“I have some ideas that will attempt to bypass some of that computational complexity that should let us make this analysis both computationally cheaper and also perhaps more insightful,” said Dawson.

Dawson said his approach to that simplification — using advances in numerical methods and machine learning — will provide a better way to understand, predict, and control turbulent systems with the hope of improving designs.

He plans to take advantage of the fact that, despite their tendency to be complex and chaotic, turbulent flows have some consistent structures and repeated patterns that they follow.

Dawson said his approach should be able to automatically detect which terms within the fluid dynamics equations are important for the generation of certain turbulent features and then single them out for a given problem that is being solved. Because the theoretical equations are well-established, this provides an opportunity to take a novel approach to using machine learning in the model by applying it directly to the equations with the aim of using it to determine which terms in the equations are relevant for a given problem.

“The idea is that we should be able to do this analysis without really needing a lot of high-fidelity data from the equations themselves,” said Dawson.

Dawson said he will be starting with simple systems and gradually introducing complexity as he gains confidence in his methods. He’ll begin with incompressible flows over flat surfaces and gradually build to compressible flows through pipes or ducts with corners and eventually to consider cardiovascular blood flows with more complex geometries, aiming to develop a modeling approach that can be used on a wide range of real-world applications.

“We’re at a point where we have a confluence of knowledge about turbulent flows because we have a lot more data and computational power that enables us to perform different analysis methods,” he said. “I think it’s an exciting time to be in the field.”

As part of this project, Dawson will also be conducting outreach to inform the community about engineering in general, fluid dynamics, and his project specifically.

“I’m looking to develop and implement educational initiatives that go hand in hand with the research,” he said. “I’ve got plans to incorporate some of the ideas I’m looking to develop within the curricula of my courses, and I have plans for broader outreach to school and community groups to increase STEM literacy more generally.”

The NSF’s Faculty Early Career Development (CAREER) program offers the foundation’s most prestigious awards in support of early-career faculty who have the potential to serve as academic role models in research and education and to lead advances in the mission of their department or organization. For more information, visit www.iit.edu.