This research will improve mathematical models and numerical methods for studying platelet aggregation during blood clotting and the flow of fluid-particle suspensions. These are large- scale problems of considerable complexity. Consequently, the research will involve a high degree of computation, much of which will require use of a supercomputer. A new particle-based algorithm for solving a multidimensional convection-dominated convection-diffusion equation will be developed. The use of localized forces to represent platelets and suspended particles will be extended to the representation of blood vessel walls. This work should shed light on a central question in aggregation: how the blood's local fluid dynamics and the geometry of the blood vessel interact to influence the location, rate, and extent of aggregate formation. It will aid in the design of nonthrombogenic artificial internal organs.
