CFD models of the bronchial airways with dynamic boundaries

Obtaining reliable CFD predictions of the bronchial flow that reflects the actual flow within a living lung requires the development of a deforming airways model, and the imposition of physiological subject-specific boundary conditions. This thesis addresses these two issues by the development of dynamic CFD models of the bronchial airways using a dynamic CT data set covering the breathing cycle of a laboratory animal. A deformation algorithm is proposed that matches the CFD mesh of the subsequent airway geometries generated from the dynamic CT data set. In addition, a novel nonlinear dynamic airway model generated from a pair of CT images is introduced. The proposed non-linear deforming model is capable of successfully capturing the non-linear motion characteristics of the bronchial airways based on the clinical measurements of the lung volume change. Furthermore, a technique to drive physiological subject specific boundary conditions for the terminal surfaces of the CFD models of the bronchial airways is introduced. The proposed technique depends on approximating the lung volume associated to each terminal surface over several time points over the breathing cycle based on the mechanical coupling between the bronchial airways and the vascular tree. The computed dynamic subject-specific boundary conditions were imposed on the terminal surfaces of the deforming airway model and the effect of wall motion on the flow features during tidal breathing is investigated for the first time. The outcome of this thesis is expected to improve the fidelity of the CFD predictions of the bronchial flow compared to the actual flow within a living lung. In addition, the availability of a new non-linear dynamic model of the bronchial airways that requires one pair of CT images as input, which complies with the radiation dosage restrictions for humans will facilitate the development of well-resolved CFD models of the human bronchial airways.