Empirical equations have been adopted widely and primarily in the design codes due to their benefit of small amount of input data required and their convenience of application. However, these equations are limited to a small number of section design shapes. For example, the empirical equations in AS4600 (the design code for cold-formed steel structures) are derived primarily from channel and zed sections. Therefore, inaccurate results may be obtained if these equations are applied to other sections. Due to this limitation and for ease of design, numerical methods are often implemented and used in computer programs. The adoption of the numerical methods allows complex problems to be analysed with a significant reduction in the amount of hand-calculation and calculation time. However, the accuracy of the numerical methods and the way the methods programmed are unknown.
Due to the unknown degree of accuracy of the empirical and numerical methods, this thesis is aimed at investigating the ultimate response of the top hat purlins under uplift and downward loadings, primarily the single, double and triple span top hat-sheeting systems, using empirical equations provided by AS4600 (empirical equations from the American Iron and Steel Institute (AISI) Specifications were also used in some sections in this thesis) and two numerical methods of analysis: the finite strip method and the Simplified Model (developed based on the elasto-plastic finite element method of analysis). The validity of these methods was also evaluated by comparing the results from these methods with the capacities provided by the thin-wall manufacturer, Stramit Corporation Limit. The two numerical methods, the finite strip method and the Simplified Model, were implemented with the programs: Thin-Wall and Purlin Analyzer, respectively.
This thesis was concluded that Thin-Wall is most suitable to use under simple structures, preferably for end-rotationally fixed members subjected to axial loads. The validity of Purlin Analyzer was proven to be of a high degree of accuracy in results for long span purlins; and it was concluded that the empirical equations used in the design code were impractical for the new sections being developed in the cold-formed steel market.