Search UNB Scholar
About UNB Scholar
UNB Scholar is an institutional repository initiative of UNB Libraries intended to collect, preserve, showcase, and promote the open access scholarly output of the UNB community. Use UNB Scholar to explore specific collections, or search all content in the repository. Material submitted to the repository will also be freely discoverable online through Google and other major search engines.
UNB Libraries welcomes submissions to UNB Scholar. Find out more about depositing your publications!
Communities in DSpace
Select a community to browse its collections.
Recent Submissions
Item type:Item, Phase transformation behavior of microalloyed medium-manganese steels(University of New Brunswick, 2026-08) Zvavamwe, Felisters; Aranas Jr, ClodualdoMedium-manganese (medium-Mn) steels are promising candidates for advanced high-strength steel (AHSS) applications due to their ability to achieve an exceptional combination of strength and ductility through transformation-induced plasticity (TRIP) and twinning-induced plasticity (TWIP) mechanisms. However, the fundamental understanding of how microalloying influences phase transformation behavior, microstructural evolution, and mechanical response remains limited, particularly regarding solute partitioning, precipitation, and grain boundary phenomena. Therefore, this work systematically investigates the role of microalloying in medium-Mn steels, with particular emphasis on vanadium (V) and molybdenum (Mo) additions. A series of steels were produced through laboratory-scale casting with controlled compositional variations (0-2 wt.% V and selected Mo additions). The experimental program combined scanning electron microscopy (SEM), electron backscatter diffraction (EBSD), energy-dispersive spectroscopy (EDS), X-ray diffraction (XRD), mechanical testing, and thermodynamic calculations to evaluate phase constitution, crystallographic characteristics, mechanical response, phase stability, and solute partitioning behavior. Following casting, vanadium microalloying significantly influenced microsegregation, prior austenite grain structure, and phase constitution. Increasing vanadium content promoted grain refinement and altered retained austenite distribution through solute partitioning and carbide formation. During intercritical annealing, austenite reversion was governed by the interplay among thermodynamic driving forces, diffusion kinetics, and precipitation behavior. Austenite nucleation occurred preferentially at high-angle grain boundaries due to enhanced solute segregation, while growth was controlled by grain boundary, dislocation (pipe), and bulk manganese diffusion. Vanadium carbide formation reduced carbon availability, increasing the nucleation barrier and decreasing density of active nucleation sites. Significant δ-ferrite formation was observed at high vanadium additions. The retention of δ-ferrite was found to be influenced not only by vanadium content but also by solidification conditions and segregation behavior. Furthermore, δ-ferrite constrained martensitic transformation, resulting in variant selection. Building upon these findings, molybdenum microalloying was explored as a preliminary investigation for future work. Molybdenum delayed martensitic transformation, refined prior austenite grains, reduced microstructural banding, and promoted a more uniform retained austenite distribution. In summary, this work provides fundamental insights into the effects of vanadium and molybdenum microalloying on phase transformation behavior in medium-Mn steels through solute partitioning, diffusion, and precipitation phenomena.Item type:Item, Investigating Autoregressive Diffusion Models: Internal guidance, training schedules, and black-box model behaviour(University of New Brunswick, 2026-08) Bourque, Jordan; Palma, FrancisAutoregressive Diffusion Models (ARDM) combine autoregressive sampling with diffusion-based learning, blending two major approaches in modern generative modelling. Although ARDM already benefits from both autoregressive factorization and diffusion-style denoising, its internal UNet representations remain difficult to interpret, and its training objective is primarily driven by scalar likelihood-based losses. This thesis investigates whether internal guidance and scheduling-based interventions can improve or alter ARDM behaviour under fixed-budget experimental conditions. The thesis evaluates one baseline ARDM and forty-six modified ARDM variants on CIFAR-10. The experimental families include basic discriminator guidance, advanced discriminator and EMA-reference reasoning guidance, counterfactual feature-credit guidance, and scheduling-based timestep sampling. All final experiments use the same dataset, model architecture, training duration, and evaluation pipeline, and, therefore, the observed differences are attributed to the tested intervention rather than to model