Doctoral Degrees (Mechanical and Mechatronic Engineering)

Permanent URI for this collection

https://scholar2.stb.sun.ac.za/handle/10019.1/780

Browse

Recent Submissions

Now showing 1 - 5 of 95
  • Thumbnail Image
    Item
    A human cyber-physical system to study the motion sickness of seafarers.
    (Stellenbosch : Stellenbosch University, 2023-11) Taylor, NC; Kruger, K; Bekker, A; Stellenbosch University. Faculty of Engineering. Dept. of Mechanical and Mechatronic Engineering.
    ENGLISH ABSTRACT: Maritime 4.0 offers technical opportunities to digitally enable ships. Systems are equipped with virtual counterparts, forming cyber-physical systems, to manage operations. However, the human element remains constrained to the physical layer. This dissertation proposes the Mariner 4.0 concept that encourages equipping seafarers with virtual counterparts. Mariner 4.0 is a contribution formulated to serve as a branch of digitalisation that addresses opportunities and challenges associated with human factors in Maritime 4.0, such as a lack in accessibility of human-centric data during operation. Human cyber-physical systems offer a promising means for human-system integration and human digital representation amidst technically-centred developments. This dissertation contributes an architecture for a human cyber-physical system that defines core elements – a physical layer, where seafarers are present, a cyber layer, which contains virtual counterparts for seafarers, and communication between the layers. A human cyber-physical system for seafarers is implemented and deployed, which is the first objective of this dissertation. The trialled human cyber-physical system is a contribution that facilitated human-centric data acquisition and processing for seafarers on a ship over the course of a 20-day long voyage. The motion sickness of seafarers is monitored subjectively and objectively in near real time, a novel feat in shipping, in a case study with 63 participants on board. Seafarers recorded their subjective experiences of motion sickness through two methods. The first, traditional method used paper-based questionnaires, while the second, novel method used the human cyber-physical system. The second objective of this dissertation is to use the human cyber-physical system as a mechanism for advancing the study of motion sickness in shipping. The human cyber-physical system enables personalised data analysis as well as conventional aggregation of results. As such, novel understanding of motion sickness and the study thereof in shipping is uncovered. Measures of motion exposure are determined objectively according to recommended procedures in ISO 2631-1 (1997) and are integrated with the percentage of motion sick individuals on board, forming motion sickness criteria. The human cyber-physical system provides a platform for inspecting the evolution of criteria development during ship operation. The results contribute diagnostic thresholds that could be used beyond the operational stage to assess the levels of sickness that individuals or a cohort may present if exposed to certain measures of ship motion. The motion sickness criteria accommodates factors, such as exposure duration, for tailoring diagnostic thresholds to voyage missions – potentially applicable for voyage planning and ship design. The human cyber-physical system provides a customised platform for addressing challenges and opportunities associated with human factors in Maritime 4.0. Moreover, the human cyber-physical system extended the knowledge basis of motion sickness on ships innovatively. The human cyber-physical system served as a human-centric platform for seafarers that empirically revealed the importance of performing data acquisition and analysis at an individual level in addition to an aggregate level.
  • Thumbnail Image
    Item
    Flow and heat transfer in packed beds of rock.
    (Stellenbosch : Stellenbosch University, 2023-11) Hassan, EBE; Hoffmann, JE; Stellenbosch University. Faculty of Engineering. Dept. of Mechanical and Mechatronic Engineering.
    ENGLISH ABSTRACT: Utilizing thermal energy storage systems in a solarized Brayton cycle increases the capacity factor of the power plant by using waste heat. The waste heat can be used to power a Rankine cycle to produce additional energy after sunset or during periods of high demand. Packed beds of crushed rock have been proposed as a promising storage solution since it is readily available, inexpensive, and able to withstand high temperatures. Dolerite is one of the rock types selected as the optimal choice for storing high temperatures. Nevertheless, predicting the pressure drop over a packed bed is a crucial parameter for estimating the pumping power of the system. There are various parameters that influence the pressure drop through the bed, such as the shapes and sizes of the crushed rocks, which affect the packing density and particle orientation. In this study, the investigation of the pressure drop through the packed bed of crushed rock depending on the flow direction was conducted. The crushed rock was represented by an