Nasal high flow (NHF) cannulae are used to deliver heated and humidified air to patients at steady flows ranging from 5-50 l/min. Knowledge of the airflow characteristics within the nasal cavity with NHF and during natural breathing is essential to understand the treatment's efficacy. In this thesis, the distribution and velocity of the airflow in the human nasal cavity have been mapped during natural and NHF assisted breathing with planar- and stereo-PIV in both steady and oscillatory flow conditions. Anatomically accurate transparent silicone models of the human nasal cavity were constructed using CT scan data and rapid prototyping. Breathing flowrates and waveforms were measured in vivo and dimensionally scaled by Reynolds and Womersley number matching to reproduce physiological conditions in vitro. Velocities of 2.8 and 3.8 m/s occurred in the nasal valve during natural breathing at peak expiration and inspiration, respectively; however on expiration the maximum velocity of 4.2 m/s occurred in the nasopharynx. Velocity magnitudes differed appreciably between the left and right sides of the nasal cavity, which were asymmetric. NHF modifies nasal cavity flow patterns significantly, altering the proportion of inspiration and expiration through each passageway and producing jets with in vivo velocities up to 20.8 m/s for 40 l/min cannula flow. The main flow stream passed through the middle airway and along the septal wall during both natural inspiration and expiration, whereas NHF inspired and expired flows remained high through the nasal cavity. Strong recirculating features are created above and below the cannula jet. Results are presented that suggest the quasi-steady flow assumption is invalid in the nasal cavity during both natural and NHF assisted breathing. The importance of using a three-component measurement technique when investigating nasal flows has been highlighted. Cannula flow has been found to continuously flush the nasopharyngeal dead space, which may enhance carbon dioxide removal and increase oxygen fraction. Close agreement was found between numerical and experimental results performed in identical conditions and geometries.

This work aims to assess human nasal blockage by investigating its influence on nasal airflow dynamics, both computationally and experimentally. An in-house CFD code (Lithium) computes the steady (mean) nasal airflow for a cavity constructed from CT images of a healthy adult, for the internal cavity and for the first time for the external flow. To account for turbulence occurrence, the low Reynolds number k-ω Reynolds-Averaged-Navier-Stokes (RANS) model is used. The flow field is calculated at different breathing rates by varying the influx rate. Blockages are introduced at various locations inside the cavity to investigate common nasal blockages. The computational results are assessed against published literature and the Particle Image Velocimetry experimental (PIV) results, carried out on a 2.54:1 scale model of the computational nasal cavity. Schlieren optical technique is also used for external nasal airflow visualizations of a human subject, to comment on using an optical system for clinical application. These computations reveal a significant dependency of both, the internal and external nasal airflow fields on the nasal cavity’s geometry. Although for this model, the flow is found to be turbulent in the inspiratory phase of 200 ml/s and higher, it is suggested that the nature of flow can vary depending on the nasal cavity’s structure which is influenced by genetics. Nevertheless, some common flow features were revealed such as higher flow rate in the olfactory region and main flow passage through lower airways during inspiration. More uniform flow passage was found in expiration. The results also suggest a possible correlation between the internal geometry of the cavity and the external nasal airflow angle and thickness. This correlation can allow an application of optical systems such as Schlieren which is shown to give accurate qualitative images of the external nasal airflow for assessment of the nasal blockage.