A simple model for lateral torsional buckling resistance of steel I beams under fire condition – comparison with experimental results
Conference Paper
Overview
Research
View All
Overview
abstract
When a beam is bent about its axis of greatest flexural rigidity it may twist before it reaches its strength
limit state. This flexural stability limit state is most commonly referred to as lateral torsional buckling of a beam.
The twisting of the beam occurs when the compression flange becomes unstable as a result of its being subjected to
flexural induced axial stresses. Lateral torsional buckling is of importance when the compression flange is laterally
unsupported as is often the case in continuous beams, cantilever beams, frame beams and frame columns.
The aim of this work is to develop a simple model, i. e., an analytical method to be used by designers for the
calculation of the resistance moment of steel I-beams failing by lateral torsional buckling when submitted to the fire.
A particular attention has been paid to the possibility to use the same model as the one proposed in
Eurocode 3 – Part 1-1, simply modifying material properties according to the temperature. This is the procedure
currently proposed in Eurocode 3 – Part 1-2, although its accuracy has never been demonstrated and can indeed be
questioned. Due to the fact that higher temperatures usually develop at the end of the flanges than in the rest of the
section, the decrease of the bending stiffness in case of fire is faster around the weak axis than around the strong axis
of the section. The method has been developed from a set of experimental results performed at elevated
temperatures.
As it is known any analytical, numerical or theoretical model is much more likely to be accepted if it is
backed and supported by a set of experimental tests. Comparison of the simple model has been made with laboratory
test results obtained in the framework of this work. A set of experimental full-scale tests were performed on IPE100
profiles at elevated temperature for several length specimens from 0.5 meter to 6.5 meters of buckling length.
Residual stresses, geometrical imperfections and material strength were measured for each tested element.
The load was applied after the heating of the beams. The beams were electric heated by means of ceramic
mat elements. Automatic control on different heating devices was presented in order to ensure a uniform temperature
distribution along the length of the elements. The temperature field has been measured with thermocouples welded
on the beams.
A Set of experimental results are presented, relating the critical load with the mid span movement of the
beam cross section, when submitted to a constant moment distribution and to a uniform distributed load, due to the
ceramic mat and the insulation material weight.