ellipsoidal shape with the same volume and aspect ratio as the average of randomly collected crushed rock samples. This is because the aspect ratio is considered one of the particle characteristics that enables one to capture the flow directional effect. Simulation models were developed to assist in deriving a pressure drop correlation as well as the effect of particle orientation on pressure drop. For simulation, a discrete element model (DEM) was used to generate the particles and computational fluid dynamics (CFD) to simulate the flow at pore scale over the particles. Firstly, a packed bed of ellipsoidal particles was developed using a DEM-CFD method. Afterward, an experimental model was developed to validate the DEM-CFD model. Following the validation, the DEM-CFD model was compared with a crushed rock packed bed to examine how well the ellipsoid particles represent the crushed rocks. The findings reveal that the model successfully captured the flow direction effect across the crushed rock bed. However, it underestimated the pressure drop through the crushed rock by 5 % in horizontal flow and overestimated it by 20 % in vertical flow. The wall has a direct effect on the particles' alignment, where at the bottom of the container the particles are aligned with their flat faces. The wall effect is high for the particles close to it; however, it does not extend deep into the bed. Additionally, it depends on the bed-to-particle diameter ratio. For a large rock bed where it is a free packing bed, the wall has an insignificant effect on the particles' alignment. Therefore, the wall effect was then eliminated from the DEM-CFD model, and a correlation was proposed to predict the pressure drop through a crushed rock packed bed using the porous media approach. The proposed correlation was used to predict the pressure drop across a packed bed of 10 MWth. The findings were compared with the isotropic Ergun model. After investigation of the outcomes, it was found that the proposed correlation captured the directional effect. Also, it was predicted that the pressure drop across a porous bed would be about 22 % lower than that predicted by the Ergun equation in the vertical direction.
  • Thumbnail Image
    Item
    Improving counting performance of densely flocked sheep in aerial imagery.
    (Stellenbosch : Stellenbosch University, 2023-11) Biggs, DR; Schreve, K; Theart, RP; Stellenbosch University. Faculty of Engineering. Dept. of Mechanical and Mechatronic Engineering.
    ENGLISH ABSTRACT: The use of machine learning and computer vision to count sheep in aerial images taken by an unmanned aerial vehicle (UAV) is explored. The aim is to develop sheep counting approaches which address the challenges related to high object densities and low object-to-image-pixel ratios encountered when using detection-based architectures and aerial images. A comprehensive review of state-of-the-art object detection and instance segmentation techniques and existing object counting approaches is presented. A novel dataset is generated and presented which contains aerial images and videos of sheep grazing in pastures, captured at an altitude of 30m. These images and videos contain scenes of non-uniform object (sheep) distributions and high object densities. A total of 13 different object detection and instance segmentation models are considered, and five of these models undergo hyperparameter optimisation and k-fold cross-validation. The five models are compared based on their crossvalidation mean average precision (mAP) scores on the novel dataset. The best-performing object detection model and instance segmentation model are shown to be Cascade R-CNN and hybrid task cascade (HTC), respectively. The non-uniform distribution of objects, high object densities and low objectto-image-pixel ratios in aerial images present challenges seen in both this dissertation and the literature. It was found that high object densities and low object-to-image-pixel ratios adversely affect counting performance. Two novel approaches are proposed that minimise these effects. The first approach, local density threshold shifting (LDTS), focuses on the challenges posed by high object densities. LDTS shifts the classification probability of each detection based on the density of that detection. This approach achieves an mean absolute error (MAE) and mean absolute percentage error (MAPE) of 26.50 sheep and 4.22%, respectively, on an unseen test dataset and reduces the overall counting error by 78.51% compared to the baseline counting approach. This approach has been published in the International Journal of Remote Sensing. The second approach, sub-window inference, focuses on increasing the object-to-image-pixel ratios of small objects taken in aerial images and videos. This approach utilises a novel cropping technique along with additional data augmentation during training to enhance performance. Sub-window inference achieves an MAE and MAPE of 3.21 sheep and 1.27%, respectively, on an unseen test video dataset, which is a 97.40% reduction in error compared to the baseline counting approach. Sub-window inference has been submitted to the journal Computers and Electronics in Agriculture.
