DSpace Repository

Welcome to the NWU Repository, the open access Institutional Repository of the North-West University (NWU-IR). This is a digital archive that collects, preserves and distributes research material created by members of NWU. The aim of the NWU-IR is to increase the visibility, availability and impact of the research output of the North-West University through Open Access, search engine indexing and harvesting by several initiatives.

Communities in NWU Repository

Select a community to browse its collections.

Now showing 1 - 5 of 20

Recent Submissions

A circular economy model for organic waste management in a South African eco-tourism reserve
(North-West University, 2026) De Wet Gideon; Coetzee R Aruwajoye N.N Roopa M
This dissertation presents the design, contextual adaptation, and validation of a circular foodwaste valorisation model tailored for the Kalahari Reserve, an isolated eco-tourism and conservation site in the semi-arid Northern Cape region of South Africa. The research addresses a dual operational challenge prevalent in remote reserves: the substantial cost associated with importing food and the environmental challenges posed by organic waste disposal. The study employed the Design Science Research methodology, advancing through iterative phases encompassing problem identification, artefact development, contextual adaptation, and empirical validation. The artefact is composed of three interdependent subsystems: livestock feeding, bokashi fermentation, and horticultural production, that collectively form a decentralised circular supply chain, transforming organic waste into food, compost, and animal feed. Pilot studies focused on bokashi processing and livestock feeding, yielding coefficient data crucial for model calibration. Quantitative validation was performed using Statistical Process Control (SPC), regression analysis, scenario testing, Material Flow Analysis, and Circular Value Stream Mapping to evaluate process stability, yield reliability, and system synchronisation. Key performance indicators were deployed to assess circular outcomes. The model demonstrated a waste-diversion rate of 100%, a conversion rate of 100%, a resource-efficiency ratio of 0.99%, and a secondarymaterial substitution rate of 53%, achieving an overall System Performance Index (SPI) of 93%.The findings affirm that the artefact achieves mass-flow closure, operational stability, and high circularity under arid conditions, illustrating that decentralised circular-economy interventions are both viable and potentially transferable within similar conservation contexts. This study makes a significant contribution to the domain of Industrial Engineering by integrating Lean and Circular principles through the combined application of Material Flow Analysis (MFA), Circular Value Stream Mapping (CVSM), and Key Performance Indicator (KPI) frameworks. Additionally, it establishes a methodological framework for engineering-based circular-bioeconomy models in the Global South, thereby facilitating localised sustainability transitions in ecologically sensitive and resource-constrained settings.
Effect of heat treatment on residual stress in Selective Laser Melted CoCrMo
(North-West University, 2026) Du Plessis Juan; Bayode A Kloppers C.P
Selective Laser Melting (SLM) of Cobalt Chrome Molybdenum (CoCrMo) alloyspresents major advantages for producing high-performance biomedical and aerospace components. However, the process induces significant residual stresses due to large thermal gradient development during the layer-by-layer manufacturing method. Residual stresses are important to mitigate as they can compromise dimensional accuracy, accelerate crack formation, and decrease component reliability. This study aims to evaluate the effect of stress-relief heat treatments on the reduction of residual stresses in SLM-fabricated CoCrMo alloys. Cantilever and cubic specimens were manufactured to evaluate the effect heat treatment has on residual stresses through the cantilever deflections (inherent strain method), XRD and microstructural analysis. Numerical simulations were implemented to predict the effects of 700°C,770°C, and 1065°C heat treatments. Experimental results showed substantial reductions in cantilever deflection, with stress relief of 54% at 700°C and 82% at 770°C. Heat treatment at 1065°C led to stress reversal, indicating significant redistribution of residual stress as inferred from cantilever deflection. Simulations highlighted the trend of decreasing residual stress with an increase in temperature, although they underestimated the magnitude of experimental stress relief, particularly at higher temperature, suggesting unmodelled phase-transformation effects. Microstructural analysis revealed microstructure development from elongated columnar grains in the as-built condition to partial recrystallised structures at 770°C and mixed elongated and equiaxed microstructures at 1065°C. XRD confirmed the presence of ɣ-phase and ε-phase across all conditions, with the ɣ-phase being the most dominant. The XRD also showed that heat treatment caused shifts and broadening in peaks, intensity changes, and the appearance of an additional peak when heat treatments were applied.The combined simulations and experiments show that heat treatment is an effective approach for residual stress reduction in SLM CoCrMo alloys. The study also found that calibrated finite element simulations can serve as a valuable predictive tool for optimising heat treatment strategies and minimising trial-and-error experiments in SLM. It should be noted that residual stresses were inferred indirectly from cantilever deflection measurements using the inherent strain method, and that the investigation was limited to selected stress-relief temperatures for a single SLM CoCrMo material system. Within these constraints, a heat temperature of 770°C provides a desirable balance between stress relief and dimensional stability.
Real-time rail surface defect detection: a neuromorphic approach
(North-West University, 2026) Esterhuizen, Rensche; Van Vuuren, P.A
