An experimental parametric study assessing the effect of a certain initial geometric imperfection on cylinder buckling

Abstract

Given the sensitivity of axially-loaded cylinder buckling to geometric imperfections, the advent of additive manufacturing provides a compelling platform to assess subtle changes in initial configuration from an experimental perspective. The cylindrical form is ubiquitous in many load-bearing contexts, from soda cans to submarines to rocket fuel tanks. Despite the relative simplicity of the shape, it is striking how the assessment of buckling loads continues to pose strong challenges to the designers of these load-bearing components, often in a situation where weight-saving is a key constraint. Furthermore, this type of buckling typically corresponds to a sudden, severe, catastrophic event, and in addition to posing serious consequences in practice, this also presents challenges for those conducting high-fidelity experiments, especially in terms of repeatability. Any conditions that deviate from a pristine cylindrical shape, and purely axial loading, are the root-cause of the buckling variability. Typically, the slight deviations from the pristine (perfect) conditions are somewhat random and unpredictable in nature. 3D-printing now provides the experimentalist with the ability to produce cylindrical specimens to a very high degree of geometric fidelity, thus allowing a degree of control of the form of subtle geometric imperfections. This paper focuses attention on a specific form of deviation from a pristine cylindrical shape, in which the longitudinal sides follow a mild cosine wave form. The amplitude of this shape is varied (in both directions, i.e., inwards and outwards), and its effect on both the buckling load and buckling mode shape assessed experimentally, using the outcome of finite element analysis as a guide.