Issue link: https://www.e-digitaleditions.com/i/1475629
Inhalation August 2022 11 phenomenological models. ese may be time-vary- ing, due to the transient nature of most devices. It is important to model air entrainment accurately, both that of air into the aerosol and of the small- est particles into any air co-flow. Flow and surface temperatures are important predicted parameters because they indicate the patient's experience. In addition, temperature influences physical properties that govern particle/surface interaction, liquid atom- ization and break-up, evaporation and condensation rates [21, 22]. Surface interaction models can assess whether particles deposit on surfaces or rebound, as a function of surface properties as well as physical particle properties including size, velocity and angle of impact. Respiratory deposition and drug delivery Modeling of airflow, particle flow and deposition inside the lower respiratory system, using 3-D geometry generated from anatomical image data (e.g., a computed tomography (CT) scan) [2, 23], is the final step in predicting the delivery of drug aerosol particles. ese simulations seek to assess the probability of deposition of particles of differ- ent sizes at various locations. Using image data to generate the CFD modeling domain is often known as image-CFD or image-based CFD modeling [19, 24]. 3-D simulation of vascular systems, including drug transport, is also possible [24, 25], with similar recent advances in ease-of-use, availability, comput- ing power and understanding. Complex fluid behavior e fluid formulations present in pharmaceutical aerosols are sometimes non-ideal, highly viscous, non-Newtonian or viscoelastic in behavior, e.g., nasal sprays with added mucoadhesives, recently discussed in Inhalation [26]. is altered viscosity can affect spray behavior including nozzle flow rate, atomiza- tion and break-up into droplets, and interaction of the liquid phase with the surfaces of the nasal cavity. CFD simulation of highly viscous, non-Newtonian or viscoelastic fluids—e.g., for nasal spray applica- tions—is possible and has been explored [27, 28]. Industrial approach to CFD modeling A business that utilizes CFD modeling for product development or research may select open-source CFD modeling software or use proprietary software. Open-source software will typically be free at point- of-use and offers full control for users, but training is recommended, as highlighted in reference 8. Product lifecycle management packages in use within organi- zations may have basic built-in CFD solvers that can be used freely or purchased at lower cost. Types of CFD simulation for pharmaceutical aerosols CFD simulations can model virtually any fluid- dynamic process between a pharmaceutical device and the targeted treatment area, mirroring the jour- ney of drug molecules towards their intended (or unintended) destination. An article by Ruzycki, et al. [10] is still useful in describing the challenges and successes of pharmaceutical inhaler simulation. Par- ticularly apt is the opinion that CFD analyses are most useful in conjunction with experimental stud- ies. Excellent reviews of most aspects of pharma- ceutical aerosol study where CFD can be effectively applied are found in references 11 and 12, and these remain relevant from a simulation perspective. Device internal flow e aim of simulating device internal flow is to pre- dict or understand the state at the device exit, for example, the study of phase change and atomization in the two-phase flow through the nozzle of a pMDI actuator [13, 14]. Such flow is difficult to simulate, being complex, 3-D, multi-phase, small-scale and transient, as illustrated in experimental observations [15-17]. For DPIs, there is complex internal flow with different phenomena, as observed in a prod- uct development study that used validated CFD to investigate powder aerosolization in a new type of DPI [18]. Near-field aerosol With this approach, the aerosol closest to the device (the near-field or near-nozzle spray) is modeled, sometimes within the same simulation as the internal flow just described [14]. Modeling only the first few millimeters from the nozzle allows high-fidelity sim- ulation of the dense particle flow, with atomization phenomena predicted. ese types of simulation aim to predict initial particle size, velocity distributions and spray angles before an aerosol interacts with its environment (e.g., a patient's oral/nasal cavity or a USP induction port throat (USP-IP)). Results might be used as initial conditions in downstream studies. Far-field aerosol e aim of far-field aerosol simulations is to under- stand where particles of different sizes travel and deposit, accounting for interaction of the aerosol with the patient or testing device. ese may involve oral or nasal cavity geometry [19], a model throat or other analytical test device [20], or simply a hypothet- ical box to visualize the aerosol's behavior. Here, local particle density is lower and particles spend more time interacting with their environment; air flow, temperature and moisture are important, as is the interaction with physical boundaries. In these sim- ulations, the aerosol initial conditions at the device nozzle (velocity, particle size, mass flow rate) may be pre-set or predicted from various physically-based or