Space Resource Utilisation (SRU) technology will enable further exploration and habitation of space by humankind. For example, oxygen produced \textit{in situ} can be used as the oxidiser in rocket propellant, or for life support systems. The production of oxygen on the Moon can be achieved through the thermo-chemical reduction of the lunar soil, also known as regolith. All reduction techniques require a consistent feedstock from this mix of fine mineral particles to produce oxygen reliably and consistently. The preparation of this feedstock, known as beneficiation, is a critical intermediate stage of the SRU flowsheet, however it has received little research attention relative to the preceding excavation, and the subsequent oxygen production stages.

Triboelectric charging and free-fall separation are attractive technologies for mineral beneficiation as they offers low mass, low power, and low mechanical complexity compared to other approaches. Tribocharging is a process by which particles (conductors, semi-conductors, and insulators) acquire charge through frictional rubbing and subsequent separation. Previous experimental studies have tested different designs of tribocharging apparatuses for terrestrial and space applications, however charge transfer modelling methods have not been employed to optimise design parameters. Furthermore, whilst modelling of the triboelectrification process has been presented in the literature using the discrete element method (DEM), these models often depend on poorly quantified or ill-defined parameters, such as an effective work function for insulating materials. Previous studies have also been restricted to either 2D or 3D domains and have not considered the impact of this on the performance of the models.

To address these knowledge and research gaps, the objectives of this thesis are as follows:
\begin{enumerate}
\item Develop a novel tribocharge modelling approach based on the discrete element method that de-emphasises the poorly-defined quantities found in the high-density limit approach that has been demonstrated previously;
\item Determine the suitability of modelling tribocharging in 2D and 3D;
\item Validate this novel tribocharge modelling method by comparing simulation outputs and experimental data;
\item Present and validate a new DEM-based method for tribocharger design optimisation; and,
\item Evaluate experimentally the impact of an optimised tribocharger design on the performance of an electrostatic separator using standard mineral processing criteria.
\end{enumerate}

A straightforward experimental method to quantify key tribocharging model parameters, namely the charge transfer limit, $\Gamma$, and the charging efficiency, $\kappa_c$, is presented herein. These parameters are then used in both 2D and 3D DEM charge transfer simulations (particle-particle and particle-wall interactions; single and multiple particles and contacts) to evaluate the suitability of faster 2D models. Both the 2D and 3D models were found to perform well against the experimental data for single-contact and single-particle, multi-contact systems, however 2D models failed to produce good agreement with multi-particle, multi-contact systems.

A novel DEM-based approach for tribocharger design optimisation using particle-wall and particle-particle contact areas as proxies for charge transfer is demonstrated. This optimisation method is used to design an optimal tribocharger for use under terrestrial conditions. The novel tribocharge modelling approach was then applied to the optimised charger design. This design was then built and validated experimentally, with good agreement found between the model outputs and experimental data.

The optimised terrestrial design was then employed to study the charging behaviour of pure silica and ilmenite, as well as binary mixtures of silica and ilmenite, and samples of lunar regolith simulant JSC-1. Ilmenite was used because it is a target mineral for oxygen production from the lunar regolith, and silica was used because of its position in the triboelectric series relative to ilmenite. The optimised tribocharger design affected significantly the movement of pure ilmenite in the electrostatic field, despite a negligible change in bulk charge. Experimental results from the binary mixtures indicate that ilmenite recovery is independent of initial ilmenite concentration and can be predicted from the mass distribution of pure ilmenite samples. For JSC-1, the tribocharger was found to increase the density of the material in certain collectors.

This thesis presents new modelling approaches for both tribocharging and tribocharger design optimisation. These techniques will facilitate ultimately the development of beneficiation technologies for SRU. The use of these modelling methods should increase confidence in the performance of tribocharger designs proposed for future SRU missions to the Moon.