School of Engineering (SEng.) Collections
Permanent URI for this collection
Browse
Recent Submissions
1 - 5 of 335
-
ItemReliability enhancement of Non-destructive testing methods for IN-SITU concrete compressive strength using Convolutional Neural Networks with destructive testing data.(Makerere University, 2025-12-17)Reliable evaluation of in-situ compressive strength of concrete is vital for ensuring safety and longevity of ageing infrastructure particularly bridges that recently are subjected to increasing traffic loads and deterioration due to climatic conditions. While traditional core extraction methods have reported high accuracy, these are intrusive, time-consuming and costly, thus limiting their practical application in structural assessment of concrete. In response to the need for less intrusive, quicker and cost-effective alternatives, this study investigated the use of the two most common Non-Destructive Testing techniques (NDT)—Schmidt Rebound Hammer (REB) and Ultrasonic Pulse Velocity (UPV)—coupled with Convolutional Neural Networks (CNN) to improve predictive accuracy. The study developed and trained three CNN model configurations using MATLAB® R2025a with data collected from five existing concrete bridge structures. The architecture consisted of a Conv1D with an input layer for two-channel sequential data, a convolutional layer with five filters, BatchNorm, ReLU activations, two fully connected layers with dropout regularization and a final dense output neuron, optimized for regression tasks. Each test method (REB, UPV and Core) generated 30 measurements which were subsequently cleaned and organized into triplicate datasets. REB-only, UPV-only and Combined REB–UPV were evaluated against core test results as ground truth. Model results revealed that the REB-only model performed reasonably well (R² = 0.75, RMSE = 1.78 MPa, MAPE = 10.8%), UPV-only model exhibited lower predictive capacity (R² = 0.31, RMSE = 2.09 MPa, MAPE = 29.6%) whereas the Combined REB–UPV model outperformed both, achieving an R² of 0.90, RMSE of 1.39 MPa and MAPE of 7.8% during training. However, signs of overfitting were observed during testing, primarily attributed to the limited dataset size. It was demonstrated that combining NDT data with CNN-based deep learning networks can significantly enhance compressive strength prediction over single-method approaches. This study offers a novel application of CNNs—typically used in image recognition—for numeric prediction-based concrete assessment using NDT and Destructive Testing (DT) data. It presents a promising, scalable yet non-invasive practical tool for structural health monitoring.
-
ItemInfluence of Geometric Curvature on the Out-Of-lane flexural capacity of Unreinforced Masonry Walls.(Makerere University, 2025-12-08)Unreinforced masonry (URM) remains a prevalent construction system in Uganda and numerous regions worldwide, yet its susceptibility to out-of-plane (OOP) loading continues to precipitate some of the most abrupt and severe structural failures. Existing analytical frameworks and design provisions are predominantly based on straight, planar walls, leaving curved wall configurations, ubiquitous in both architectural practice and vernacular construction, insufficiently characterized within prevailing theory. This study interrogates the combined influence of geometrical configuration and material properties on the flexural resistance of URM walls subjected to OOP loading. Laboratory characterization of locally manufactured clay bricks and selected mortar mixes revealed pronounced variability in stiffness relative to compressive strength, reflecting heterogeneous microstructural features and production inconsistencies. While augmented mortar strength enhanced compressive performance, its effect on unit-mortar interface friction remained marginal, highlighting persistent limitations in bond behaviour. These empirically derived parameters were incorporated into a finite element model developed in ABAQUS and validated against published experimental benchmarks and mathematical metrics, successfully reproducing crack initiation, stiffness degradation, and ultimate failure mechanisms. Parametric analyses revealed that straight walls exhibit predominantly bending-dominated responses, wherein improved mortar grades elevate load capacity without altering the governing failure mode. Conversely, curved walls manifested fundamentally distinct structural behaviour: increasing projection distance markedly improved OOP strength and initial stiffness through the development of compressive thrust lines and an enlarged effective lever armphenomena not adequately captured by classical plate theory. Geometry-induced arching conferred performance gains even for lower-grade mortars, whereas higher-grade mortars accentuated this effect, albeit with reduced ultimate displacement. Using projection height as a practical geometric descriptor, a simplified expression is proposed to estimate the flexural capacity of curved URM walls, providing an accessible alternative to full numerical modelling. Collectively, the findings underscore wall geometry as a dominant, geometry-driven resilience mechanism operating synergistically with material enhancements, challenging the conventional reliance on planar bending theory for predicting URM OOP behaviour.
