DISCRETE ELEMENT METHOD SIMULATION OF GRINDING MEDIA MOTION IN BALL MILLS: METHODOLOGY, VERIFICATION, AND APPLICATION TO WEAR PREDICTION

Abstract

Predicting the wear of grinding media and liners in ball mills requires detailed knowledge of how the granular charge moves and impacts the mill shell, information that cannot be obtained directly from full-scale industrial operation. This paper presents a Discrete Element Method (DEM) simulation methodology, developed as part of a doctoral study on mining equipment durability and condition diagnostics, for modeling grinding media motion and the resulting contact forces in a tumbling ball mill. The model treats each grinding ball as a discrete body governed by Newton's second law and resolves ball-ball and ball-liner interactions using a linear spring-dashpot contact model with Coulomb friction, integrated in time with an explicit leapfrog scheme and implemented in the ROCKY DEM software. Simulations were carried out for a factorial set of operating conditions spanning rotational speed, fill level, and ball diameter, and the resulting charge kinematics and contact forces were coupled with a finite-element model of the mill drive train to evaluate impact-induced gearbox vibration. The DEM model was verified against full-scale flash-stop experiments on an industrial mill, with predicted charge angles (shoulder, toe, departure, and impact-point) agreeing with measured values to within 4%. Results show that fill level and rotational speed strongly govern contact force magnitude, fall height, and the transition from cascading to cataracting motion, and that DEM-derived impact statistics correlate with acoustic indicators of liner wear. The findings demonstrate that verified DEM simulation provides a reliable basis for wear prediction, operating-regime optimization, and condition-based diagnostics in industrial ball mills.