Multi-Laboratory Uncertainty Analysis of PIV-Measured Flow Quantities Relevant to Blood Damage in the FDA Nozzle Model

Particle image velocimetry (PIV) has been used in regulatory submissions to the FDA for pre-clinical and post-market evaluations of flow fields in medical devices, such as artificial heart valves, blood pumps, and stents. The velocity and shear fields obtained from the PIV experiments are also used to validate computational fluid dynamics (CFD) data accompanying the submissions. However, previous studies have questioned the accuracy of PIV measurements in regions of high shear and low velocity (regions prone to hemolysis and thrombosis). Currently, there is no clear estimate of the amount of uncertainty involved in measuring various flow parameters in these high-risk regions. The objective of this study was to perform an inter-laboratory PIV study in a simplified nozzle model and quantify the uncertainties involved in measuring flow quantities relevant to blood damage, such as near-wall velocity, viscous and Reynolds shear stresses, size and velocity within recirculation regions, and for estimating an index of hemolysis.