On the Discretization of the Power-Law Hemolysis Model

Area of Science:

• Biomedical Engineering
• Computational Fluid Dynamics (CFD)
• Hemolysis Modeling

Background:

• Flow-induced hemolysis is a critical issue in blood-contacting medical devices.
• Accurate computational prediction of hemolysis can accelerate device development.
• Convergence of computational fluid dynamics (CFD) velocity fields is standard, but hemolysis calculation convergence also needs evaluation.

Purpose of the Study:

• To compare the convergence of the power-law hemolysis model across various flow conditions.
• To identify factors influencing the required number of timesteps for hemolysis convergence.
• To assess the accuracy of converged hemolysis predictions against analytical solutions.

Main Methods:

• Simulated simple flow scenarios: exponentially varying stress pathlines, Couette flows, sudden radial expansion, and the FDA channel.
• Analyzed convergence behavior of the power-law hemolysis model by varying timesteps.
• Compared converged hemolysis values with analytical solutions for exponential stress cases.

Main Results:

• Hemolysis convergence required one to tens of thousands of timesteps, dependent on stress exponent.
• Rapidly increasing (large exponent) and rapidly decreasing (small exponent) stress demanded more timesteps.
• Converged solutions showed potential errors exceeding 10% for both rapidly increasing and decreasing stress scenarios.

Conclusions:

• Hemolysis calculation convergence is essential for accurate CFD predictions in blood-contacting devices.
• Complex flow geometries, like radial expansions, can present challenging pathlines for convergence.
• The power-law hemolysis model may yield significant errors even with converged solutions under certain stress conditions.