Experimental validation, numerical modelling and design comparison for steel beams against lateral torsional buckling
Although steel is commonly used in construction, mainly for slender structures, it does present a significant risk of instability. One of the main instability phenomena is lateral torsional buckling (LTB). It occurs when a beam is loaded in bending and has a significant difference in inertia between its two cross-sectional axes. This phenomenon can lead to significant damage and loss of stability. The aim of this project is to evaluate the construction standards used to deal with lateral torsional buckling, in particular the upcoming new Eurocode, which has not yet been published. To achieve this, a large-scale laboratory test is carried out. The selected specimen is a hot-rolled IPE 330 section subjected to 4-points bending. In order to predict the beam’s behavior appropriately and design the test setup, numerical modeling is used. The finite element model includes a residual stress model, an advanced material model and initial geometric imperfections. Furthermore, in order to check the modelling assumptions and make the test as complete as possible, additional destructive tests are carried out: a residual stress measurement test using the sectioning method, monotonic and kinematic tensile tests on standard and round coupons, measurements of the beam’s initial dimensions and initial out-of-plane imperfections. These experimental results are then inserted into the finite element model, enabling the comparison between the simulated and real-life LTB experiment. If the results are consistent, the finite element model can be used to extend the study of LTB to other steel profiles. The results obtained from monotonic tensile tests on standard coupons show that the assumptions made about the mechanical properties are closed to the expected values. What is more, the finite element model showed that the results obtained are more or less the same as those initially estimated. The test-setup is therefore correctly conceptualized. Given the information available today, the new Eurocode is closer to the peak LTB resistance estimated by finite elements than the current Eurocode, but not as close as the Swiss standard.
École Polytechnique Fédérale de Lausanne
2024-01-19
Lausanne
EPFL