Phil Owen

Bio: Phil Owen is an academic researcher from Commonwealth Scientific and Industrial Research Organisation. The author has contributed to research in topics: Mill & Discrete element method. The author has an hindex of 10, co-authored 14 publications receiving 346 citations.

Quasi-static fall of planar granular columns: comparison of 2D and 3D discrete element modelling with laboratory experiments

Abstract: We investigate the fall of planar granular columns in a quasi-static regime using a combination of laboratory experiments and simulations using the discrete element method (DEM). Columns made of alternately coloured sand layers are initially laid out in a box and then released when a retaining wall is set in slow motion with constant speed. The flow of the sand involves failure of the granular material along multiple slip planes at different stages of the deformation, leading to complex flow patterns. DEM simulations, in 2D and 3D, are able to qualitatively and quantitatively reproduce a range of essential elements of this flow, with particle shape being an important contributor to the mobilisation of specific slip planes. We also examine the sensitivity of the DEM predictions to changes in the operating, material and simulation parameters. We find that DEM predictions are invariant to changes of moving wall speed, spring stiffness, particle–particle friction and particle–wall friction. Conversely, they a...

Prediction of mill liner shape evolution and changing operational performance during the liner life cycle: Case study of a Hicom mill

Abstract: The prediction of wear and performance change over the life cycle of the consumable liners used to protect grinding mills is an important element of their optimization. Poor wear behaviour can lead to higher replacement costs and production loss during liner replacement. Prediction of wear behaviour using discrete element method (DEM) can provide valuable information about the impact of the design on the life cycle of the liner. Here, the DEM method is extended to predict the evolution of the shape of the liner throughout the life cycle. Using a Hicom 110 nutating mill as a case study, we demonstrate a process for constructing a wear model that is shown to be able to quantitatively predict the spatially varying wear rates over the surface of the liner. The functional dependence of wear rate on micro-mechanical flow quantities is not currently understood; hence, two model variants for each of the normal impact damage and the shear abrasion damage are considered. This model is validated by comparison with the actual erosion depths of a real liner. It is then used in a multi-step DEM simulation to predict the liner evolution and the operational changes in the mill performance over the full life cycle of the liner. For the Hicom mill, it is demonstrated that the wear is entirely dominated by abrasion and that the best measure of this was the shear energy dissipation of wall contacts. It was also to identify that the floor of the mill was constructed of a more abrasion resistance material. The life cycle analysis showed a 20% decline in power draw as the increasing grinding chamber volume and eroding lifters lead to less efficient operation of the mill.

Discrete Element Method (DEM) for Industrial Applications: Comments on Calibration and Validation for the Modelling of Cylindrical Pellets

Abstract: As a consequence of increasing computer power and more readily useable commercial and open source codes, Discrete Element Method (DEM) is becoming widely used across a range of applications to simulate increasingly complex processes. This is exacerbating the challenge of setting up simulations for industrial applications. The literature on input parameter selection is divided. A number of papers report methods for their direct measurement. Others, by contrast, propose a “calibration” approach where the particle properties are derived as adjustable parameters by quantitative comparison of experimental and simulation results. This paper reports on the calibration and validation of the input parameters for the specific case of cylindrical tablets represented by conjoined spheres. The initial steps are to not only assign and optimise the DEM input parameters but also optimise the shape representation; what degree of linearity of the edges and angularity of the corners are required to accurately reflect the cylindrical shape. The model was used to simulate two configurations of a rotating drum: an “attrition tester” with a single longitudinal baffle and an un-baffled drum. The results were compared qualitatively and quantitatively with experimental data. While the qualitative comparison was good in most cases, detailed quantitative comparison fared less well, with some significant errors. This study highlights some of the key issues for a wider-spread of industrial applications for DEM.

The packing properties of superellipsoids

Abstract: We investigate the properties of packings of frictionless non-spherical particles utilizing a dynamic particle expansion technique. We employ superquadric particles (superellipsoids), which allows us to explore how a broad range of particle shapes affect both the macroscopic and the local configurational properties of the system. We smoothly transition from spherical particles possessing only translational degrees of freedom to large aspect ratio non-spherical grains where rotational degrees of freedom are highly important. We demonstrate that the degree of anisotropy and local surface curvature of the particles have a profound effect on their packing properties, determining whether a random or an ordered packing is readily formed.