ARGONNE, Ill. – Every science experiment and every mathematical model faces the same challenge: uncertainty.

In particularly complex systems like engines, these uncertainties can mount and multiply. From fuel injectors to combustion chemistry, every part of the process of combustion has some level of uncertainty associated with it. These uncertainties are extremely difficult for scientists to measure.

“There are lots of unknowns that are involved,” said mechanical engineer Sibendu Som of the U.S. Department of Energy’s (DOE) Argonne National Laboratory. “We’re looking at how all of them interact to affect overall uncertainty.”

Scientists at Argonne, as part of the new Virtual Engine Research Institute and Fuels Initiative (VERIFI), are looking at a number of parameters in the internal combustion process. These include the relationships between the diameter of the nozzle in the fuel injector, the dynamics of the fuel spray, the proportion of fuel to air in the combustion chamber, and the exhaust. By gaining a better understanding of how these parameters interact, the Argonne researchers seek to create cleaner and more efficient engines.

Overall, Som and Argonne colleagues Yuanjiang Pei and Michael Davis have investigated 32 different parameters simultaneously, trying to establish how the uncertainties vary under different conditions. “If we can find a way to reduce uncertainty, we can take a step toward developing a more predictive simulation,” Som said.

The process is known as uncertainty quantification. It’s been around for several decades, but scientists had never applied it to engine dynamics to such an encompassing extent. Previous analyses of uncertainties have mainly examined the chemical process separately.

“This is the first time we’ve applied these methods in such a complicated system,” said Argonne mechanical engineer Doug Longman. “This statistical tool has been around for a while, but we’re applying it in a brand new way.”  

VERIFI researchers are taking an iterated approach in which data gathered from the simulations can be fed back to both engine modelers and combustion chemists to further reduce uncertainty and create more predictive engine simulations.  By taking advantage of the incredible computational power available today, the VERIFI team will then identify the most important engine and fuel parameters.  This unique engine simulation and analysis will enable optimized engine combustion at any operating condition.

In the near future, the VERIFI team plans to run diesel engine simulations of unprecedented scale on Mira, Argonne’s 10-petaflop IBM Blue Gene/Q supercomputer.

Funding for this work is provided by DOE’s Office of Energy Efficiency and Renewable Energy.