Numerical analysis of airflow and particle deposition in multi-fidelity designs of nasal replicas following nasal administration.

An improved understanding of flow behaviour and particle deposition in the human nasal airway is useful for optimising drug delivery and assessing the implications of pollutants and toxin inhalation. The geometry of the human nasal cavity is inherently complex and presents challenges and manufacturing constraints in creating a geometrically realistic replica. Understanding how anatomical structures of the nasal airway affect flow will shed light on the mechanics underpinning flow regulation in the nasal pharynx and provide a means to interpret flow and particle deposition data conducted in a nasal replica or model that has reduced complexity in terms of their geometries. This study aims to elucidate the effects of sinus and reduced turbinate length on nasal flow and particle deposition efficiencies. A complete nasal airway with maxillary sinus was first reconstructed using magnetic resonance imaging (MRI) scans obtained from a healthy human volunteer. The basic model was then modified to produce a model without the sinus, and another with reduced turbinate length. Computational fluid dynamics (CFD) was used to simulate flow in the nasal cavity using transient flow profiles with peak flow rates of 15 L/min, 35 L/min and 55 L/min. Particle deposition was investigated using discrete phase modelling (DPM). Results from this study show that simplifying the nasal cavity by removing the maxillary sinus and curved sections of the meatus only has a minor effect on airflow. By mapping the spatial distribution of monodisperse particles (10 μm) in the three models using a grid map that consists of 30 grids, this work highlights the specific nasal airway locations where deposition efficiencies are highest, as observed within a single grid. It also shows that lower peak flow rates result in higher deposition differences in terms of location and deposition quantity, among the models. The highest difference in particle deposition among the three nasal models is ∼10%, and this is observed at the beginning of the middle meatus and the end of the pharynx, but is only limited to the 15 L/min peak flow rate case. Further work demonstrating how the outcome may be affected by a wider range of particle sizes, less specific to the pharmaceutical industries, is warranted. A physical replica manufactured without sections of the middle meatus could still be adequate in producing useful data on the deposition efficiencies associated with an intranasal drug formulation and its delivery device.