  • Thumbnail Image
    Item
    The influence of eye model complexity and parameter variations on the predictive power of simulated eye-tracking data
    (Stellenbosch : Stellenbosch University, 2023-11) Fischer, JD; Van den Heever, D; Van der Merwe, J; Stellenbosch University. Faculty of Engineering. Dept. of Mechanical and Mechatronic Engineering.
    ENGLISH ABSTRACT: Gesimuleerde oogvolgdata vergemaklik die vinnige, herhaalde, en onafhanklike evaluasie van oogvolgmetodes sonder dat daar 'n hulpbron-intensiewe gebruikersstudie vereis word. Gesimuleerde data is veral voordelig gedurende die vroeë ontwikkeling van 'n oogvolger waarin verskeie konfigurasies van apparatuur en algoritmiese komponente oorweeg word. Die meeste van die gesimuleerde oogvolgdata wat in die oogvolg-verwante literatuur gebruik word, is gegenereer deur gebruik te maak van lae-kompleksiteit oogmodelle met parameters gebaseer op bevolkingsgemiddeldes wat die kornea simuleer as 'n enkele sferiese oppervlak. Hierdie dissertasie het ondersoek ingestel of die verhoging van die oogmodel se kompleksiteit en parameters wat gebruik word om gesimuleerde data te genereer, kan lei tot verbeterings in werklike oogvolguitkomste. Die bevindinge van die dissertasie toon dat die gebruik van 'n oogmodel wat 'n a-sferiese voorste korneale oppervlak en realistiese variasies in die asferisiteit van die voorste korneale oppervlak insluit in 'n gesimuleerde oogvolgomgewing, die voorspellende vermoë van die gesimuleerde data sal verbeter. Dit sal moontlik op sy beurt lei tot die ontwikkeling van oogvolgmetodes met beter resultate.
  • Thumbnail Image
    Item
    The effect of a crossover diffuser configuration on the performance of a mixed flow compressor stage for a micro gas turbine engine.
    (Stellenbosch : Stellenbosch University, 2023-11) Van Eck, H; Van der Spuy, SJ; Stellenbosch University. Faculty of Engineering. Dept. of Mechanical and Mechatronic Engineering.
    ENGLISH ABSTRACT: Mixed flow compressor configurations have become increasingly popular for use in Micro Gas Turbine (MGT) engines. Mixed flow configurations have the advantage of smaller frontal areas while still attaining high pressure ratios from a single stage configuration. The performance of these compressors is largely influenced by the performance of the diffuser, which in turn is limited by inherent geometric restrictions. Research has shown that, compared to legacy diffuser configurations, crossover diffusers provide superior performance in such cases. It does however suffer from inferior operating ranges. This study investigates various crossover diffuser configurations aimed at expanding the operating range and increasing the performance of a mixed flow compressor stage of a MGT engine. To this end, three baseline test compressors, featuring single vane crossover diffusers, are designed using the MATLAB® based 1D Application (App) V2.0. The baseline compressors cover a wide range of design mass flow rates, rotational velocities, and impeller meridional exit angles (mixed flow angles). To evaluate the effect of diffuser configuration on compressor performance, the diffuser configurations of the baseline compressors aremodified, with the achieved performance being compared to that of the baseline configuration. Preliminary performance prediction results for all designs are provided by the 1D App V2.0, which is based on mean line loss model theory. These results are validated using NUMECA/Fine™ Turbo CFD software. Diffuser modifications include various unique single and tandem vane crossover diffuser configurations. A splitter vane configuration provides a marked increase in choke margin, albeit at a slightly reduced stall margin. A relative choke margin increase of up to 106% is observed. Tandem vane configuration evaluations include the investigation of relative vane length, as well as tangential shift of the second vane row relative to the first. A 75% relative tangential shift tandem configuration provides the best overall performance results, regardless of relative vane length. A tandem vane configuration with a reduced first vane row number (half) is proposed. It provides a relative choke margin increase of up to 155% compared to the baseline configuration. As a final validation, the findings of this study are tested on a real-world application. A compressor upgrade is proposed for the CAT250TJ – Gen1 MGT engine. A baseline single vane crossover diffuser and four additional modified vaned diffuser configurations are presented. An on-engine evaluation is conducted using GasTurb™ gas turbine modelling software. The baseline configuration provides a 10.74% static thrust increase compared to the Gen1 configuration at maximum operating conditions. Compared to the Gen1 configuration, a reduced first vane row number diffuser configuration display a design point performance increase of 7.18% and 6.27% for total-to-total efficiency and pressure ratio respectively. An operating margin increase of 9.8% compared to the baseline crossover diffuser is also achieved.