Comparison of analytical and numerical particle deposition using commercial CFD packages: impaction and sedimentation

Whole-lung dosimetry codes and computational fluid dynamics (CFD) techniques have been used extensively to predict particle deposition in the respiratory tract of animals and humans. Although these predictions implement three well-known deposition mechanisms (impaction, sedimentation, and diffusion), validation of deposition due to each deposition mechanism in isolation has been difficult. In the current work, impaction deposition predictions using equations from the leading whole-lung dosimetry codes were compared to experimental data for the same Stokes and Reynolds numbers. In addition, impaction was predicted numerically using two commercial CFD packages (CFX and Fluent) and compared to experimental particle deposition, for the same geometry, and flow conditions that were overwhelmingly impaction dominated as measured by the Stokes number. Significant differences were found between CFD predicted deposition due to impaction and the analytical equations contained in whole-lung dosimetry models (NCRP, Trumpet, MPPD). Of the two CFD software packages, CFX typically had the best agreement with the experimental data; however, neither software package agreed well for all Stokes numbers examined. In addition, predicted impaction deposition from whole-lung dosimetry code equations did not agree well with experimental data for all Stokes numbers. These discrepancies highlight the current state of uncertainty in particle deposition predictions and indicate that any single technique or equation may be unsuitable to accurately explain the flow and particle behavior in an airway bifurcation.