Lakehead University Knowledge Commons

Knowledge Commons is an open access repository for scholarship and research produced at Lakehead University. It is a free and secure repository for LU faculty, students, staff, and researchers to preserve and present their scholarship.

Communities in KnowledgeCommons

Select a community to browse its collections.

Now showing 1 - 3 of 3

Recent Submissions

A data-driven machine learning framework to model turbulence modulation and preferential concentration by particles
(2026) Waseem, Maryum; Tarokh, Ali; Wang, Wilson; Elyasi, Saimak
Particle-laden turbulent flows appear across a wide range of engineering and environmental systems, from spray combustion and aerosol dispersion to sediment transport and industrial multiphase reactors. Yet accurately predicting particle behaviour within these flows remains a persistent challenge, particularly in resolving how particle inertia couples to turbulent flow structures and how that coupling can be retained at computational scales where full resolution is prohibitive. This dissertation addresses both challenges through two connected studies. The first study establishes a quantitative framework linking particle inertial properties to the topological features of homogeneous isotropic turbulence that govern preferential concentration. Using high-fidelity DNS with two-way Eulerian–Lagrangian coupling on a 1283 grid at Reλ = 120, we systematically investigate the influence of the Stokes number St and the particle-to-fluid density ratio ρp/ρf on particle clustering. The centrifuge mechanism is quantitatively confirmed: clustering intensity peaks at St = 1, where particles are most effectively expelled from vortical regions and accumulate in straindominated zones identified by the Q-criterion. A non-monotonic dependence on density ratio is observed, with the maximum correlation between Q and particle number density occurring at ρp/ρf = 500, reflecting an optimal balance between inertial decoupling and fluid responsiveness. A split-correlation analysis further reveals that vortex exclusion is a robust and nearly universal mechanism, whereas strain-field accumulation exhibits stronger sensitivity to particle inertia and density. The relatively weak magnitude of the overall linear correlations demonstrates that the Q-criterion alone is insufficient to fully characterize preferential concentration, indicating that turbulence intermittency, multiscale particle–structure interactions, and trajectory memory effects also play important roles. The second study builds on these physical insights by developing a machine-learningenhanced framework for particle-laden turbulence simulations. The objective is to recover the small-scale velocity structures that are removed by filtering operations in coarseresolution simulations but are essential for accurately predicting inertial particle dynamics. To achieve this, a multilayer perceptron (MLP) model is trained using DNS data from the first study to reconstruct subgrid-scale velocity corrections from filtered flow fields. The reconstructed velocity field restores a substantial portion of the gradient intensity and high-wavenumber energy content that is absent in the filtered representation. The reconstructed field is then used to drive an Eulerian–Lagrangian particle-tracking solver with full two-way coupling, allowing particle drag forces to be consistently returned to the carrier phase as momentum source terms. Because particles evolve within a velocity field that contains reconstructed small-scale structures, the simulation recovers key features of particle-turbulence interaction that are typically suppressed in filtered simulations. In particular, the ML-enhanced framework restores stronger preferential concentration, more realistic particle slip velocities, and spatially intermittent particle source terms associated with strain-dominated regions of the flow. Together, the two studies form a coherent progression. The first establishes the physical mechanisms governing particle clustering and identifies the flow-topology conditions under which preferential concentration occurs. The second demonstrates that these critical small-scale flow structures can be reconstructed using a data-driven model, enabling improved particle tracking and two-way coupling predictions in simulations where full DNS resolution is not computationally feasible. The combined framework therefore provides a computationally efficient pathway toward higher-fidelity prediction of particle-laden turbulent flows, with direct relevance to industrial and environmental multiphase systems.
Energy-efficient deep neural networks using Domainwall Memory cache for general-purpose graphics processing units
(2026) Namvarmotlagh, Alireza; Atoofian, Ehsan; Deng, Yong; Wu, Kaiyu; Zhou, Yushi
Deep neural networks (DNNs) have become the dominant computational paradigm across computer vision, natural language processing, and generative modelling, yet achieving state-of-the-art accuracy increasingly requires models with billions of parameters and commensurately large memory footprints. These models place extreme bandwidth and capacity demands on the on-chip memory hierarchy of modern general-purpose graphics processing units (GPGPUs), making the shared L2 cache a major contributor to energy consumption. At the same time, aggressive SRAM scaling leads to rapidly increasing leakage power, presenting a fundamental challenge for future high-performance computing architectures. Domain Wall Memory (DWM) is a promising alternative for large on-chip caches due to its ultra-high density and near-zero leakage, but its shift-based access mechanism introduces variable and often high access latency that must be addressed before practical deployment. This thesis presents a hardware–software co-design framework that integrates a DWM-based L2 cache into tensor core (TC)-equipped GPGPUs while mitigating DWM’s shift penalty. On the hardware side, the conventional SRAM data array is replaced with DWM, and tape-head prediction policies are employed that