What is the force needed to crush a soda can? Under what loading is a rocket shell guaranteed to never collapse? Surprisingly, there is no great answer to these seemingly basic questions. The predictions based on linear stability analyses strongly overestimate the critical loads observed in the experiments over the last 50 years. In real systems, defects weaken these thin-walled cylinders and cause them to collapse prematurely. As a result, it is quite difficult to guarantee their stability.
Traditionally, the role of defects in the stability of cylinders is investigated by measuring the linear instability of an imperfect shell. However, our fearless leader and collaborator in Switzerland, Tobias Schneider (EPFL) suggests a conceptually different approach. Schneider speculates that the fully nonlinear dynamics framework that he recently used to revolutionize our approach to the transition to turbulence is applicable to elasticity too. However, this requires studying the stability of perfect shells, which as we mentioned are not easy to find.
With the Schneider lab, we combine theoretical and numerical investigations with high-precision experiments to map out critical perturbations as a function of loading conditions. Our goal is to fully describe the buckling phenomenon as a nonlinear instability of a perfect shell, rather than as a linear instability. Experimentally, we uniaxially compress homemade polymer cylinders, as well as the most famous aluminum shells in the world (a.k.a. coke cans), while ‘gently’ poking them from the side. The poking triggers the collapse of the cylinder at relatively low forces, before any defect ‘kicks-in’, in such a way that we are indeed probing the nonlinear stability of a perfect shell.