Freely readable over here: dl.acm.org/doi/pdf/10.1...

Fun collab, with a fair share of geometry, fabrication and computer graphics! Multilayered pneumatic chambers open up numerous avenues: 3+ networks, non-developable, differential pressurization… still lots of mechanics to be understood and geometries to be programmed!

With computer geometry wizards Ofir & Mélina, we also introduce a purely geometric inverse design algorithm. Usable to program any developable geometry, without any costly optimisation!

Staggering the networks while tuning the widths of alternating tubes allows for a precise control of the curvature of the inflated object through differential contraction. It’s just like a bilayer, but continuously programmable!

Through successive soldering, we managed to fabricate two distinct, bound networks of tubes. Our new inflatables are far stiffer than usual thin-sheets inflatables due to their homogeneous thickness — no free rotation along the welding lines!

Last insight: when we design objects under tension such as inflatables we often disregard bending stiffness under the assumption of large tensile stress. But this junction angle is constant and very much macroscopic! Bending stiffness matters regardless of the loading!
www.pnas.org/doi/10.1073/...

Moreover, we discuss the rotationnal behavior of the junction, there is some cool mechanics with a “universal” behavior in that discussion.

(bending stiffness is usually, like, really hard to measure precisely for thin/complex membranes and sheets)

Since the shape scales as a function of the tensile force in the system, the junction angle between the two rods does not change regardless of the loading. We show how we can harness this cool property for a precise, purely visual measurement of bending stiffnesses!

In it, we show how this junction is a combination of two self-similar, classical solutions of Euler’s Elastica. The shape of this boundary layer is determined by the (a)symmetry of the bending stiffness