proactively reposition track heads based on predicted access patterns. A hybrid predictor combining stride and two-level context-based prediction achieves the lowest shift overhead among all evaluated strategies. On the software side, structured pruning is applied to representative CNN and transformer models to reduce parameter count and regularize memory accesses, and TC-optimized kernels are implemented that efficiently exploit the pruned structures. Across a suite of seven convolutional and attention-based DNN models, pruned DWM-based L2 caches achieve an average energy saving of 73.2% compared to an unpruned SRAM-based L2 cache, while delivering an average performance improvement of 13.5% that effectively mitigates performance degradation across all evaluated models. Under iso-area conditions, the DWM-based L2 cache achieves 17× more capacity than SRAM, enabling it to outperform SRAM by 7% to 37.4% in execution time and reduce energy consumption by 53.3% to 71.6%. The resulting Energy–Delay Product (EDP) of SRAM is 2:3× to 4:58× higher than that of DWM. These results demonstrate that carefully cooptimizing emerging non-volatile memories at both software and hardware levels can deliver energy-efficient DNN acceleration without sacrificing performance.
Investigation of Solar-Powered Geothermal Heat Pump System for decarbonizing space heating at Thunder Bay Regional Health Sciences Centre, Northwestern Ontario
(2026) Nagi, Anjali; Ismail, Basel I.; Tarokh, Ali; Salem, Sam
This thesis investigates the technical feasibility, economic viability, and environmental impact of implementing a Solar-Powered Geothermal Heat Pump (SP-GHP) system to decarbonize space heating at the Thunder Bay Regional Health Sciences Centre (TBRHSC) in Northwestern Ontario. Decarbonizing the heating sector in extreme cold climates is a critical challenge, as conventional systems in such regions are typically carbon-intensive and fossil-fuel reliant. The research employs an integrated methodology combining experimental laboratory testing, regional geological characterization, building energy modeling, and solar photovoltaic (PV) optimization. Experimental results from a lab-scale geothermal heat pump simulator at Lakehead University validated the system's thermal responsiveness, demonstrating that supply air temperatures and Coefficient of Performance (COP) increase linearly with entering water temperatures. Utilizing regional soil data from the nearby Musselwhite mine, the study identified that ground temperatures are most favorable at a shallow depth of approximately 5 meters, which avoids the high costs of deep drilling. A detailed heat load analysis for the 650,000 ft² facility estimated a peak heating load of 640 kW and an annual demand of 1.43 GWh. To meet this demand, a horizontal slinky-loop ground heat exchanger was designed, requiring approximately 20 km of pipe. Furthermore, location-specific solar simulations determined an annual average optimum tilt angle of 39°, while a load-responsive winter tilt of 67° was proposed to synchronize solar electricity generation with peak winter heating requirements. Techno-economic and environmental assessments conducted via RETScreen Expert software revealed that the proposed system would achieve a 96.4% reduction in annual heating-related greenhouse gas emissions. While the project requires a substantial initial investment of approximately $3 million, sensitivity analysis suggests it becomes financially viable with a payback period of 12 to 15 years when supported by clean energy incentives and carbon pricing. This study concludes that integrated SP-GHP systems represent a technically sound and environmentally advantageous solution for advancing low-carbon transitions in large-scale institutional buildings like hospital, within extreme cold climatic regions.
Health equity in action: assessing the impact of the Healthy Kids Family Program using the RE-AIM Framework
(2026) Pearson, Hannah; Pearson, Erin; Klarner, Taryn; Boynton, Heather
Background: Families living in equity-deserving communities experience disproportionate barriers to engaging in health-promoting behaviours such as healthy eating and physical activity due to intersecting social determinants of health. Community-based health promotion programs are well positioned to address these inequities; however, there remains a need for comprehensive, real-world evaluations. The RE-AIM Framework (Reach, Effectiveness, Adoption, Implementation, and Maintenance), is a robust tool for evaluating community-based initiatives, yet its application within family-focused, equity-oriented interventions remains under- represented. Purpose: The purpose of this study was to conduct a summative evaluation of the Healthy Kids Family Program (HKFP), a four-week, community-based health promotion program delivered in equity-deserving neighbourhoods in Thunder Bay, Ontario. Guided by the RE-AIM Framework, this study aimed to assess the utility, impact, and sustainability of the HKFP as a model for fostering health behaviours in families and children. Method: A pragmatic mixed methods case study design was employed, using a combination of quantitative and qualitative data collected from 2021 – 2025. Involving a priori and new data and in-line with the RE-AIM Framework, this entailed an assessment of the: (1) Reach associated with the target population; (2) Effectiveness of the program; (3) Adoption by program providers and resident participants; (4) Implementation fidelity with respect to anticipated versus actual program delivery; and (5) Maintenance of program-related outcomes among recipients over time. Quantitative data were obtained from validated health measurement tools assessing health behaviours, self-efficacy for nutrition, and quality of life via two summary scores involving Physical and Mental health. These surveys were administered at baseline, immediately post-intervention, and 6-weeks post-intervention across 11 iterations of the HKFP. Analysis of Variance (ANOVA) testing, dependent t-testing, and descriptive statistics were employed