Modeling of aerosol deposition with interface devices.

Various approaches can be used to mathematically model the performance of different masks, mouthpieces, and aerosol delivery devices. The sophistication of such models can vary widely, from the use of simple algebraic empirical correlations to advanced computational fluid dynamics simulations. Bench-top testing is also often used to model aspects of devices, since it is difficult to capture certain aspects of device behavior with mathematical models. These various approaches to modeling differ in their limitations. Empirical correlations exist for predicting the effects of varying mouthpiece diameter and mouth-throat dimensions on extrathoracic losses, but are restricted to stable, nonballistic aerosols in certain flow rate ranges. Computational fluid dynamics (CFD) simulations that solve the Reynolds-averaged Navier-Stokes (RANS) equations typically require near-wall turbulence corrections in order to adequately model mouth-throat deposition, while Large Eddy Simulation (LES) removes this deficiency. Bench-top models that use replicas of the extrathoracic airways vary in their accuracy and generality in replicating the filtering properties of these airways. Choosing and using these various modeling approaches for evaluating patient-device interfaces requires knowledge of their merits and pitfalls, a brief discussion of which is given here.