Computational particle hemodynamics in the rabbit abdominal aorta

The goal of this research is to more fully characterize the fluid mechanics in the abdominal aorta of the rabbit and to explore possible correlations between various hemodynamic wall parameters and early atherosclerotic lesion development. The flow fields are simulated computationally using Newtonian and Quemada models for whole blood with the geometries based on aortic castings. The resulting disturbed flows are characterized by the velocity fields, fluid element pathlines and hemodynamic wall parameters such as the wall shear stress (WSS), oscillatory shear index (OSI), wall shear stress gradient (WSSG), wall shear stress angle gradient (WSSAG), and the normalized deposition fraction (chidf). Near the aorto-celiac junction, the influence of non-planarity is systematically investigated by examining three geometries: a planar bifurcation with assumed mid-plane symmetry, a planar aorta with a celiac branch incorporating a transverse branching angle, and a non-planar geometry that includes aortic curvature as well as an out-of-plane celiac branch. The inclusion of non-planar geometric features imparts a significant bias to the secondary flows and the hemodynamic wall parameters spreading them further from the junction region. The non-planar celiac junction is a subset of a complete abdominal aorta that includes the celiac, superior mesenteric, left and right renal arteries. Resting, post-prandial and exercise conditions are simulated. For resting conditions, the secondary flows lead to a simulated deposition pattern that traces a helical path down the aorta and qualitatively matches the observed formation of a "ventral streak." In the inertia dominated flow through the aorta, fluid element pathlines are found to approximate particle trajectories computed from the particle momentum equation very well. Likewise, the inclusion of stochastic forces modeling red blood cell interactions are found to be a secondary effect. Near the celiac branch, focal increases in permeability to low density lipoproteins (LDL) and intimal white blood cells (WBC) correspond to local elevated values of the WSSG, WSSAG and chidf; the WSS inversely correlates with the regional frequency of LDL permeable sites suggesting a possible atheroprotective mechanism for local increases in shear stress. In the complete aorta, segmental averages of the WSSG, WSSAG and chidf correlate with intimal WBC densities again suggesting a possible role in focal increases in permeability. The OSI inversely correlates with the frequency of mitotic endothelial cells (EC) and the WSSAG nearly correlates which, when combined with other results in a backward-facing step, suggests that changes in the angle of the shear stress vector unscaled from the magnitude of the shear stress may also regulate EC division in disturbed flows. Regional distributions of chidf around the branch orifices correlates with both the frequency of horseradish peroxidase permeable sites and the distribution of intimal WBC. Regional averages of the WSSG and WSSAG can correlate with the distribution of intimal WBC. The merits of the Lagrangian hemodynamic wall parameter, chidf, are demonstrated by these results complimenting the shear stress-based hemodynamic wall parameters and as the only one capable of describing gross permeability patterns not associated with a junction region.