scaling or additional data. The results show that no intervention universally dominates the baseline across all metrics. EMA-reference reasoning guidance achieves the strongest likelihood-oriented results, producing the best evaluation NELBO. Low-weight end-feature discriminator guidance produces the best FID, adaptive scheduling produces the best Inception Score, and scalar counterfactual guidance produces the best SSIM. These findings suggest that ARDM improvement is metric-specific: likelihood, sample quality, structural similarity, and training stability do not always move together. Beyond final metric comparison, the thesis provides controlled probes of internal ARDM behaviour through feature placement, EMA feature matching, counterfactual channel perturbation, gradient-alignment diagnostics, and timestep scheduling. The main contribution is therefore a controlled empirical analysis of how internal guidance and scheduling interventions shape ARDM behaviour, rather than a claim of a single universally superior model.Item type:Item, Privacy-preserving data analytics on trusted execution environments(University of New Brunswick, 2026-08) Argho, Ankon Ghosh; Ray, Suprio; Mandal, KalikinkarModern data-intensive applications often run on platforms that cannot be fully trusted. Sensitive data must therefore be protected during communication, storage, and processing. We study privacy-preserving data analytics in two application areas. First, we present ASKIndex, a concurrent, disk-aware learned index with a lock-free core. We also present ASKIndexSE, which adds authenticated encrypted persistence, secure client communication, and SGX-enabled execution. ASKIndex achieves up to three times higher throughput than the evaluated disk-aware learned-index baselines. Under the stated threat model, ASKIndexSE protects encrypted-mode query payloads and persistent page contents. Second, we study federated LSTM forecasting over decentralized smart-energy data. The system protects federated messages with authenticated encryption. It also evaluates local training and server coordination in a Gramine protected-execution simulation. The selected 1-hour configuration reports MAE of 0.4342±0.0016 and RMSE of 0.6856 ± 0.0008 across three random seeds. Together, the two systems show that privacy mechanisms must match the workload. They also require explicit threat models, careful key management, clear leakage boundaries, and measured performance costs.Item type:Item, Feasibility study: A multi-source data approach to healthcare travel demand analysis in southeastern New Brunswick(University of New Brunswick, 2026-08) Ijeh, Ekene Splendor; Hanson, TrevorHealthcare-related travel is an important but often underrepresented component of transportation planning because it is infrequent, underreported, or hidden within broader trip-purpose categories. This thesis examines whether multiple data sources, three in this case, can complement each other to support healthcare travel demand analysis in southeastern New Brunswick. The Destination 2040 Household Travel Survey identified 32 healthcare-related trips, representing 0.31% of more than 10,000 trips, but contributed mode, distance, purpose, and tour-structure information needed for behavioural interpretation. The Administrative Health Data provided the strongest basis for trip generation by converting confirmed healthcare encounters into trip rates by service stream, population group, geography, and facility. Cellphone mobility data supported trip distribution and facility-catchment analysis by revealing observed facility activity, repeated presence, and broader trip sequences. The study concludes that a multi-source framework provides a stronger basis for healthcare transportation planning than any single dataset alone.Item type:Item, Designing tangible technologies to support seniors social connections(University of New Brunswick, 2026-08) Short, Taylor; Bateman, ScottAs mobility and health decline with age, and as technology becomes more complex, many seniors face barriers to staying socially connected with friends and family at a distance. Games are a key way to create social connections, but online games often require complex interfaces that pose barriers to seniors’ participation. Tangible interfaces that provide physical objects as input and output for interactive systems may provide an intuitive alternative, but their suitability for this demographic has not been well studied. To address this, we iteratively designed tangible interfaces for an online card game, moving from an initial AR-based system (limited by tracking issues) to a final e-ink playing card design that reduced accessibility barriers. Seniors believed the e-ink system was fun and could support remote play. Our work demonstrates how novel play interfaces can be designed and built to reduce the barriers seniors face in connecting with others around social play.
