Red blood cell partitioning at simulated bifurcations in physiologically realistic microvascular networks

Partitioning of red blood cells (RBCs) at vascular bifurcations is studied using direct numerical simulations of cellular-scale blood flow in physiologically realistic microvascular networks. The in silico networks are constructed following in vivo data, and are comprised of multiple bifurcations. Flow of deformable RBCs at physiological hematocrit is considered, and the dynamics of each individual cell are accurately resolved. The simulations predict the classic disproportionality in cell partitioning, as observed in vivo, where a daughter vessel with a higher flow fraction receives a disproportionately higher fraction of cells. We also observe a reverse partitioning where the branch receiving a higher flow fraction receives a lower fraction of cells. The cellular-scale mechanisms underlying the diverse types of partitioning are presented. The sequential nature of the bifurcations is shown to result in a positive or negatively skewed hematocrit profile leading to the classical or reverse partitioning. Two underlying mechanisms, namely the plasma skimming and cell screening mechanisms, are quantified. The plasma skimming mechanism is shown to under-predict cell partitioning, leaving the cell screening mechanism to make up for the difference.