-
ItemAssessing the impact of land use/land cover and climate change on environmental flow requirement in river systems with multiple uses(Makerere University, 2025-10)Water resources systems like the R. Namatala, shaped by the hydrological cycle, involve complex processes impacted by both human activities and climate change. The pressures of commercialization, industrialization, and increasing water demands exert immense stress on water resources with 65% of the world's rivers now at risk. Diminished river flows, driven by more water withdrawals, land use changes, and climate change, are compromising environmental flow requirements and threatening ecosystem health. Challenges such as inadequate water management practices and limited data on river systems hinder the maintenance of ecological flows, leading to unsustainable water resources exploitation. Consequently, reliance on flow estimates that overlook essential ecological factors worsens the threat to freshwater ecosystems. The purpose of this study was to assess the impact of land use change and climate change on environmental flow (e-flow) requirements. Specifically, the study; i) determined the current environmental flow requirements, ii) examined land use change trends and, iii) determined the impact of land use / land cover and climate change trends on the environmental flow requirements of R. Namatala. Three approaches were that included; (1) Hydrological method, (2) Hydraulic method and (3) Holistic approach were selected to estimate the current E-flow requirements. Land use trends between 1995 to 2023 were determined using image classification tools in google earth engine and trends for 2023 to 2040 were projected using Terrset Land Change Modeler. Hydrologiska Byråns Vattenbalansavdelning (HBV) model was used to determine impact of Land use/land cover and climate change on environmental flow under RCP 2.6 and RCP 8.5. River Namatala had a mean annual flow of 2.65 0.08 m3/s obtained over a period of 73 years. The river normally experiences low flows in the months of December to March and high flows in the months of April to June. Average annual Low flows of 0.65 0.45 m3/hr and max average annual flows of 18.41 8.79 m3/hr. The current E-flow requirement for R. Namatala was determined based on the flows over a period of 73 years, consideration of the different water demands. The estimated current e-flow requirement at the outlet of R. Namatala catchment was 1.072 , 1.036 and 1.103 m3/s as determined from the hydrological, hydraulic and holistic methods, respectively. The results from the historical land use/ land cover trends estimated over a period of 26 years (1995 – 2020) indicated an incremental dominance by cropland (0.471% yr-1), followed by grassland (0.158% yr-1) at the expense of forestland (-0.466% yr-1) and wetlands (-0.176% yr-1). The projected LULC trends estimated over a period of 17 years (2023 – 2040) indicated that forestland would have the predominant increment (0.263% yr-1) followed by cropland (0.071% yr-1) while grasslands would decrease (-0.343% yr-1). Utilizing projected land use data and climate change most likely scenario - RCP 2.6, the projected e-flow requirement was estimated to be 0.880, 1.082 and 1.1591 m3/s for the hydrological, hydraulic and holistic methods, respectively. On the other hand, for the most unlikely scenario - RCP 8.5, e-flows were estimated as 0.910, 1.076 and 1.153 m3/s for the hydrological, hydraulic and holistic methods, respectively. The study concluded that land use together with climate change will have an incremental impact to the future e-flow requirement for R. Namatala of between 4.53% and 5.08% as predicted by the most unlikely and most likely scenarios, respectively. Further to this, the ecosystem is foreseen to have water quantity challenges due to abstractions with months of December to February cited as critical months with reduced flows. Government interventions, including wetland gazettement and forest protection, contributed to increased forest cover but there was a reduction in cropland, a factor that could in itself contribute to food scarcity in the area. Climate projections under RCP 2.6 and RCP 8.5 indicated increased precipitation and seasonal flow shifts, with minimal variation in environmental flow requirements. The study assumed socio-economic dependence of the communities on the river based on the communities’ proximity and accessibility to the river. The e-flow determination also depended on measured flows and no groundwater component. It is thus recommended that further socio-economic and groundwater assessments, improved water management, enhanced policy enforcement, and continued climate-focused research are done to ensure sustainable catchment management. With there being no universally accepted e-flow determination method, this study notes a minimum e-flow requirement of 0.880 m3/hr as determined using the hydrologic method but recommends a value of 1.153 m3/hr as determined using the holistic approach for sustainable exploitation of the water resource since the holistic approach considers a number of ecosystem functions that are not considered by the other methods.