This study explores the implementation of a neuromorphic-based approach for real-time rail surface defect detection, addressing the growing demand for accurate and efficient maintenance solutions in rail systems. Leveraging a dataset of 1514 rail surface images collected under real-world operating conditions and annotated with binary labels ("defect" and "no defect"), the research applies a comprehensive data augmentation pipeline - including rotation, flipping, zooming, and contrast adjustments - to expand the dataset to 248374 images. This augmented dataset ensures class balance and greater variability, enhancing model generalisation and preventing overfitting. A consistent data split (70% training, 15% validation, 15% testing) is employed to maintain fairness across evaluations. The study presents a comparative evaluation of spiking neural networks (SNNs) and convolutional neural networks (CNNs) in terms of classification performance, energy efficiency, and real-time feasibility across three hardware platforms: a 12-core AMD Ryzen 9 CPU (3.8 GHz), a NVIDIA RTX3060 GPU (12 GB VRAM), and the BrainChip AKD100 neuromorphic processor. The analysis further investigates the influence of hand-crafted features - including Grey-Level Co-occurrence Matrix (GLCM), Local Binary Pattern (LBP), and Sobel and Canny edge detection - on model performance and computational efficiency across different hardware configurations. Experimental results show that CNNs consistently outperform SNNs in precision and recall, particularly when implemented on GPUs, where the CNN model achieves a precision of 94.82%, recall of 94.88%, and F1-score of 94.49%. The Akida neuromorphic processor using a CNN model, with its 8 neural cores and sub-1W power consumption, maintains a high precision of 94.74% while achieving latency below 5 ms and throughput exceeding 250 frames per second - demonstrating strong balance of accuracy, speed, and energy efficiency. In contrast, SNNs, though more energyefficient, show diminished accuracy, with the CPU-based SNN model achieving an F1-score of only 83.75%. The research also outlines optimal hardware configurations for deployment, considering factors such as sensor quality, data acquisition methods, power supply, and cooling requirements. The findings indicate that combining CNNs with the Akida processor offers an ideal solution for real-time rail defect detection, balancing classification performance with resource efficiency. The final chapter synthesises these insights to offer practical guidance for system deployment and highlights future directions to advance neuromorphic-based approaches in railway maintenance.
Investigating Behavioural Intelligibility In Reinforcement Learning
(North-West University, 2026) Fouché , Janu; Van Vuuren, PA;
In this dissertation, an investigation is launched into the fundamentals of reinforcement learning, aiming to answer the question: How can first-principles implementations of RL agents enhance our understanding of their behaviour and improve intelligibility in controlled environments? This investigation is addressed by outlining the overarching mathematical concepts, the theoretical agent framework, the structure of the controlled environment, and the intelligibility methods employed. Subsequently, four selected agents-PPO, Q-Learning, REINFORCE, and DDQN-are implemented using transparent, first-principles approaches. Each implementation is described in detail. The agents are then deployed within the environment, trained, and experimentally evaluated with respect to key features of their decision-making and learning behaviour. These experiments demonstrate that, through these investigative techniques, it is possible to observe and confirm theoretical aspects of the agents in a practical and transparent manner, thereby contributing to answering the research question. In conclusion, by uniting the mathematical concepts with the theoretical understanding of the agents and presenting them in a transparent manner, one can train agents in a controlled environment and investigate them using intelligibility techniques. This approach provides significantly improved insight into the behaviour of these agents and lays the groundwork for future tests to be conducted on these fundamental implementations.
Simulation-based verification of stope heat loads in deep-level mines
(North-West University, 2026) Galamadien, Phillip; Vosloo, J.C
As gold mines continually increase their mining depths and distances from shafts, the difficulty of providing effective cooling also increases. This increases the need for effective planning, which requires accurate knowledge of the contributions of different areas to the mine's total heat load, especially the production areas (stopes). Few heat load quantification methods isolate the heat of production areas. A simulation-based approach can be used, but such a method's accuracy must be validated. The systematic literature review process revealed a gap in the literature which is if simulations as a predictive tool can be used to determine stope heat loads and whether it can be used to compare with heat loads determined by industry accepted methods. Therefore, this study focuses on validating whether the simulation method can be used to compare calculations done on stope heat loads. In this study, the heat loads of three stopes of an operational mine were quantified by considering their energy balances. Focus was placed on the main heat sinks - ventilation air and service water. Thermo hydraulic simulations were set up to duplicate the real-world setting as closely as possible. Certain simulation parameters were fixed based on real-world knowledge, while unknown parameters were adjusted to match real-world measurements as best as possible. The simulated heat loads were compared to the energy balance method. The Van der Walt's method was determined and also compared to the simulation method. The entire process was repeated for two case studies from literature.For the in-field measured stopes, the total stope heat generated was approximately 700 kW at a depth of approximately 3.6 km. The heat removed by the ventilation air and service water was 254 kW and 441 kW, respectively. Overall (including both the in-field measured stopes and case studies), the mean absolute errors produced by the simulation and Van der Walt's method were 10.9% and 29.6%, respectively. The simulation method was validated by applying the method to external case studies. Overall, it was demonstrated that the simulation method can be used to make predictions about expected stope heats and can compare with industry methods.