Design and characterization of a Taylor–Couette device for studying factors affecting turbulence-induced hemolysis

Blood-perfused medical devices operate predominantly under turbulent flow conditions, associated with increased flow-induced hemolysis. However, current hemolysis models are based on laminar flow experiments and do not account for turbulence. This study establishes an experimental setup to quantify turbulence-induced hemolysis and enables the development of hypotheses regarding interactions between turbulent flow and erythrocytes. A Taylor–Couette-device with counter-rotating cylinders was developed, operating with 20 ml fluid volume and capable of generating, visualizing, and measuring laminar and turbulent flow properties at identical mean shear rates. Flow regimes were characterized using an image-based, custom-trained neural network to ensure consistent classification across fluids. Experiments were performed with water and water–dextran solution as mono-phase blood-analogue fluids at physiologic mean shear rates of 256, 396, and 537 s−1, with one laminar and three turbulent measurement points. Laser Doppler anemometry (LDA) provided flow characterization, while torque measurements at the inner cylinder were used to estimate energy dissipation. The system reliably achieved laminar and turbulent flow regimes. LDA measurements provided velocity profiles, turbulence intensities, and Reynolds shear stresses in good agreement with literature. The intended increase in turbulence intensity across the turbulent points was realized at 396 and 537 s−1 but not at 256 s−1. Torque-, wall shear stress-, and LDA-based estimates of dissipation and Kolmogorov length scales agree in order of magnitude, with a near-constant offset attributable to bearing friction and motor losses. The Taylor–Couette-device is suitable for controlled study of turbulence-related hemolysis. Higher mean shear rates will require a reduced gap width, and two-phase blood-analogue fluids should be considered in future studies.