-
ItemDevelopment of mobile carbon capture system for motorcycles’ emissions(Makerere University, 2025)This dissertation investigated the effectiveness of mobile carbon capture technology for reduction of emissions from road transport specifically targeting motorcycles. The study focused on the integration of soda lime and activated carbon as sorbents within a custom-fabricated muffler, which was then attached to the motorcycle's exhaust system. The primary objective was to determine the reduction of carbon dioxide (CO2) and carbon monoxide (CO) emissions from motorcycle exhaust gases before their release into the atmosphere. The four custom-fabricated mufflers were designed to ensure optimal flow of exhaust gases through the sorbent materials. Considerations included the density and surface area of the activated carbon and soda lime, the duration of gas exposure to the sorbents, and the durability of the materials under the high temperatures and pressures found in motorcycle exhaust systems. The design also accounted for minimal interference with the motorcycle's overall performance, ensuring the exhaust system remained functional while maximizing emission reduction. A series of experimental setups were designed to simulate real-world operating conditions of motorcycles. Fabricated muffler capture devices integrated with activated carbon and soda lime, were affixed to the exhaust systems of motorcycles. The effectiveness of these sorbents in reducing CO and CO₂ emissions was assessed by measuring the concentration of these gases before and after exhaust treatment under various operating conditions, such as idling, increased flow rate, and cruising. The muffler with activated carbon showed a significant capacity for adsorbing CO, with an average reduction of 60% across different operating conditions for the focal motorcycles (newer motorcycle). The muffler with soda lime, on the other hand, demonstrated a robust capacity for absorbing CO₂, achieving an average reduction of 78%. The combined use of these sorbents resulted in an overall reduction in carbon emissions by approximately 65% for CO and 80% for CO₂. The findings suggested that mobile carbon capture (MCC) using activated carbon and soda lime could be a feasible strategy for decarbonizing road transport in urban areas, particularly where motorcycles dominate the transportation landscape. However, the long-term effectiveness of these sorbents, potential impacts on motorcycle performance, and the economic viability of widespread implementation require further investigation. This study lays the groundwork for future research on optimizing MCC systems for motorcycles
-
ItemAssessment of battery energy storage systems for reliable power supply on Uganda’s electricity grid(Makerere University, 2025)Uganda’s grid still faces significant reliability challenges, with frequent prolonged outages at times lasting up to 15 hours, far beyond the 4-hour target. While several remedies have been deployed to combat reliability like solar energy integration and infrastructure upgrades, these interventions also face their own challenges. The intermittency limits solar energy’s effectiveness and infrastructure upgrades suffer from vandalism plus a constrained government budget. As a solution, this study explored deploying a utility-scale battery energy storage system to improve network reliability by smoothing power supply during outages. Supply-demand curves of selected feeders were analyzed to identify the critical outage times for battery discharge, focusing on the distribution network. The study identified vital feeders such as Mutundwe-Masaka Central 33 kV, Masaka Central- Mitala Maria 1 33 kV, among others where the battery energy storage system deployment would yield the most significant impact towards network reliability improvement. Using data of 3 years for load flow analysis, this thorough process helped determine the optimal size and location of the battery energy storage system using a particle swarm optimization algorithm. The algorithm focused on reducing the cost of energy not served and improving key reliability indices like system average interruption duration index, system average interruption frequency index, loss of load probability and energy not served. The optimal placement was found at buses with the highest loss reduction along critical feeders, which resulted in reduced reliability interruption indices and a significant reduction in the cost of energy not served. The optimized reliability indices from MATLAB were validated against those from DIgSILENT Power Factory using the IEEE 14-bus system as a bench mark, showing acceptable error margins. < 10%. A cost benefit analysis comparing the deployment of a battery energy storage system and a feeder upgrade option showed that while both strategies were viable; the financial analysis portrayed the battery energy storage system having a slightly shorter payback period of 7.9 years, IRR-7%, ROI-2.73% compared to 8.06 years, IRR-5%, ROI-2.5% for the feeder upgrade option. Hence the battery energy storage system offered a more profitable investment which would be recovered in a shorter period, making it a more cost-effective solution in the long run. Consequently, the economic analysis which shows the overall socio-economic benefit of the society as a whole is expected due to increased production through improved power reliability.