Optimize SAG Mill Wear Using DEM Simulation

Avoiding machine stoppage is a continual challenge in the mining industry, since downtime hours cost valuable production time and money. In the case of semi-autogenous (SAG) mills, these losses can be excessive. SAG mills are the technology of choice for reducing primary hard-rock ore to feed size for use in a secondary crusher. The mill’s continuous operation is interrupted when the mill liner — which comprises riser bar plates that protect the original casing and provide required lift required for grinding action — wears out after months of processing abrasive material.


Overall, SAG mills are expensive to operate due to several factors, including high power consumption and the need to frequently replace the liner, which comes with a steep price tag. Stopping mill equipment is not a desirable alternative, so the best way to resolve SAG mill challenges is to use simulation technology for equipment design alterations. Replacing the SAG’s worn-out liner can occur every six to nine months and cost between $30,000 U.S. and $200,000 U.S. per hour, and each repair takes between 40 hours and 120 hours. Obviously, optimizing liner design along with operating processes can result in significant savings.

To achieve SAG mill equipment optimization, you should combine practical experience with pass-or-fail evaluation studies that modulate project resources or operational conditions. We do not recommend that you carry out a project only with in-the-field trial- and-error-testing.

To help predict liner wear and adjust conditions for SAG mill operation, an engineering team can easily assess change-of-process variables, such as speed and fill level, or create a new elevator design that changes the face’s angle and height, exemplified by excellent matching between experimental observations and computational predictions via Rocky DEM software. Discrete element method (DEM) simulations, based on first-principle physics, are extremely useful in these cases, allowing a greater understanding and evaluation of multiple solutions with a high degree of precision.

Mill Showing Surface Wear Modification



Consider a large SAG mill with a diameter of 10.97 m, as shown in Figure 1a. The mill has a high-definition line design with alternating lines of high and low elevators (Figure 1b). The mill processes rocks from 200 mm to 300 mm, using 350 mm steel balls as the grinding medium. The equipment works under these conditions: speed 10 rpm (80 percent of critical speed), volumetric filling 14.5 percent, ratio of balls to rocks 4: 1 in mass. Based on practical experience, mill throughput decreases substantially as a result of loom wear by 60 percent or more. The mining company decided to replace the mill liner based on this 60 percent target. Under typical operating conditions, the mill liner wears out in five months, causing frequent and huge downtime losses.


                                                                                                                          (a)                                                                                                                                      (b)

Figure 1: Mill slice used for DEM calculations


The mining company’s engineering team needed to know if the mill operated under optimum conditions; if not, they would consider modifying the lifter design if needed. Given the operating schedule and costs involved, conventional approaches seemed unfeasible. The team determined that the best solution was to streamline optimization using DEM. Engineers chose Rocky because it provides physically robust, tried-and-tested models with state-of-the-art hardware capabilities and post-processing tools to quantify boundary wear and evaluate milling performance.


Predicting Geometry Wear: Quantification of boundary wear in Rocky is done using the validated Archard’s wear model. The governing equation is A dh = Cmodel dw, in which A is the surface area of a boundary element and dh is the incremental loss in depth when subject to shear work dw. While specification of the wear parameter Cmodel should achieve accelerated wear, care must be taken to avoid very fast and abrupt changes. Often, determination of Cmodel is done after calibration against known experimental data, such as loss in height or mass of the lifters with time. Once a robust wear parameter is obtained, potential process and design changes can be evaluated with a high degree of confidence.

Evaluating Mill Performance Using Energy Spectra:  Rocky provides models to predict performance, but these are computationally very intensive. The software’s energy spectra tool enables reliable predictions of breakage and attrition rates for continuous processes — without running breakage calculations. The tool collects energy for normal and tangential collisions gathered between user-specified (experimental or literature-based) minimum and maximum energy (a) per particle type per time step, and (b) per collision type, which accounts for all combinations of particle–particle and particle–boundary collisions. These calculations are used to predict power draw, process throughput, and milling efficiency. (Details can be found in Rocky literature.)

Want to explore details about Rocky’s custom features — case setup, calibration, post-processing and more? Download the free white paper below.



Related Post


Your email address will not be published. Required fields are marked *