Aerospace structures must be lightweight yet highly flexible — able to survive launch loads, thermal cycling, and the vacuum of space, then deploy reliably in orbit. We pair the lab's scientific machine-learning methods with advanced finite-element mechanics to design and analyse deployable space structures and the composite materials they are built from.
Deployable structures — booms, reflectors, solar sails — must fold compactly for launch and reliably deploy in orbit. Tensegrities offer a compelling structural paradigm: lightweight, stiff, and deployable. We have developed the mathematics of stable tensegrity configurations and bring scientific ML surrogates to bear on the large-deformation dynamics that govern deployment — replacing repeated, expensive finite-element runs while respecting the underlying physics.
See Scientific Machine Learning for the methods behind these surrogates.
Composite materials are ubiquitous in aerospace — combining high stiffness and strength with low weight. Thin-walled open-section composite beams appear in aircraft spars, wind turbine blades, and launch vehicle structures. We apply variational asymptotic methods to derive dimensionally reduced models that retain the accuracy of 3D analysis at a fraction of the computational cost.