2876 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
the center to the sides and mixed with the downward wash
water flow. The wash water flow then dominated the liq-
uid motion close to both sides of the reverse fluidized bed
and a downward flow could be observed. When this down-
ward flow encountered with the liquid discharged from the
downcomer near the downcomer outlet, it flowed from the
side to the center and rejoined the uprising flow dominated
by the rising bubbles. As shown in Figure 4 for tests 1 and
3, the uprising flow dominated by the rising bubbles incor-
porated with the downward flow dominated by the down-
ward wash water to create a large recirculation loop within
the reverse fluidized bed. Through this large recirculation
loop, the wash water could quickly mix with the uprising
liquid containing entrained particles near the downcomer
outlet. Furthermore, this large recirculation loop could also
provide a fast drainage of entrained solid when it recircu-
lated back from the uprising flow to downward flow near
the wash water inlet. Hence, in tests 1 and 3, the gangue
entrainment was lower compared to test 2.
In addition, a small side flow at the wash water inlet
was found in test 3 at the highest gas flux. This flow joined
the flow leaving the concentrate outlet (indicated by the
two red arrows near the wash water inlet in the plot for
test 3). As this side flow directly came from the wash water
inlet, it mainly contained clean wash water. Hence, when
it joined the flow leaving the concentrate stream, it could
further dilute the liquid containing entrained particles to
reduce gangue entrainment.
Hence, the CFD simulations identified that when the gas
flux was decreased by 20% in test 2 from the baseline gas flux
in test 1, the bubbles were likely to accumulate in the reverse
fluidized bed due to the lower rising velocity, which dimin-
ished the reduction in air fraction. Meanwhile, the weaker
uprising flow due to the smaller gas flux in test 2 also limited
the formation of a large recirculation loop in the reverse flu-
idized bed, slowing down the drainage of entrained particles
and the mixing of wash water with the liquid containing
entrained particles. This resulted in higher gangue entrain-
ment in test 2. In contrast, a large recirculation loop was
formed in tests 1 and 3 due to higher gas fluxes, promoting
the fast mixing of wash water and the liquid with entrained
particles and the fast drainage of these particles, enhancing
Figure 4. Vector plots for the liquid phase near the downcomer outlet and wash water inlet in RFC at different gas fluxes
the center to the sides and mixed with the downward wash
water flow. The wash water flow then dominated the liq-
uid motion close to both sides of the reverse fluidized bed
and a downward flow could be observed. When this down-
ward flow encountered with the liquid discharged from the
downcomer near the downcomer outlet, it flowed from the
side to the center and rejoined the uprising flow dominated
by the rising bubbles. As shown in Figure 4 for tests 1 and
3, the uprising flow dominated by the rising bubbles incor-
porated with the downward flow dominated by the down-
ward wash water to create a large recirculation loop within
the reverse fluidized bed. Through this large recirculation
loop, the wash water could quickly mix with the uprising
liquid containing entrained particles near the downcomer
outlet. Furthermore, this large recirculation loop could also
provide a fast drainage of entrained solid when it recircu-
lated back from the uprising flow to downward flow near
the wash water inlet. Hence, in tests 1 and 3, the gangue
entrainment was lower compared to test 2.
In addition, a small side flow at the wash water inlet
was found in test 3 at the highest gas flux. This flow joined
the flow leaving the concentrate outlet (indicated by the
two red arrows near the wash water inlet in the plot for
test 3). As this side flow directly came from the wash water
inlet, it mainly contained clean wash water. Hence, when
it joined the flow leaving the concentrate stream, it could
further dilute the liquid containing entrained particles to
reduce gangue entrainment.
Hence, the CFD simulations identified that when the gas
flux was decreased by 20% in test 2 from the baseline gas flux
in test 1, the bubbles were likely to accumulate in the reverse
fluidized bed due to the lower rising velocity, which dimin-
ished the reduction in air fraction. Meanwhile, the weaker
uprising flow due to the smaller gas flux in test 2 also limited
the formation of a large recirculation loop in the reverse flu-
idized bed, slowing down the drainage of entrained particles
and the mixing of wash water with the liquid containing
entrained particles. This resulted in higher gangue entrain-
ment in test 2. In contrast, a large recirculation loop was
formed in tests 1 and 3 due to higher gas fluxes, promoting
the fast mixing of wash water and the liquid with entrained
particles and the fast drainage of these particles, enhancing
Figure 4. Vector plots for the liquid phase near the downcomer outlet and wash water inlet in RFC at different gas fluxes