The development and application of a simultaneous two-way coupled, discrete element method and smoothed particle hydrodynamic model for spatial scale-up dynamics of laboratory vertical stirred media detritors

Ssebunnya, Divine

The development and application of a simultaneous two-way coupled, discrete element method and smoothed particle hydrodynamic model for spatial scale-up dynamics of laboratory vertical stirred media detritors

Thesis / Dissertation

2025

Publisher

University of Cape Town

Department

Department of Chemical Engineering

Faculty

Faculty of Engineering and the Built Environment

Abstract

Stirred grinding processes are becoming more common in industries spanning from ceramics and pigment production to mineral processing. Within the minerals processing industry, stirred milling to produce fine grinds is required to liberate minerals within the more complex ore types. Processing of complex low-grade ore is on the rise due to the depletion of easy to process, high grade ore bodies. Stirred mills are becoming the dominant mill type for achieving ultrafine grinds within the minerals industry due to their energy efficiency. These mills agitate grinding media, in a chamber with a concentric impeller to facilitate particle size reduction. The use of laboratory scale systems in the preliminary assessment of the grinding capabilities of larger industrial scale mills is common practice within the industry. For other mill types such as ball mills, methods such as the Bond test are used on laboratory scale mill systems to assess the scale-up effects of the system from laboratory to pilot and industrial scales. However, apart from the signature plot approach, there is no established methodology that incorporates the use of laboratory scale stirred mills in the size reduction scale-up process to larger systems available in the literature. This is because the spatial flow dynamics within stirred mill systems are complex, due to dependencies on multiple parameters spanning from the mill dimensions and design to the grinding media and fluid characteristics. Therefore, the scale-up dynamics, based on media and fluid dynamics within stirred mill systems are not well understood, within the literature. This study serves to address this through an assessment of spatial scale-up effects on a system of laboratory scale stirred media detritors (SMD). The SMD is a common vertical stirred mill type used to achieve size reduction for fine and ultrafine wet grinding applications. Three scales of laboratory mill systems are assessed as three separate cylindrical vessels using the same concentric impeller shaft and configuration of pins, which vary in length proportionate to the vessel diameter. The flow and energy responses of the mill system are simulated, analysed, and compared across scales. In the advent of the discrete element method (DEM) and smoothed particle hydrodynamics (SPH), the computational simulation of free surface, particulate, and fluid flow in wet fine grinding mill systems has been executed, with DEM accounting for the grinding media, impeller and mill shell, and SPH accounting for the slurry, modelled as a quasi-compressible fluid. This study presents a two-way coupled DEM-SPH model that computes the DEM and SPH portions of the mill system simultaneously and resolves their interaction forces through a two-way coupling interface. The simultaneous nature of the model provides a significant reduction in the simulation runtime and the feasible SPH resolution. Additionally, a methodology to model media shapes as super quadric surfaces is developed and implemented. The DEM-SPH model is assessed through analysis of the effects of key variables to gauge its performance prediction capabilities on the largest mill size before its application in the scale-up study. The key variables assessed are media size and density, slurry density and viscosity and media fill. The effects of mill speed and media shape are considered in the scale-up study, with simulations conducted at each of the three mill sizes. The scale-up study considered the spatial distributions within the flow and energy environments and assesses the use of single value quantities, such as dimensionless numbers and intensive variables, as metrics for mill performance. These quantities showed sensitivity to different attributes of mill performance.