Implementation and Comparison of Non-Newtonian Viscosity Models in Hemodynamic Simulations of Patient Coronary Arteries

Abstract

Realistic numerical simulations of blood flow in patient-specific coronary arteries constitute a challenge in the study of hemodynamics. Several blood viscous models are available; yet their direct comparison has not been carried out. This work innovates by programming and implementing six viscosity models (Carreau, Carreau-Yasuda, Casson, Cross, generalized power-law and power law) as user-defined functions in ANSYS® Fluent to compare the major hemodynamic parameters, the time-averaged wall shear, the oscillatory shear index and the relative residence time, and to evaluate atherosusceptibility in coronary arteries. The study used the left coronary arteries of an apparently healthy patient and an unhealthy patient, with 40% stenosis in the left anterior descending. Flow simulations comprised two sets (steady-state and pulsatile flows), each based on Newtonian and non-Newtonian fluid, i.e., a total of four cases. Results indicate that the Casson model originates larger areas of atherosusceptibility and the generalized power-law models returns the most healthy results. The non-Newtonian pulsatile cases show less critical areas than the remaining studied regimes.