Challenges and solutions associated with the simulation and design of cold-formed steel structural systems

Abstract

The treatment of cold-formed steel sections in design codes is very largely restricted to individual members under ideal conditions. More efficient design is possible if the complexities of the structural response caused by the thin plating and complex shapes, together with the actual conditions of load introduction and restraint arising from practical situations can be recognised. Traditionally this has only been possible by resorting to full-scale testing. This is, of course, time consuming and expensive; moreover, the impossibility of covering all variations of all the important problem parameters means that developing a comprehensive understanding of all aspects of the physical behaviour is unlikely. Numerical analysis offers the promise of an alternative approach. However, for this to be reliable there must be confidence that it accurately models the physical situation. For the past decade a programme of research has been underway aimed at the provision of a more complete understanding of the structural behaviour of cold-formed steel sections when employed in particular practical situations. Three such cases are addressed herein: purlins as used in the roofs of industrial buildings, beams used to support floors and columns forming part of a stud wall framing system. In each case the process has been to firstly identify all the important structural components including fastening arrangements, then to develop numerical models using ABAQUS that represent each of these physical features to a sufficient degree of accuracy, then to validate the models by comparison with all available test data, then to conduct parametric studies covering the full range of variables found in practice and, finally, to use the pool of results and the improved insights into behaviour as the basis for improved design approaches that, by more accurately capturing the key physical features, provide better predictions of performance. An important feature of this has been to ensure that the resulting design procedures were the simplest possible consistent with reliable predictions.