Computational modelling of pulmonary drug fluidisation

A dry powder inhaler (DPI) relies on patient inspiration to fluidise drug particles for transport into the lungs. This study uses computational fluid dynamics (CFD) to develop a two-fluid model which focuses on this process. Closures are provided through the kinetic theory of granular flow (KTGF) where particle energy fluctuations are captured through granular temperature. Simulation results are compared to experimental evacuation data for Lactohale©206 across different inlet Reynolds numbers. During model development, a systematic analysis is conducted to understand how modelling parameters affect fluidisation. k-ε and k-ω turbulence closures were tested, with the k-ε RNG model most aligned to experimental data. In addition, effects of packing limit, coefficient of restitution and turbulence dispersion were studied. Results show inlet turbulence intensity and dispersion models have marginal effects on evacuation, whereas maximum packing limit significantly influences powder fluidisation. With optimum parameter values, the two-fluid model generates excellent agreement with experimental data for all Reynolds numbers. Additionally, the powder size distribution is studied. Results show no significant difference in evacuation time between monodisperse simulations of the dominant diameter and a polydisperse particle size distribution. Hence, when capturing fluidisation features, a simplified monodisperse simulation is sufficient, bringing a significant reduction in computational cost.