to analyze quantitative data to examine changes over time. Qualitative data were derived from multiple sources including semi-structured interviews with program participants and staff, as well as program documentation including program fidelity notes, and were analyzed using reflexive thematic analysis. Data sources were triangulated and organized deductively via RE-AIM dimensions to provide a comprehensive evaluation of the program and its impact. Results: Data from 60 participants were obtained from HKFP assessments. Participant demographics revealed that the majority who participated were female (n = 52), aged 22-48 years (M = 35.52), with a mean monthly income of $3,396. Results for Reach showed that the HKFP engaged primarily female individuals across diverse areas of Thunder Bay, though participation was influenced by structural barriers including childcare and scheduling conflicts. Significant improvements were noted in Effectiveness. Quantitative results indicated a significant increase in Self-Efficacy for Nutrition scores between baseline (M = 3.43, SD = 0.89) and follow-up (M = 3.93, SD = 0.72), p < 0.001, η² = .11. No significant improvements in health behaviours or quality of life took place. Assessment of Adoption revealed that the HKFP was supported through strong community partnerships and collaboration. Implementation revealed that program delivery was shaped and adapted by both facilitators to engagement (i.e., supportive environments and resource access) and barriers to engagement (i.e., duration of program session, pandemic-related restrictions). Regarding Maintenance, participants shared that they applied knowledge learned through program delivery into their daily lives; however, longer-term sustainability was influenced by ongoing structural challenges. Qualitatively, participation in the HKFP was described by participants as influenced by contextual barriers including conflicting schedules, transportation, and childcare. Qualitative findings obtained from HKFP participant interviews (n = 7) highlighted themes related to structural barriers (i.e., childcare, financial constraints), social support and community (i.e., peer connection, sense of belonging), as well as empowerment and increased confidence in health behaviours (i.e., self-efficacy and feeling capable of making change). Other qualitative findings obtained from the administrative team of the HKFP (n = 3) contextualized the realities of program offerings and described the importance of collaboration between organizations offering health promotional programming. Conclusion: This study demonstrated the importance of employing a robust, well-rounded framework when considering study design, specifically when working with equity-deserving communities. For example, statistically significant gains were made in self-efficacy related to nutrition, which may be attributed to the program’s emphasis on skill-building, resource access, and supportive learning environments. Qualitative findings suggest that a meaningful shift in perception surrounding healthy eating took place. They also highlight the importance of contextualized measures of success that extend beyond traditional behavioural outcomes to include empowerment, feasibility, and sustainability. The HKFP represents a promising model for community-based health promotion in equity-deserving communities, because the program offerings provided an opportunity for participants to learn information and make informed decisions surrounding their health. Future research should prioritize strategies to enhance long-term maintenance, including longer or tapered intervention sessions over time, address structural participation barriers including transportation, childcare, and conflicting schedules. This study demonstrated that short-term, community-based health promotion programs can produce meaningful impact for an equity-deserving community through gains in self-efficacy and health behaviour framing, even in the absence of immediate, measurable behaviour change.
The paradox of care: service providers’ experiences of delivering opioid-related care in rural Northwestern Ontario
(2026) Ambury, Sydney; Sprakes, Abigale; Kortes-Miller, Kathy; Sanderson, Kathy; Potvin, Leigh
The opioid epidemic continues to impact rural communities across Canada, with Northwestern Ontario (NWO) experiencing disproportionately high rates of opioid-related harms relative to the rest of the province. Despite this, little research has examined how service providers in rural NWO understand and navigate opioid-related care in practice. Grounded in a Social Determinants of Health (SDOH) framework, this qualitative descriptive study explored how providers understand the influence of the SDOH on opioid-related harms and service engagement, what strategies they use to deliver and adapt care for clients who use illicit opioids, and what their experiences suggest for sustaining and strengthening rural opioid-related service systems. 14 semi-structured interviews were conducted with service providers across health, social service, and justice sectors in rural NWO, and the data were analyzed using reflective thematic analysis. Six themes were generated and organized into three conceptual pairings: Structural Inequities and Reconfiguring Service Delivery, Environmental Vulnerabilities and Relational Care, and Professional Strain and Workforce Sustainability. Across these three pairings, the findings revealed that the conditions constraining rural care simultaneously produced the practices that sustained it. Participants absorbed service gaps through role expansion and informal coordination, built trust within community and institutional contexts that caused harm, and sustained their commitment through the same attachments that depleted them. Together, these findings point to a service system whose functioning depends largely on invisible forms of labour unevenly distributed across the workforce. The findings have implications for social work practice, rural service policy, and the conditions needed to sustain providers.