| Summary: | The current standards for design of steel beams are divided into four classes based on the slenderness of cross-sectional parts. Depending on the section class a beams load capacity is either determined by a plastic, elastic or a reduced elastic model. The division between cross-sections calculated by a plastic and an elastic model is today done in a very harsh way with no smooth transition. This leads to a large difference in the calculated resistance for the sections located near the class limits. Elasto-plastic sections are in the current situation calculated as purely elastic and their large plastic reserve is not considered in the current standards. The ambition of this study was to develop an appropriate interpolation model for elasto-plastic sections which gives a finer transition between the plastic and elastic region. To verify the proposed interpolation model were a large range of beams with varying slenderness modeled in the FEM program Abaqus. The beams have been restricted to include only simply supported HSQ-beams. The results obtained were found to indicate that Eurocode generally seem to overestimate the capacity of plastic cross-sections in section class 1 and 2, in some cases by up to 10%. With this insight focus ruled over from the interpolation model to instead look closer into the accuracy of the limit values in Eurocode. Overall, the results points to that Eurocode overestimate the bearing capacity the closer the elastic area you get. In the elastic area the load capacity instead shifts into being underestimated, which in turn shows the lack of an interpolation model. There are a number of possible reasons why Eurocode seems to overestimate the bearing capacity in some cases. The largest impact on the results seems to be the initial deflection of the internal parts. It is possible that the introduced deflection in Eurocode is too far from the allowable manufacturing tolerances that can actually occur in reality. Other factors such as choice of buckling lengths when the slenderness is determined and the effect of residual stresses may also affect the ultimate load bearing capacity. Too big conclusions should not be made by the results of this study given the limitations made, but the results nevertheless indicate that there are deficiencies in the classification of cross-sections in Eurocode and that further research should be carried out.
|
|---|
