Abstract
This study employs response surface methodology (RSM) with a custom optimal
design to develop and optimize iron (Fe) grade and recovery through magnetization
roasting followed by induced dry-roll magnetic separation. The relationships between the
independent and dependent variables are investigated. The effect of roasting temperature,
magnetization roasting time, magnetic field intensity, rotor speed, and product splitter
position for the induced dry-roll magnetic separator on Fe grade and recovery are studied.
Suitable models are generated to predict the optimum operating conditions. An analysis of
Variance (ANOVA) is employed to validate the developed regression models’ adequacy
and assess the main and interaction-related effects on Fe grade and recovery. During
magnetization roasting, a satisfactory Fe grade of 66.8% with a recovery of 16.7% was
obtained under optimal conditions of 1050 ◦C for 97 min. Conversely, after the induced
dry-roll magnetic separator optimization, an Fe grade of 66.1% with a recovery of 60.2%
was achieved under optimum conditions of 0.105 T for magnetic field intensity, 70 Hz for
rotor speed, and an 11 mm product splitter position. This study effectively illustrates how
RSM can model the processes of magnetization roasting and induced dry-roll magnetic
separation, particularly concerning the operating parameters used for treating iron ore
plant tailings. Furthermore, it highlights the efficiency of this methodology in generating
substantial insights in a short timeframe while minimizing the number of experiments
conducted.