Madnia's results are drawn from direct numerical simulations (DNS) used to dissect the burning vortex rings. They are possible only with a fully parallel computer code optimized for the SGI Origin2000 platform. The code performs very well—95 percent parallel efficiency is achieved in many of the simulation runs, according to Madnia. Due to the size and precision of the models, however, a typical simulation takes between 10 and 20 days running on 16 processors.

When a run is completed, the simulations created on the Origin2000 are compared to experimental results obtained by W. J. A. Dahm's group at the Laboratory for Turbulence and Combustion at the University of Michigan. There are some limitations to a model's ability to capture the physics of a realistic combustion process, Madnia says, but the upside is that "DNS allows us a degree of control in isolating specific physical phenomena that is just inaccessible in experiments."

This level of control lets Madnia and his team use their simulations to answer the researchers' basic questions concerning the key physical characteristics of flames. Madnia says his next step will be to identify just how detailed these simulations can become.

While Madnia and his team focus on the basic character of flames, the implications of their work may someday answer questions in combustion-engine design. Hopefully, he says, the models could be used to replicate the basic features and details of flames in complex combustion systems.

"The ultimate goal of understanding this is to come up with models that are ingenious and that people in industry can use," Madnia says. He adds that, while this is basic research, it is also fundamental. "This is something that you do in research. First find the problem, then understand it, then add more and more difficulty."

In a decade or two he hopes his work will have a significant effect on combustion engines. "You see, first we understand this problem. Then later we model the chemistry, and then the soot. Which of course is of a great importance in terms of pollution, and our atmosphere." This seminal research will allow engineers to look at models of how turbulence might influence the chemistry of a flame and how heat released from the flame might influence turbulence. Using this information, they will be able to change the parameters of different engine designs before the engines are produced.

This research is supported by the National Science Foundation and the Petroleum Research Funds administered by the American Chemical Society.

Team Members
Stefan Enachescu
Daniel Livescu
Cyrus K. Madnia
Cosmin Safta