ABSTRACT
Implantable ventricular assist devices are an effective temporary solution to heart failure, especially in the context of long wait times for heart transplants. With further progress, these devices may provide a permanent clinical cure for end-stage heart disease. Traditionally, device designers have relied on experimental analysis, and it is only recently that computational fluid dynamics has emerged as a cost-effective and reliable design tool. We present here a model for hemolysis prediction, designed to be used with computational fluid dynamics analysis of any blood-handling device. The model is based on experimentally measured physical properties of human blood cells, and hemolysis measurements in simple shear flows. An implementation of the model and numerical predictions of hemolysis are presented in the context of an on-going computational analysis of the GYRO blood pump being developed at Baylor College of Medicine. The numerically predicted hemolysis index matches well with experimentally measured values.