Understanding fluid flow in the human body is an important component in the prevention and treatment of disease. For example, detailed models of the human heart and blood clotting are used in the design of artificial heart valves, and similar models are needed for many other aspects of biological research. Simulation of these systems requires large-scale parallel machines, which are difficult to manage and program. To simplify programming, the Titanium group at Berkeley has developed a high performance dialect of Java for programming large-scale parallel machines, and they have an implementation of the human heart model in Titanium. The goal of the proposed research is to implement a second application within this framework and to analyze the performance of the systems and the generality of the framework.