The use of a factorial experimental design to predict the flotation performance of fluorite ore

摘要

The effects of process variables on the recovery and grade of fluorite were determined using "design of experiment" [DOE] and Box-Behnken approaches. These models allowed the construction of the response surface graphs which presented the effects on the recovery and grade of the collector, depressant and frother dosages. The preliminary results indicate that DOE led to a rougher concentrate of 60% CaF_2, with a recovery of 98%. This showed a remarkable improvement compared to a plant baseline of 60% recovery and a grade of 42% CaF_2. However, additional test work employing a cleaner flotation stage needs to be done in order to produce a consistent final grade concentrate. This study shows that the response surface methodology can be used to model and optimize fluorite flotation.