Comments on effective use of numerical modelling and extended classical shell buckling theory

Abstract

This paper argues that in the understandable enthusiasm to use the now readily available high-fidelity numerical software allowing the analysis of the complexities of shell buckling, and especially the notorious case of the axially loaded cylinder, there has been insufficient consideration of classical shell buckling theory and its more recent reduced stiffness extension. If future design is to be able to effectively exploit the potential benefits of the use of advanced composite materials and/or the use of rib reinforcement, it is suggested that these numerically driven research programmes would benefit from more explicit integration within the framework of classically extended buckling theory. This suggestion is reinforced in what follows by highlighting the results from a longstanding research programme in which classically extended buckling theory has been demonstrably enhanced through its integration with reliable numerical experimentation. This enhancement has helped to resolve a number of important issues including: identification of the most critical forms of imperfection affecting the buckling loads; how the most critical knockdown factors are influenced by changes in the shells' geometric and material characteristics and how safe design might be more explicitly related to the prescribed levels of geometric tolerances. Recommendations are made as to how such integrated research programmes could in the future further enhance our ability to design materially efficient shells against buckling.