As we head deeper into the digital age, computers appropriate an ever greater share of the work of designing and testing physical systems, spanning the gamut from nuclear components to personal care products. Engine design is another important, but very complex, application that is benefiting from computational modeling.
To optimize the performance of combustion engines, chemical models are incorporated into computer simulations. These simulations incorporate millions of variables and require extensive uncertainty analysis. A just-announced simulation project, undertaken by Argonne’s Virtual Engine Research Institute and Fuels Initiative (VERIFI), will utilize 60 million computer core hours on Argonne’s Mira supercomputer to help resolve this uncertainty.
It’s a level of computational power access that is unprecedented for VERIFI, which describes itself as “the first and only source in the world to focus on high-fidelity, three-dimensional, end-to-end, combustion engine simulation/visualization, and simultaneous powertrain and fuel simulation, with uncertainty analysis.”
As the lab’s public information officer Greg Cunningham explains in a recent posting, VERIFI will be tasked with dispelling uncertainties related to the complex interplay of variables — these include temperature of cylinders and pistons, pressure inside the cylinder and flow rates of fuels — thus paving the way to more effective engine simulations.
Mira is a 10-petaflops IBM Blue Gene/Q system that is currently the fifth-fastest supercomputer in the world based on TOP500 rankings. This represents a big step up in computational power for the project and the participants expect this increased capability will ultimately help them to develop a better combustion model. With more powerful hardware, the Argonne team will be able to go from single variable assessment to exploring how all the variables interoperate to effect targets such as performance and emissions. Resolving these variables is likely to require about a million calculations.
“This has the potential to be pioneering work, because we haven’t seen anyone really trying to understand these boundary conditions, model parameters and uncertainties at this level of detail,” said Sibendu Som, principal investigator and principal mechanical engineer at Argonne’s Center for Transportation Research. “You really need access to these types of computing resources to resolve these questions.”
Cunningham writes that the primary focus of the project will be enabling a new generation of gasoline compression engines that have characteristics of gasoline and diesel engines. This kind of engine is notable for using compression at low-temperatures to ignite the fuel in the same manner used by diesels. The overall aim is to facilitate the design of more efficient engines.
This research, like VERIFI’s other work, is intended to help vehicle manufacturers and engine designers meet new fuel efficiency and emissions mandates in less time and with less investment. The research is funded by the Vehicles Technology Office of DOE’s Office of Energy Efficiency and Renewable Energy, and the results from the study will be made public. Findings will also be integrated into the CONVERGE engine simulation software, available from Convergent Science, Inc.