تجزیه و تحلیل حساسیت نقص از ستون های کانال راز دار در درجه حرارت بالا

An imperfection sensitivity analysis is carried out on cold formed lipped channel columns integrated in wall structures. The equations given in Eurocode 3: Part 1.3 and finite element analyses are used to evaluate the flexural buckling strength and ultimate strength of the sections. The analyses are performed at both room and fire temperatures by introducing corresponding material data values obtained from high temperature transient tensile tests into the models. The purpose of the study is to evaluate the possibility of using normal temperature design formulae directly at high temperatures and to evaluate the influence of the choice of initial geometric imperfections on the modelled behaviour of the columns. The analytical estimations are compared to the results of the finite element analyses, which in turn are validated using available test results. It is shown that the magnitude of the modelled types of local imperfections has an effect on the compression stiffness of the members, whereas the magnitude of global flexural imperfections has more influence on the ultimate strength obtained in the analyses. The influence of the choice of buckling curve at high temperatures is discussed.

Cold-formed steel members are generally quite sensitive to geometric imperfections. Although this is commonly acknowledged, no general rules exist for the modelling of imperfections, mainly because of lack of extensive and accurate data. The guidelines given in design codes or product standards usually provide only conservative upper limits for the magnitude of imperfections to be used in design. An alternative approach to the problem is the probabilistic method for the determination of a suitable maximum imperfection presented by Schafer and Peköz [1]. In the use of advanced analysis methods, such as finite element techniques, imperfection sensitivity is often characterised by a large number of very closely spaced eigenvalues representing more or less similar buckling modes. A choice of imperfection mode combinations has to be made in order to continue the analysis into the post-buckling range, where also material imperfections come into play. The current paper attempts to shed light on the way different imperfection types and their combinations affect the behaviour of cold-formed steel columns by way of finite element analyses with the purpose of finding suitable imperfection magnitudes to be used in further research. The study is part of a larger research project concentrating on the behaviour of cold-formed steel columns and beams at high temperatures. The analyses are performed at both room and high temperatures, which are introduced into the models directly using corresponding material properties obtained from tests [2]. Another objective is to evaluate the basic assumptions needed for the use of Eurocode 3: Part 1.3 [3] formulae also in fire conditions. The studied column cross-section is C 100×40×15 with wall thickness t=1 mm and length L=2500 mm. The columns are designed to be integrated into a wall structure so that they are attached to wall panels at each flange. The connection is assumed to give sufficient restraint against torsional and torsional–flexural buckling modes. Therefore only flexural buckling about the strong axis is considered in this paper. In this study, these boundary conditions are assumed to be valid also at the temperature T=600°C, although gypsum wall panels used in practice have usually already lost their restraining capacity at such high temperatures.

The results of the parametric study show a constant pattern for the way the behaviour of the compressed member changes when the initial imperfection conditions are varied. When buckling curve c is used, all analyses give values higher than the ultimate load values calculated using Eurocode 3 [3] and a stress value corresponding to 0.2% plastic strain. When buckling curve b is used also at high temperatures, the results are relatively close to each other and the Eurocode 3 prediction is still slightly on the safe side in most cases, when compared to the analysis results. However, when compared to test results, the finite element analyses give conservative values. The results of this study hint that the use of buckling curve b could also be appropriate at high temperatures. The use of buckling curve c tends to produce unnecessarily conservative results, especially considering that partial safety factors were not included in the calculations. If a stress value corresponding to a higher total strain were used in the analytical predictions, the EC3 values would eventually become higher than those obtained from the FE-analyses. The threshold value has not been determined in the scope of this study, but on the basis of this parametric study, the use of the stress corresponding to 0.2% plastic strain for lipped channel sections at all temperatures seems appropriate and gives results that are on the safe side. It was seen from the results of the parametric imperfection study that the increase of the magnitude of the local imperfections leads to a relatively straightforward decrease of initial stiffness of the member and that, on the other hand, the magnitude of global imperfections has more influence on the ultimate load of the member. Overall, the analyses give a fairly good understanding of the basic imperfection sensitivity of these members. A combination of both local and global imperfections should be used in ultimate strength analyses. Moreover, it is worth considering to model the local imperfection as a combination of two or more local buckling modes. A suitable value for global imperfections is found to be approximately L/500 and for local imperfections h/200.