Development of computational mass and momentum transfer models for extracorporeal hollow fiber membrane oxygenators

摘要

Simulation of mass and momentum transfer in hollow fiber membrane (HFM) oxygenators has remained a topic of interest for decades. The current work reports upon efforts toward modeling the transport phenomena within these devices using a computational fluid dynamic (CFD) approach. The results and findings form a basis for future efforts in computational model development, design refinement and the investigation of other HFM systems.The main purpose of the present work was to determine the validity and applicability of the Darcian porous media model for three-dimensional flow simulations of a commercial membrane oxygenator design. Close agreement between experiment and simulation was found for pressure losses within the device, although not within the standard deviation of the experimental data. The divergence could be attributed to the governing equations, which incorporated the Darcian model parameters into a less restrictive model, or the model parameters themselves, which were assumed isotropic and uniform.The porous media model developed was used to predict the superficial velocity field of the membrane oxygenator during operation. For model verification, the experimental and CFD-predicted flow fields were compared using optical flow calculations performed on radiographic and simulated radiographic images, respectively. The data obtained were adequate for qualitative but not quantitative comparison and suggested reasonable agreement between experiment and simulation. Methods for improving the data and the interpretation were also described.It was hypothesized that appropriate CFD models would predict the local oxygen concentration and gas exchange in a HFM oxygenator. A mass transfer correlation for a non-reactive fluid was derived using parameters from the literature. The correlation was converted to a reactive form for blood through dimensional analysis and incorporated into a CFD simulation. The results from these simulations and an analysis of the experimental process indicated that the parameters used are subject to error when data collection is performed in a limited flow range.In conclusion, CFD-based simulations hold promise for HFM evaluation and design. Future researchers should consider the appropriateness of models and parameters during simulation development. Proper validation techniques are also critical for model assessment and evaluation.