Jennifer Franck’s first foray into computing was on the venerable, if rudimentary, Commodore 64. As a child, she tapped out simple looping programs that sent a series of numbers to her printer. Since those early days, Franck’s programs have gotten considerably more complex.
The new lecturer in engineering is an expert in computational fluid dynamics. She writes programs that simulate how fluids and gases flow around objects. Specifically, she codes what are called large-eddy simulations, a class of code designed to study turbulence. She mostly uses her model to investigate the dynamics of flight — how wind interacts with wings.
After earning her Ph.D. in mechanical engineering from Caltech in 2009, she came to Brown as a postdoc to work with Kenneth Breuer in engineering and Sharon Swartz in ecology and evolutionary biology, who are widely known for their research on the mechanics of bat flight. “What I was interested in was to see if I could explain some of the characteristics of animal flight using my models on the computer,” Franck said.
One of the questions Franck looked at is why bats flap their wings, as opposed to using them for soaring flight. “There’s a theory that bats evolved from passive gliders to actively flapping their wings,” she said. “The question was, what’s the benefit of flapping.”
Franck’s models helped to show that flapping creates vortices — tiny pockets of low air pressure — above a bat’s wings. Those vortices create extra lift and may be part of the reason flapping is worth the effort.
Franck has also used her models to explore applications that might improve aircraft flight. “Say you want an airplane to have more lift,” she said. “Could you apply some sort of device on the wing that would pump some extra energy into the flow and give you better performance? I’m interested in applying code to those types of flow control questions.”
There are significant advantages to the computational approach, Franck says. It’s much easier, for example, to modify the parameters of an experiment on a computer than it is to design new physical models for wind tunnel tests. Another advantage is that computer models help to isolate the specific aspects of a problem that researchers are trying to address.
“We generally model a very simple airfoil that’s often just two dimensional because it simplifies the problem,” Franck said. “If we’re looking at the basic physics behind a problem, we don’t want to make things too complicated.”
Though the models may be simple, the code that generates them is not. Most of Franck’s programs require computer clusters that string together multiple processors. For some of her research, Franck has used a cluster at Brown’s Center for Computation and Visualization. For other projects she’s used the Department of Defense’s Army Research Lab cluster in Maryland.
It’s a long way from the Commodore 64, but Franck is right at home. “Coding has always just come naturally to me,” she says.
She and her husband Christian, professor of engineering at Brown, live in Providence with their two kids.