The reason is that in turbulent combustion, the difficulties of modeling turbulence -- by itself one of the most challenging problems in physics -- are further compounded by the complexities of strong, non-linear interactions between turbulence and the chemical reactions that occur during combustion.

Using a computational method called direct numerical simulation, Cyrus K. Madnia, Ph.D., associate professor of mechanical and aerospace engineering at the University at Buffalo, and his colleagues in UB's Computational Fluid Dynamics Laboratory have performed simulations that are the closest to date to a true model of the physics of chemically reacting turbulent flows.

The work, published in the current issue of the Journal of Fluid Mechanics, comes closest to mimicking the turbulent reacting flows that occur in hydrocarbon combustion without taking into account complex chemistry.

Knowledge of how these turbulent flows affect internal combustion could greatly improve the efficiency and environmental impact of all kinds of engines.

"In reacting turbulent flows, this will be the benchmark," said Madnia, lead author on the paper.

The simulations by the UB researchers come the closest to mimicking hydrocarbon combustion since they demonstrate a defining feature of combustion: how the heat of reaction affects the exchange of energy in a turbulent system.

In every engine or furnace, combustion causes the release of heat, which results in an increase in internal energy. At the same time, the turbulence in the engine is producing kinetic energy.

"Between internal and turbulent kinetic energy, there is a continual exchange," said Madnia. "How these two fields exchange energy with each other constitutes the basic physics of turbulent reacting flows."

Madnia cautioned that the flows that he and colleagues have simulated do not account for the incredibly complex chemistry of combustion, which involves hundreds of chemical reactions.

"Several decades from now, we still will not be able to simulate internal combustion," remarked Madnia. "But with this research, we are pushing the limit of direct numerical simulations for simulating turbulent combustion."

Unlike the group's previous work with nonreacting flows, where the researchers could "check" their results against those of laboratory experiments, these simulations cannot be compared with laboratory experiments because none exist.

According to Madnia, the goal of the research is to gain a better understanding of the two-way interaction between chemistry and turbulence in order to develop more realistic models of the fluid mechanics and chemistry involved in combustion.

"With this work, we have helped push that frontier a bit further," he said.

Co-authors on the paper are Daniel Livescu, Ph.D., formerly a doctoral candidate with the UB Department of Mechanical and Aerospace Engineering who now is at Los Alamos National Laboratory, and Farhad A. Jaberi, Ph.D., formerly a post-doctoral researcher at UB who is an associate professor of mechanical engineering at Michigan State University.

This work is sponsored by the National Science Foundation and by donors to the Petroleum Research Fund administered by the American Chemical Society. - By Ellen Goldbaum


[Contact: Ellen Goldbaum]

10-Jan-2002