Using discrete element modelling (DEM) and breakage experiments to model the comminution action in a tumbling mill

Master Thesis

2008

Permanent link to this Item
Authors
Journal Title
Link to Journal
Journal ISSN
Volume Title
Publisher
Publisher

University of Cape Town

License
Series
Abstract
The Discrete Element Method (DEM) is a powerful modelling tool that characterises the system at the individual particle level. This makes it particularly well suited for simulating tumbling mills whose charge is principally individual particles (steel balls, rocks and fines). The use of DEM to simulate tumbling mills has proliferated since the early 1990s and been successfully employed to predict important milling parameters such as charge motion, power draw, liner wear and impact energy distribution. The ultimate aim of any model of the tumbling mill is to predict the product of the milling process. Current DEM simulations of the tumbling mill however do not simulate the breakage of the particles and as such can not directly predict the product. In order to predict the performance of industrial-scale tumbling mills, laboratory-scale mills are used to experimentally obtain data, which is then scaled up using black box mathematical models. In this thesis a tumbling mill model that utilises the power of DEM to provide the mechanical environment and the energies available for breakage is proposed. The incorporation of DEM eliminates the need to scale up because DEM is able to simulate the actual industrial-scale device. Data from breakage experiments on the ore being treated is also incorporated into the model to determine the breakage functions. Population balance techniques are applied in the mathematical framework of the model to predict the product of the comminution process. In order to test the proposed tumbling mill model, DEM simulations of a 1.695m diameter pilot SAG mill using charge based on actual operation data were performed and analysed. Results from the DEM simulation and Drop Weight Tester breakage experiments were then used in the proposed tumbling mill model to predict the evolution of the product size distribution.
Description

Includes abstract.


Includes bibliographical references (leaves 147-153).

Reference:

Collections