Oklahoma State University
Computational Fluid Dynamic (CFD) models were employed for lung aerosol dynamics studies, which can provide in-depth knowledge based on the natural laws of physics in a noninvasive manner. However, existing CFD models assumed that the lung airway walls are rigid, which limited the modeling of the lung compliance effect on airflow and particle transport dynamics. Thus, determining real-time lung deformation was hampered during airflow and particle transport simulations. Additionally, due to the imaging resolution limits of CT/MRI scans, the reconstructed human respiratory system can only reach generation 6 (G6). Therefore, it was necessary to develop an elastic lung model to simulate airflow and pressure distributions, as well as the respirable particle dynamics in human respiratory systems containing more generations.
To address the knowledge gaps mentioned above, the Computational Biofluidics and Biomechanics Laboratory at Oklahoma State University (CBBL@OSU) developed a unique CFD-based elastic lung model, which can accurately simulate anisotropic lung motions simultaneously with the transient airflow velocity and pressure fields in human respiratory systems. Specifically, integrating one-way coupled fluid-structure interaction (FSI) with the SST Transition model in ANSYS Fluent and ANSYS Mechanical, we successfully simulated the transient laminar-to-turbulence transition airflow patterns in a representative respiratory system in the CBBL virtual human system. Such an approach also has an optimized balance between computational accuracy and efficiency.
It is also worth mentioning that the representative human respiratory system employed models lung airways in unprecedented detail, covering the conductive region from nose and mouth to generation 13 (G13). The virtual respiratory system configuration was reconstructed with realistic anatomical characteristics using a stochastic algorithm, realized in SOLIDWORKS 2015 and Materialise® Mimics 19.0 and 3-Matic STL 11.0.
The elastic lung model presented here provides a generalized and computationally efficient method to utilize pulmonary function test (PFT) results for lung motion characterization. The model is also ready to be integrated into our well-validated Euler-Euler and Euler-Lagrange models, to enhance the fundamental understanding of how realistic lung movement can affect inhaled particle deposition predictions. The model will be further developed into a unique in silico study tool to provide high-resolution digital evidence of subject-specific pulmonary drug delivery, thereby advancing personalized medicine and precise pulmonary healthcare.