Abstract
The Newtonian model has commonly been used to represent the viscosity of blood in the aorta, despite blood itself being a non-Newtonian fluid. This is justified where shear rates tend to be large. However, we hypothesized that using the Newtonian model to predict the hemodynamics on the aortic valve, particularly in those with severe calcifications, is inaccurate owing to valve leaflet geometry irregularities inducing multiple regions of low shear rates, <100 s−1, where a Newtonian model is invalid. We investigated the utility of three fluid viscosity models via quasi-static simulations: Newtonian, Carreau, and Quemada on a severely calcified aortic heart valve and compared their ability to capture important hemodynamic parameters of wall shear stress (WSS) and the oscillatory shear index (OSI). Our findings indicate that when the shear rates were large enough, >100 s−1, the use of a Newtonian model was justified. However, in spatial regions of relatively low shear rates, <100 s−1, specifically on the inner cusps of the fibrosa side of the valve, WSS calculations under a Newtonian model were found to be noticeably different when compared with their non-Newtonian, Carreau and Quemada counterparts. We hereby conclude that to facilitate more accurate computational flow simulations in severe aortic heart valve calcification, which is subjected to relatively large spatial regions of low shear (<100 s−1), a non-Newtonian model should be applied.