Date of Award

January 2025

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Chemical Engineering

First Advisor

Gautham Krishnamoorthy

Abstract

Hemolysis, the rupture of red blood cells (RBCs), is a critical concern in blood-circulating devices, requiring accurate predictions to ensure patient safety. This study investigates the sensitivity of hemolysis predictions to variations in viscosity models and hemolysis coefficients under complex flow conditions. Computational Fluid Dynamics (CFD) simulations were conducted using Ansys Fluent and OpenFoam on identical meshes with the FDA benchmark nozzle model. Under laminar flow conditions, CFD frameworks accurately predicted global variables, but variations in derived quantities, such as strain rate and vorticity, emerged due to differences in numerical solvers and gradient evaluation methods. These variations affected predictions of blood damage and non-Newtonian flow behavior. To assess this, blood properties, including flow symmetry indices, vortex characteristics, and hemolysis—were evaluated using Newtonian and four non-Newtonian viscosity models (Casson, Cross, Power Law, and Carreau-Yasuda). The Normalized Index of Hemolysis (NIH) vs. Reynolds number analysis highlighted that hemolysis is significantly influenced by viscosity model choice. At low Reynolds numbers, models with higher viscosity at low shear rates (e.g., Casson, Carreau-Yasuda) predicted elevated NIH values, indicating increased hemolysis risk. At higher Reynolds numbers, model predictions converged, reducing NIH variations. Furthermore, absolute values of NIH were very sensitive to the empirical power law coefficients employed, highlighting the need for updated coefficients for accuracy. These findings emphasize the importance of viscosity model selection and hemolysis power law coefficient accuracy in blood damage predictions. Optimizing these parameters is essential for improving CFD-based hemolysis models and minimizing RBC damage in medical devices.

Download

Share

COinS