Justin Sirignano, James Allison join new DOE center to research hypersonic engines

Zack Fishman

ISE professor Justin Sirignano brings his expertise on machine learning to a new research center, funded by the U.S. Department of Energy, to help develop a new engine that can travel several times the speed of sound.

The Center for Exascale-Enabled Scramjet Design is a collaboration of about 20 engineers and scientists to research scramjets, a type of engine that can reach hypersonic speeds, or greater than Mach 5 (3700 mph). The center has received about $17 million in funding from the DOE.

A scramjet, which stands for “supersonic-combustion ramjet,” is an engine that can propel vessels efficiently at speeds above Mach 6. All engines need to compress incoming air to burn fuel more effectively — but rather than use compressors and turbines like standard jet engines, ramjets simply fly faster than Mach 3 to compress the air to subsonic speeds. Scramjets work similarly but allow air to stay supersonic throughout the engine, allowing for an upper speed limit of more than Mach 10.

Engineers from around the world have been researching and developing scramjets since the 1950s. Although recent programs have had briefly successful experiments, such as the Indian Space Research Organization’s five-second scramjet test in 2016, it has not yet seen wide deployment. Sirignano says some countries are interested in developing hypersonic technology for applications in military and national defense, such as missiles.

“The United States is in a bit of an arms race in terms of hypersonic flight with places like Russia and China,” Sirignano says. “The government is investing a lot of money into it, so it’s a pretty important technology.”

For his part, Sirignano will work with physicists to run simulations of scramjet prototypes, which avoid the prohibitive cost of real-world experiment — up to $100 million per test. It will be the first time machine learning techniques are used to model engineering systems, and Sirignano says the new application is promising for future research.

“My interest would be that machine learning can be used to design complex engineer systems, and the scramjet is one example of that,” he says.There’s promise, and that’s why people are very interested.”

ISE professor James Allison, who joins Sirignano at the new center, says the scramjet research will use machine learning to increase simulation times and create designs that would otherwise be inaccessible. It will also implement advanced design frameworks to create optimal designs by understanding how one design decision affects others within a simulation.

Recent News