Applications of Data-Driven Learning Models in Fluid Mechanics: Solid-Fluid Multiphase Systems and Bat Flight

dc.contributor.authorRaj, Neil Ashwinen
dc.contributor.committeechairTafti, Danesh K.en
dc.contributor.committeememberPalmore, John A.en
dc.contributor.committeememberKarpatne, Anujen
dc.contributor.committeememberPitchumani, Rangaen
dc.contributor.departmentMechanical Engineeringen
dc.date.accessioned2025-06-10T08:02:51Zen
dc.date.available2025-06-10T08:02:51Zen
dc.date.issued2025-06-09en
dc.description.abstractFluid-solid interactions are central to numerous industrial, environmental, and biological systems, encompassing particle-laden flows in chemical processing, urban airflow, and biological locomotion. Understanding and studying these interactions is essential for optimizing engineering designs and advancing scientific knowledge. This thesis explores three distinct yet interconnected topics within this domain: (1) deep learning-based drag force modeling for particulate suspensions, (2) reduced-order modeling (ROM) for flow predictions in randomly arranged solid arrays, and (3) data-driven analysis of bat flight kinematics. First, we address the challenge of accurately predicting individual particle drag forces in dense suspensions. Meso-scale models such as CFD-DEM often rely on empirical drag correlations that neglect microscale particle-particle interactions, leading to inaccuracies. To bridge this gap, we develop deep learning models—including convolutional neural networks (CNNs) and graph neural networks (GNNs)—that incorporate both local particle neighborhood information and global suspension parameters. We demonstrate the effectiveness of these models in predicting particle-scale drag forces more accurately than existing empirical methods, paving the way for improved meso- and macro-scale simulations of particulate flows. Additionally, we employ a physics-based learning paradigm and see that incorporating physical constraints to traditionally "black box" models improves performance, especially in cases of data paucity. Next, we investigate the application of deep learning and data-driven approaches for reduced-order modeling of complex flow fields. Traditional CFD methods, while highly accurate, are computationally expensive and unsuitable for real-time control or parameter sweeping for optimization tasks. Here, we compare convolutional autoencoder (CAE) models combined with recurrent architectures such as long short-term memory (LSTM) networks and temporal convolutional networks (TCNNs) against dynamic mode decomposition (DMD) for predicting the evolution of flow fields over time. Our analysis focuses on two-dimensional flow across randomly arranged infinite cylinders, highlighting the strengths and limitations of each method in predicting the long-term flow dynamics. Finally, in the domain of solid and fluid phase interactions, we shift our focus to biological fluid-structure interactions, specifically relating to bat flight. Unlike birds, bats possess highly flexible membrane wings, enabling enhanced maneuverability and complex aerodynamic adaptations. We analyze high-resolution bat flight data using proper orthogonal decomposition (POD) to identify key motion patterns and independent marker points on the bat's wing surface. Additionally, we develop data-driven models to map bat wing kinematics in the body-fixed or local frame to its global trajectory and vice versa, offering insights that could inform the design of bio-inspired flapping-wing unmanned aerial vehicles (UAVs). Together, these studies contribute to advancing computational methods for fluid-solid interactions across multiple domains, from industrial flow systems to biological locomotion. By integrating deep learning with physics-based modeling, this work provides novel frameworks for understanding and predicting complex flow phenomena, with broad applications in engineering and bio-inspired robotics.en
dc.description.abstractgeneralThe interaction between fluids and solid structures plays a crucial role in science, engineering, and nature. Whether it's refining petroleum-based fuels, improving the quality of biomass-based fuels, improving air pollution control, designing efficient waste management systems or designing efficient heat exchangers, understanding how fluids and solids particles interact and how fluids flow around obstacles is vital for technological advancements. This research explores three distinct but interconnected aspects of fluid mechanics using cutting-edge deep learning techniques. First, we tackle a long-standing challenge in modeling particle-laden flows—predicting the forces exerted on individual particles suspended in a moving fluid. Traditional simulations are highly accurate but computationally expensive, limiting their practical use. By leveraging deep learning, we develop models that efficiently predict particle drag forces while maintaining high accuracy, helping to bridge the gap between complex simulations and real-world applications. Next, we investigate how deep learning can accelerate the prediction of fluid flow behavior in complex environments, such as air moving through urban landscapes or fluids passing through industrial filters. Instead of running costly simulations, our models learn from existing data to predict future flow states, offering a powerful tool for engineers to optimize designs and improve efficiency. Finally, we turn to the natural world, studying the intricate wing movements of bats in flight. Unlike birds, bats have flexible wings that allow them to maneuver with remarkable agility. By analyzing bat flight kinematics, we develop machine learning models that map wing movements to flight trajectories and vice versa. This research not only deepens our understanding of bat aerodynamics but also has potential applications in the development of bio-inspired flying robots. By combining physics-based insights with modern machine learning approaches, this work advances our ability to simulate, predict, and optimize complex fluid-structure interactions across engineering, environmental, and biological systems.en
dc.description.degreeDoctor of Philosophyen
dc.format.mediumETDen
dc.identifier.othervt_gsexam:43478en
dc.identifier.urihttps://hdl.handle.net/10919/135438en
dc.language.isoenen
dc.publisherVirginia Techen
dc.rightsIn Copyrighten
dc.rights.urihttp://rightsstatements.org/vocab/InC/1.0/en
dc.subjectDeep Learningen
dc.subjectParticle-Fluid Multiphase Flowsen
dc.subjectPhysics Informed Machine Learningen
dc.subjectData Driven Model Order Reductionen
dc.subjectBat Flight Kinematicsen
dc.titleApplications of Data-Driven Learning Models in Fluid Mechanics: Solid-Fluid Multiphase Systems and Bat Flighten
dc.typeDissertationen
thesis.degree.disciplineMechanical Engineeringen
thesis.degree.grantorVirginia Polytechnic Institute and State Universityen
thesis.degree.leveldoctoralen
thesis.degree.nameDoctor of Philosophyen

Files

Original bundle
Now showing 1 - 1 of 1
Name:
Raj_NA_D_2025.pdf
Size:
45.33 MB
Format:
Adobe Portable Document Format