Prof. Michael Burke Wins NSF CAREER Award

Feb 12 2020 | By Holly Evarts | Photo Credit: Jeffrey Schifman

Michael Burke, assistant professor of mechanical engineering, has been given the National Science Foundation’s (NSF) prestigious Faculty Early Career Development (CAREER) award for his project to design a new predictive computer model that will help create faster, cheaper designs of cleaner, more efficient engines. The NSF’s highest honor awarded to early-career faculty, the five-year $500K grant will support his proposal, “Extrapolatable, Uncertainty-Quantified Modeling of Nitrogen Kinetics Informed by Data Across Multiple Scales.”

Burke’s project will produce predictive tools for minimizing the formation of nitrogen oxides (NOx), which are responsible for smog, ground-level ozone, acid rain, and other conditions unhealthy for both people and the environment. With the NSF grant, he will create and validate a predictive model for nitrogen oxides (NOx) formation during combustion. As part of the project, Burke is partnering with industry colleagues at Siemens to enable his predictive tools to lead to better engine designs right away.

“These new predictive tools will be immensely powerful for engine design,” says Burke, who is an affiliated professor of chemical engineering and a member of Columbia’s Data Science Institute. “It’s critical that design engineers have models that can make accurate predictions with known uncertainty so they can accurately assess engineering risk, especially for new cutting-edge engine designs that have not yet even been tested or manufactured.”

Combustion, a chemical process essential to all engines, from gas turbine engines used for electricity to piston engines used for transportation, plays a large role in our energy landscape and it is clear that future combustion technologies will need to use fuel more efficiently, produce fewer emissions, and operate on a wider range of fuels, including alternatives. Predictive modeling could enable faster, cheaper designs of cleaner, more efficient engines our planet so urgently needs.

Burke, who explores a variety of problems at the interface between fundamental physical chemistry and practical engineering devices, is taking an innovative approach to this project. He will leverage both modern uncertainty quantification and computational chemistry to select, create, and exploit multiscale data that span from the molecular to the macroscopic, and design predictive models with known uncertainty.

His approach is to pinpoint the exact molecular and macroscopic data that are most needed to reduce uncertainties, generate those molecular data via computational chemistry and macroscopic data via experimental measurements, and use those data to create models with lower uncertainties. His models will form the backbone for future studies of NOx formation during combustion of all conventional and alternative fuels and will lead to improvements in engineering control of NOx pollution.

Mitigating the effects of NOx to meet increasingly strict regulations has been a major constraint in engineering design. In fact, the difficulties in limiting NOx formation while maximizing fuel economy led to the 2015 Volkswagen emissions scandal.

“A major challenge in mitigating NOx in engines is that its formation is least understood at the high pressures and low peak temperatures of modern engines,” says Burke. “We think there could be important pathways to forming NOx in engines that are missing from current NOx formation models.”

Burke will address the key outstanding issues in the current understanding of NOx formation at high pressures and will produce the first uncertainty-quantified NOx kinetic model constrained by multiscale data. As part of this, he will examine previously undiscovered pathways that he hypothesizes to comprise a major NOx route at the high pressures and low peak temperatures of high-efficiency, low-NOx engines.

Burke also expects his new rate laws and mixture rules that describe pressure dependence in reactive mixtures – another outcome of the project – will improve predictive design tools for cleaner, more efficient engines and, more broadly, understanding fundamental gas-phase kinetics phenomena affecting planetary atmospheres, including the Earth’s atmosphere and climate modeling, chemical weapons destruction, and hypersonics.

“In addition,” says Burke, “our scientific findings for NOx will contribute to reducing NOx pollution and its harmful health and environmental impacts and, more generally, improving understanding of nitrogen kinetics relevant to biomass combustion, renewable fuels, and Earth’s nitrogen cycle.”