Purpose

Blood is commonly treated as single-phase homogeneous fluid in numerical simulations of blood flow within fiber bundles of blood oxygenators. However, microfluidics tests revealed the presence of hematocrit heterogeneity in blood flowing across such geometries. Given the significant role of red blood cells (RBCs) in the oxygenation process, this study aims to propose a multiphase blood model able to correctly describe the experimental evidence and computationally investigate hematocrit heterogeneities inside fiber bundles.

Methods

The experimental results of microfluidics tests performed in a previous study were processed and based on quantitative data of image intensity, a two-phase blood model following the Eulerian–Eulerian approach was calibrated and evaluated in its predictive ability against the experimental data. The two-phase model was then used to study the RBCs distribution inside different fiber bundles at average hematocrit values of 25% and 35%, representative of hemodilution in extracorporeal blood circulation.

Results

The numerical model proved to be able to describe and predict the experimental phase separation between plasma and RBCs within the microchannel geometry at different test conditions. Moreover, blood flow simulation in commercial fiber bundles revealed the presence of specific patterns in hematocrit distribution and their dependence on variations in bundle microstructure.

Conclusion

The two-phase blood model proposed in this study provides a tool for advanced evaluation of local fluid dynamics and identification of optimal bundle microstructure allowing further gas transfer simulations to account for a reliable heterogeneous distribution of RBCs around the oxygenating fibers.