, a quadratic curve was fitted to
all the air recovery data, considering every repeat. These
curves are arbitrary and are not intended as a representative
model for air recovery but are useful to visualise the results.
The results show that all the funnels generate an
improvement in froth stability, especially at higher airflows,
where this improvement is statistically significant for both
funnels. Funnel 2 resulted in the most stable froth, fol-
lowed by Funnel 1 and then the base case without a funnel.
The improved froth stability obtained with Funnel 2 could
be linked to changes in the air fluid dynamics as this design
redirects the air towards the centre of the tank, decreasing
the fast flow near the walls, while allowing bubbles to rise
Figure 5. Froth height (blue), froth velocity (red) and air recovery (green) measured for the base case scenario. The black lines
represent the filtered data for each case, while the mean and confidence interval are shown in each respective colour. The
greyscale represents the Jg used during the corresponding time. The white spaces between the greyscale bars represent the time
left for the system to stabilise after changing Jg values