Computational Issues in the Simulation of Blood Clotting

Elijah Newren, University of Utah

Computational Biofluid Dynamics (CBioFD) problems such as blood clotting are very challenging. They involve complex flows; interactions between flow and moving, deformable objects (such as red blood cells or platelets); chemical reactions occurring within the fluid and on cell surfaces; chemical and cell tranport; and chemically induced phase transitions (polymerization). Models of these phenomena typically involve coupled nonlinear PDEs, dynamic fluid-structure interactions, and complicated networks of kinetic equations for chemical reactions.

Blood coagulation involves a long series of enzymatic reactions containing many chemicals. These reactions involve feedforward and feedback loops and display thresholding behavior. As our understanding of the dynamics of this system is poor, it is important to simulate and analyze models to determine which include the relevant physiological behavior (which, unfortunately, seems to exclude all but the most elaborate models).

There are a number of tools and methods which need to be utilized to simulate such CBioFD systems. Algorithms must be chosen and implemented which are fast and scalable on distributed memory systems. Methods must also account for the highly local nature of the fluid-structure interactions and for the clumping of cells and chemicals in certain regions of the domain.

Abstract Author(s): Elijah Newren