SAG Mill Prediction (Axb Method)

The JK Drop Weight Test Axb parameter captures the ore's resistance to impact breakage. A higher Axb indicates a softer ore that breaks more readily, leading to more steel-on-steel contact and higher media wear. The linear correlation (R² = 0.79) was derived from 8 Newmont operations; Mill 5, Lone Tree, and Juniper were excluded as outliers.

SAG Mill Prediction (Steel Fraction Method)

Steel fraction in the mill charge is an operational parameter (not an ore property). Higher steel loading means more ball-on-ball abrasion, increasing wear. This correlation (R² = 0.85) is stronger than the Axb method and useful when the JK Drop Weight test has not been performed.

Ball Mill Prediction

The Bond (1963) formula is the industry standard but consistently overpredicts actual wear by 2–3× in modern operations. This is likely due to improvements in ball quality (chemistry, hardness, manufacturing) since Bond's era. The Newmont power fit (~50% of Bond) better matches 2009-era operational data.

Limitations

• Models are empirical fits to Newmont operations (2007–2010 data). Applicability to non-Newmont circuits should be validated.
• The SAG Axb model has moderate scatter (R² = 0.79). Individual operations may deviate 20–40% from predicted.
• Ball quality (chemistry, hardness) significantly affects wear but is not explicitly modelled — the correlations assume modern forged/cast media at 58–63 Rockwell C.
• Feed size (F80), ore mineralogy, and slurry chemistry are not captured in these simplified correlations.

References

• Bond, F.C. (1963). Metal wear in crushing and grinding. AIChE, 54th Annual Meeting.
• Giblett, A. & Seidel, J. (2011). Measuring, Predicting and Managing Grinding Media Wear. Newmont Mining Corporation.
• Napier-Munn, T.J. et al. (1996). Mineral Comminution Circuits. JKMRC, University of Queensland.
• Radziszewski, P. (2002). Exploring total media wear. Minerals Engineering, 15(12), 1073–1087.