Hemolysis is a major concern in blood-circulating devices, which arises due to non-physiological stresses on red blood cells from ambient flow environment or moving mechanical structures. Computational fluid dynamics (CFD) and empirical hemolysis prediction models have been increasingly used for the design and optimization of blood-circulating devices. The commonly used power-law models for hemolysis prediction often use Reynolds stress to represent effective stress, tend to over-predict hemolysis and fail to capture trends of flow-related hemolysis. This study proposed a new power-law formulation for the numerical hemolysis prediction. The new formulation related hemolysis to the energy dissipation rate, which could be readily obtained from CFD simulations. The model constants were regressed from existing hemolysis data. The new formulation was tested for three benchmark cases and compared to conventional power-law models. The results showed that the new formulation improved prediction of hemolysis for a broad range of flow regimes. The deviations of the predicted hemolysis from experimental results were within one order, and better correlated with experimental results. This study confirmed that Reynolds stress is the main cause of over-prediction of hemolysis for conventional power-law models. Proportionally, the blood damage predicted with Reynolds stresses is more than one order higher than viscous stress, in terms of energy dissipation.
Keywords: Blood damage; Blood pump; CFD; Energy dissipation; Hemolysis; Reynolds stress.