Differences Between Blood and a Newtonian Fluid on the Performance of a Hydrodynamic Bearing for Rotary Blood Pumps

Abstract

Assuming that blood has a constant viscosity is a common practice when designing rotary blood pumps (RBPs), where shear stresses are generally higher than in the human body. This eases the design and allows numerical simulations and bench top experiments to be performed with Newtonian fluids. However, specific flow conditions may cause a change in cell distribution leading to an apparent lower blood viscosity. It has been observed that decreasing the vessel diameters and increasing flow velocities contribute to this effect. Because a hydrodynamic bearing operates under flow conditions following this pattern, it is important to verify whether this effect also takes place when this type of bearing is applied to a RBP. Because the operation of a hydrodynamic bearing depends directly on the fluid viscosity, a local change in cell distribution in the bearing gap can be reflected in changes in the bearing performance. In this work, a spiral groove hydrodynamic bearing was tested with porcine blood in a specially built test rig. The generated suspension force, cross flow, and bearing torque were recorded and compared with the reference response when using a solution of water and glycerol. Experiments with porcine blood yielded lower suspension forces, lower flows, and lower bearing torques than when using the glycerol solution. An explanation could be a lower apparent viscosity due to inhomogeneity of blood cell concentrations. Therefore, it is crucial to consider the effective blood viscosity when designing hydrodynamic bearings for RBPs and performing experiments.