Computational fluid dynamics simulation of airflow and aerosol deposition in human lungs

Computational fluid dynamics (CID) simulations of airflow and particle deposition in geometries representing the human tracheobronchial tree were conducted. Two geometries were used in this work: (1) based on the Weibel A model, and (2) based on a CT scan of a cadaver lung cast. Flow conditions used included both steady-state inhalation and exhalation conditions as well as time-dependent breathing cycles. Particle trajectories were calculated in each of these models by solving the equations of motion of the particle for the deterministic portion of particle displacement, and adding a stochastic Brownian term at each step. Me trapping of particles on the wall surfaces was monitored, and the locations of trapping in each generation were recorded. The results indicate that there are dramatic differences in the predicted deposition between the two models. The intragenerational. deposition locations show that in regions where the deposition mechanism is inertial impaction, the predominant deposition seems to be at the airway bifurcations. The results of this study suggest that under most conditions, an idealized model based on the Weibel dimensions is not sufficient to predict deposition, and an accurate model, such as those based on imaging techniques may be required. (C) 2003 Biomedical Engineering Society.