Aim

This research broadly aims to better understand the nature of thin shell conical roof behaviour.

Principle Objectives

To relate variable ring area to idealised pinned and roller boundary conditions for a variety of relevant roof geometries. 
  • One can imagine a given conical roof and applied load supported by a continuum of boundary conditions ranging from no lateral restraint (roller supports, or no ring) to full lateral restraint (pin supports, or an infinite ring) result in a corresponding continuum of pre-bifurcation structural response, as measured in membrane stress resultants.
  • It is hypothesised that peak membrane stress resultants plotted against ring area (on a log scale) will produce an "S" curve, where the resultants associated with very large rings asymptotically approach the pinned boundary solution.

To obtain predictions of linear bifurcation buckling loads and stress resultants for varying roof geometries under uniform external pressure.
  • Similarly to stress resultants, it is expected that changing any one of the geometry variables will alter the structural response in terms of buckling behaviour as well. However, it is not expected that buckling behaviour will follow the same analogue continuum response behaviour. For example, circumferential buckling mode (number of waves around the circumference) must be a natural number. Then for some change in a geometric variable, there may be no change in buckling mode! The results of a study into this behaviour are proving hard to hypothesise at the moment.
To characterise all analysed relationships such that they may be used in design.
  • It is hoped that this research will lead to a better understanding of thin shell conical roof behaviour, such that any suggested improvements to design standards that may result from this research appeal to the reader's inner scepticism and inner Occam's Razor. This will hopefully be achieved by including all necessary variables (notably the addition of realistic boundary conditions.