Optimization of Particle Size and Specific Surface Area of Pellet Feed in Dry Ball Mill using Central Composite Design

Objective: The dimensional properties of iron ore pellet feed including specific surface area and particle size distribution in the ball mill was studied using response surface area method. Methods/Statistical analysis: The effect of the operational parameters of dry ball mill including ball charge (20-40%), grinding time (30-50 min) and balling distribution (Small, Mixed and Large) on dimensional properties of pellet feed was meticulously examined and optimized using response surface methodology based on Central Composite Design (CCD). Responses were 80% passing size (D80) and Blaine (BL). A total of 30 grinding experiments were designed and carried out in the CCD method. Regression models and response surfaces were obtained for each response. Findings: The results predicted by these models showed good agreement with experimental values. In the studied range it was deduced that D80 is decreased and BL increased when grinding time and ball charge increased. Also, the production of fine particles improved using balling distribution of level B (mix balls), instead of one ball size distribution. 3D graphs were applied to visually evaluate each parameter effect on responses. The optimum conditions of ball mill operational parameters to reach the best pellet feed for each blast furnace burden and direct reduction iron (DRI) pellet feed were revealed using CCD optimization. The predicted values BL and D80 were found to be in a reasonable agreement with the experimental values, with R2 as correlation factor being 0.91 and 0.92 for BL and D80, respectively. Improvements: The optimum condition for having a suitable blast furnace burden was: Ball distribution level B, Ball charge= 21.8-22.4 %, Grinding time= 40.1-42.9 min. Also, the optimal condition for DRI pellet feed preparation was: Ball distribution level B, Ball charge= 21.8-22.1 %, Grinding time= 42.1-44.6 min.