Hybrid quantum-classical machine learning for canonical fluid dynamics and heat transfer problems

This thesis introduces the implementation of some hybrid frameworks for quantum and classical machine learning techniques associated with the fluid dynamics and heat transfer problems. Chapter 1 discusses an introduction concerning the high-performance computing, physics-based CFD simulations, classical machine learning, and quantum machine learning. In Chapter 2, the study further concentrates on the exploration of advanced physics-base computational methods emphasizing data generation to be considered as input into the following machine learning models. Chapter 2 delves into the numerical analysis of fluid dynamics and heat transfer problems such as airborne particle dispersion, deep-sea sponge, inverse heat conduction problem, and turbulent reacting flow. This chapter highlights the use of simulation to understand complex problems in precise settings, providing insights into the interaction between various fluids and the impact on heat transfer processes. Chapter 3 shifts the focus to the machine learning approaches into the behavior of hybrid fluids. This exploration reveals significant enhancements in analyzing the fluid dynamics and heat transfer problems due to the unique characteristics which the machine learning and quantum computing can offer to the research community. Together, these chapters (chapter 2 and chapter 3) present a comprehensive study that integrates computational fluid dynamics (CFD) simulations with machine learning research to advance the understanding of fluid behavior and heat transfer mechanisms, contributing to the development of more efficient and effective thermal management solutions. Finally, the thesis is concluded by Chapter 4 outlining the major findings along with this thesis.