A Computational Fluid Dynamics Model to Predict Turbulent Flow Damage in Medical Devices

Turbulence is a random, highly three dimensional, chaotic, irregular and multiscale flow condition that results in strong vorticity, high rate of mixing, with pressure and velocity variations occurring over a wide range of time and length scales. [1] Rapidly moving impeller blades of heart devices cause highly disturbed turbulent flow in/near these devices. It is commonly accepted that turbulence contributes to RBC damage. One of the most common and widely studied effects is hemolysis. While an increase in hemolysis is observed when cells are exposed to turbulent stresses, [2] the structure of turbulence in proximity to the blood cells and the fundamental mechanism by which cells are injured remain unclear. [2-4] Therefore, understanding and predicting the effect of turbulent stresses on erythrocytes is a major concern when designing prosthetic heart devices. [2-4]. In this work, a computational technique (using the CFD software FLUENT) that employs Reynolds-Averaged Navier-Stokes models of turbulence (k-ε and k-ω SST models) to simulate the classic Couette flow, the capillary flow and the jet flow hemolysis experiments was used. The simulations predicted the rate of dissipation of turbulent kinetic energy and the size of corresponding turbulent eddy structures. Idealized structures were assumed to be spherical and to have sizes that correspond to the KLS of the flow at different locations of the devices. The study investigated the relation between the surface area of the KLS structures and predictions of cell damage following a procedure detailed elsewhere [5]. Extensive measures like total eddy areas for specific KLS were used to provide a prediction of hemolysis. Depend on the results of three systems, regression analysis was performed and a new model was proposed to predict hemolysis in turbulent flow that takes into account the complexity of turbulence by giving varying weight to eddies of different sizes. Results showed that dissipative eddies comparable or smaller in size to the RBCs are related to hemolysis, and that hemolysis corresponds to the surface area of eddies that are associated with KLS smaller than about 10 µm [5]. This is roughly the length scale of an RBC. The model was applied to results from three systems. Reasonable agreement has been obtained for flow fields and exposure times of these three distinctly different experiments. Predictions are also to be compared to the experimental findings on release of a centrifugal blood pump that has been identified as a case study through the critical path initiative of the U.S. Food and Drug Administration (FDA) [6].

References

[1] Pope, S. B., 2000, Turbulent flows, New York, USA.

[2] Kameneva, M. V., Burgreen, G. W., Kono, K., Repko, B., Antaki, J. F., and Umezu, M., 2004, "Effects of turbulent stresses upon mechanical hemolysis: experimental and computational analysis," ASAIO J, 50(5), pp. 418-423.

[3] Aziz, A., Werner, B. C., Epting, K. L., Agosti, C. D., and Curtis, W. R., 2007, "The cumulative and sublethal effetcs of turbulence on erythrocytes in a stirred-tank model," Ann. Biomed. Eng., 35(12), pp. 2108-2120.

[4] Bludszuweit, C., 1995, "Three-dimensional numerical prediction of stress loading of blood particles in a centrifugal pump," Artif. Organs, 19(7), pp. 590-596.

[5] Ozturk, M., O'Rear, E. A., and Papavassiliou, D. V., 2015, "Hemolysis related to turbulent eddy size distributions using comparisons of experiments to computations," Artif. Organs, 39(12), pp. E227-E239.

[6] US Food and Drug Administration, 2013, "Computational fluid dynamics: an FDA Critical Path Initiative project," https://fdacfd.nci.nih.gov.