An analysis of the structural behaviour of axially loaded full-scale cold-formed thin-walled steel structural panels tested under fire conditions

This paper presents a theoretical analysis to predict lateral deflections and failure times of six full-scale cold-formed thin-walled steel structural panels, tested under in-plane loads and exposed to fire attack on one side. The main objectives of this study are to investigate the effects of thermal bowing deflection and to check the relative merits of using either ENV 1993-1-2 or a modified version of ENV 1993-1-3 to perform design calculations for axially loaded steel studs under non-uniform temperature distributions. In general, the design of steel studs should also consider the effect of shift of neutral axes in both principal directions of the cross-section at elevated temperatures. However, such calculations can be rather time-consuming. Therefore, a secondary objective of this study is to assess how to effectively and accurately deal with this aspect in routine design calculations. It can be seen that thermal bowing deflections will have substantial effect on the fire resistance of steel structural panels exposed to fire on one side. However, for design calculations, it is not necessary to consider the effect of increasing thermal bowing deflections due to axial compression and reducing elastic modulus of steel at elevated temperatures. Both ENV 1993-1-2 and ENV 1993-1-3 may be used in design calculations of the ultimate load of this type of construction. It appears that ENV 1993-1-3 gives slightly better agreement with test results, but ENV 1993-1-2 is easier to implement because it does not require additional calculations of effective areas of thin-walled cross-sections at non-uniform elevated temperatures. The effect of shift of the minor axis is very small on the prediction of panel failure times and can be ignored to simplify routine design. Neutral axis shift of the major axis has some effect and may change the panel failure position from at the mid-height to the support. However, ignoring this neutral axis shift seems to give the best agreement to test results in term of panel failure location and panel failure times.