Fundamentals of Gas-Particle Flows and Applications in the Human Respiratory Airway

Accurate simulation of aerosol transport and deposition in the human respiratory airway is essential for optimising intranasal drug delivery, assessing environmental exposure risks, and improving surgical outcomes. Computational fluid dynamics (CFD) has become a powerful tool for investigating these flows, yet many studies prioritise geometric fidelity while overlooking the fundamental mechanisms of particle transport. These mechanisms, governed by turbulence–particle interactions, numerical integration accuracy, and stochastic dispersion effects, strongly influence deposition predictions but are often not systematically examined. Without such analysis, the reliability and reproducibility of respiratory CFD results are limited. This thesis develops and validates a computational framework for gas–particle flows that integrates both anatomical realism and fundamental particle transport physics. High-resolution CFD simulations were performed in Ansys Fluent using constant inhalation flow rates of 15, 30, and 60 L/min, as well as pulsatile flows to aid sinus ventilation and aerosol penetration. The gas phase was resolved using a hybrid RANS-LES approach to capture both unsteady flow structures and turbulent features. Particle trajectories were modelled using an Eulerian-Lagrangian approach coupled with stochastic turbulence dispersion for a range of particle diameters relevant to pharmaceutical aerosols. The numerical methodology was first examined using simplified cylindrical tubes and 90$^{\circ}$ bend configurations as analogues to localised regions of the respiratory airway. Such well-defined geometries allow for direct comparison between modelling strategies and validation of numerical findings. These benchmark studies quantified the effects of turbulence model selection, particle time-integration scheme, and dispersion treatment on near-wall particle behaviour and deposition outcomes. The results provided clear guidance on numerical settings that balance accuracy and computational efficiency. Validated modelling strategies were then applied to patient-specific nasal airway geometries reconstructed from medical imaging. Steady and pulsatile flow simulations were used to characterise mixed laminar–turbulent regimes, secondary flow structures, and their effects on aerosol transport pathways. Pulsatile flows were found to enhance maxillary sinus ventilation and particle penetration compared with steady flows, revealing flow-particle behaviours not captured under constant inhalation conditions. The outcomes of this research demonstrate that resolving the fundamentals of particle transport is as critical as anatomical detail in predicting respiratory aerosol deposition. Gas-phase model, particle integration accuracy, and dispersion model selection were shown to significantly influence deposition patterns, particularly in near-wall regions where impaction and capture occur. By establishing a validated, reproducible methodology and providing best-practice recommendations, this thesis advances both the fundamental understanding of particle-laden respiratory flows and the applied methods used to simulate them. The findings offer practical value to biomedical engineers, clinicians, and researchers developing inhalation therapies, assessing exposure risks, or designing surgical interventions aimed at improving nasal airflow and sinus function.<p></p>