Effective models of vibratory screens which can capture the true response characteristics are crucial in the understanding of faults and failures which occur in vibratory screens. However, the current available models are usually simplified and have limited validation to that of a physical screen. Much research has been conducted to optimise the screening efficiency of screens. The optimisation includes screen geometry, material processing of the screen and the dynamic response of the screen. These investigations have not been furthered to investigate the effects of different faults on the dynamic response of a vibratory screen. To model a vibratory screen which can replicate the dynamics of a physical vibratory screen it is important to create a model with enough complexity to capture the dynamics of the screen. The model of the screen was validated using both modal analysis and the transient response of the screen. The modal analysis was used to ensure that the physical characteristics of the model are consistent with that of the physical screen. Once this was completed, the second validation aimed to investigate if the model of the screen could capture transient faults which are measured experimentally. It was found that it was not possible to conclusively determine if the finite element methods model could Finally, an intelligent method was used to distinguishing between different faults and classifying them accordingly. The intelligent method was also trained using the FEM data and then used to classify the physical screen data. / Dissertation (MEng)--University of Pretoria, 2020. / Mechanical and Aeronautical Engineering / MEng / Unrestricted