Remember the pre-pandemic days when travel was possible? As he pursued dissertation research for a Ph.D. in Civil & Environmental Engineering, Clay Sanders went to Paris last year to study a new method of solving “topology optimization” problems in structural designs.
Working with the POEMS (Wave Propagation Mathematical Analysis, and Simulation) team at ENSTA Paris Tech, Sanders researched design optimizations that would determine the best structural design option prior to construction.
This opportunity provided Sanders with a significant component of his dissertation work and allowed him to explore other interests in art, architecture, and structural design. He was among 11 Duke students who received 2019-2020 Graduate Student Training Enhancement Grants (GSTEG) from the Office of the Vice Provost for Interdisciplinary Studies. His faculty mentor was Wilkins Aquino.
A summary of his GSTEG experience is excerpted below.
I utilized my GSTEG for a research trip in June 2019 to ENSTA Paris Tech to investigate a new computational optimization technique to design structures. I worked with Professor Marc Bonnet, a researcher at ENSTA-Paris Tech, a small engineering university in Palaiseau, France, outside Paris. Professor Bonnet is a leader of the POEMS research group, which specializes in numerical methods to simulate wave propagation and solve physics-based optimization problems.
Topology optimization describes a class of structural design problems that seek to determine the optimal shape or form a structure so that they exhibit superior performance with respect to a performance metric. A common example would seek the optimal shape of a bridge, under a maximum weight constraint, to have maximum stiffness.
Our new approach, known as the “adaptive eigenspace basis method”, borrowed from computational techniques used to solve medium imaging problems for ultrasound or geological imaging applications. We showed that our new method could equivalently represent designs usually parameterized by thousands or millions of design variables with only a few dozen variables, enabling significant computational efficiency improvements.
Following the GSTEG trip, we refined the method and recently submitted a manuscript on the work to the International Journal of Numerical Methods in Engineering.
Beyond the research work conducted, I was able to explore Paris’s sites, and tastes, throughout my trip. ENSTA-Paris was only a short train ride outside of Paris, so I was able travel into the city each evening to explore the city. Other highlights of my trip included viewing Monet’s Water Lilies at the Musée de l’Orangerie, roaming the sculpture gardens at the Musée Rodin, sketching in the Luxembourg Palace gardens, visits to the Musée d’Orsay and the Louvre, and stops in as many Parisian pâtisseries as I could find.
Learn more about Graduate Student Training Enhancement Grants (GSTEG), see other 2019-2020 grantees and learn who received grants for Summer 2020.