Issue link: https://www.e-digitaleditions.com/i/1469783
12 June 2022 Inhalation wall y + is a non-dimensional distance parameter that is controlled for accurate modeling of near-wall tur- bulence characteristics in wall-bounded flows). Setting up a CFD nasal spray simulation Nasal sprays have a characteristic spatio-temporal variation from a dense spray droplet regime adjacent to the spray nozzle to a dilute regime further away from the spray nozzle. Spray simulations reveal that spray droplets injected into the nasal airway cavity impart significant momentum to the fluid phase, which induces a cloud-based gas velocity and signif- icant coupling between the phases (Figure 1). e cloud-based gas velocity as a whole imparts significant momentum on the smaller droplets and transports them further into the flow field, as fully captured in a two-way coupled simulation [1, 5]. Based on the validated nasal spray model, nasal spray transport with two-way coupled momentum exchange and the related impact on the dispersed phase transport is realistically captured by CFD simulations using an Euler-Lagrange framework [1, 5]. To simulate a nasal spray product, the CFD model requires initial and boundary conditions, and ideally benchmark validation data. CFD nasal spray models typically require input parameters such as polydisperse spray droplet size distribution, spray cone angle and plume shape, spray mass flow rate, spray velocity and formulation density. ese metrics are typically first evaluated in an in vitro setting to characterize the nasal spray using bench-top experimental setups [14]. In a two-way coupled simulation, spray droplets must be injected with a predetermined mass flow rate so that the net mass of the liquid droplets injected is equal to Nasal spray modeling—best CFD modeling practices Developing and meshing nasal airway geometries A time-consuming but critical step in the process of building a CFD model is developing and mesh- ing the geometries associated with the system. For nasal spray drug delivery simulations, the nasal air- way geometries are typically extracted/reconstructed from computed tomography (CT) or magnetic res- onance imaging (MRI) scans by specifying appro- priate radio-density thresholds that can delineate the air-tissue boundary [34]. e extracted nasal air- way geometries from scanned images are typically exported in stereolithographic (STL) format. ese STL-formatted geometries may need further refin- ing to remove surface irregularities from the recon- struction process and to simplify the geometry, for example, by removing the sinus cavities at the narrow ostium inlets. It can be easier to work with a comput- er-aided design (CAD) accessible geometry format since editing/trimming STL surfaces is usually cum- bersome. e STL-formatted surfaces are typically refined and translated into common CAD formats by "skin-overlay" with reasonably large polygonal patches consisting of b-splined surfaces, conforming to the faceted surfaces. is step reduces thousands of STL triangles to approximately 100 CAD-format surfaces and makes it possible to perform CAD pro- cessing with any third-party CAD software package [5, 28]. While the spray bottle geometry may not be an important factor for simulating the nasal spray dynamics, it is necessary to account for the presence of the spray pump inside the nasal geometry. is can be achieved by subtracting the volume occupied by the spray pump from the nasal airway volume. is step can realistically capture the partial occlusion of the nostril inlet by the spray pump and its effect on nasal inhalation airflow. e CFD simulations resolve the flow field within millions of representative micro-to-millimeter scale discrete volumes (i.e., control volumes) inside the geometry and predict aerosol transport as the flow progresses through these volumes. In order to suc- cessfully resolve flow within an airway geometry, a computational mesh (or grid) needs to be created with sufficiently small control volumes (or elements) in regions of high flow gradients to minimize com- putational errors [21]. Accurate spray simulations in nasal airways have been shown to require ~2 mil- lion polyhedral elements per nasal cavity, excluding the nasopharyngeal region, together with near-wall prism elements and a refined mesh in the vicinity of the spray nozzle tip. Additionally, to capture the near-wall flow dynamics, near-wall regions require approximately ~5 near-wall prism layer mesh ele- ments with a first prism layer thickness resulting in an average wall y + of approximately one [5] (where Figure 1 CFD nasal spray model with spray droplets injected from the spray nozzle orifice with a solid conical injection shape and a turbulent velocity profile. The momentum- jet due to the momentum exchange from the spray droplets is shown in the sub-figure.