Leaping forward in SAG and AG mill simulation using a mechanistic model framework


L.M. Tavares, R.M. de Carvalho




SAG 2011



Empirical and phenomenological models of SAG and AG mills have dominated the scene for the last few decades, being used nearly every time a mill is scale-up or an attempt is made to improve its performance. However, in spite of their success, these models fail to describe the details of both the grinding environment and the breakage mechanisms. The result is a generally good track record, with a few – but very important – instances when these models fall short in their predictions, which can be very costly. Further, nearly all of these models require detailed calibration to industrial and pilot plant data, being costly in their application, although still with limited predictive capabilities. Recognizing these challenges, a new model framework has been proposed which overcomes the limitations of currently used models by coupling the microscale population balance model, the discrete element model and functions that describe internal classification at the mill grate. This model framework has already been applied to describe breakage in ball mills, with showing very promising results. However, in the case of SAG and AG mills a much greater level of complexity is involved in the model, given the interaction between the mechanical environment that is formed totally or, in part, by the ore itself, and the breakage response in the mill. The paper describes the model framework, demonstrating how it can address appropriately effects such as mill speed, mill filling and size distribution, on the response from the mill. Some of the advantages of the model are related to its intrinsic multi-component and dynamic nature, which allows simulating appropriately mixtures of ores of different competences in a transient manner.


Discrete element method, fracture energy, Mechanistic model, SAG