In regular design practice, when it comes to conventional loading, such as uniaxial compression and bending, the local buckling of thin plates is taken care of through cross section classification. The effect of warping torsion, which also gives rise to normal stresses in the section, however, is typically not considered in the process.

Present work aimed to uncover the influence of warping torsion on the phenomena of local plate buckling and to investigate the limitations of the effective width method when it was applied against its intended use. In the case of varying results, a simple correction to the calculation method was to be developed to improve accuracy.

The examined I-sections were tested to failure and results were obtained with two different approaches: with finite element method and a Eurocode-based hand-calculation. The finite element models were refined to closely mimic physical experiments and their results were accepted as the true resistance of the sections, while the calculation method tried to capture the structural response in a practical, easily understandable way.

The calculated results showed reasonably good accuracy with that of the finite element analysis. However, what really stood out was how similar the change in resistance was when the section parameters were manipulated. Through a properly chosen function, this allowed for the creation of an exponent that could modify the calculated results to achieve an even greater accuracy.

The eccentricity of the applied load on the system was also manipulated to alter the proportion of normal stresses due to the two examined effects. It became clear that the stresses from warping in the applied calculation model were underestimated and the otherwise conservative method of effective width lost much of its safety margin when its application was extended to warping as well. Consequently, the consideration of stresses from warping in the regular design process and stability control of commonly used thin walled open sections seemed to be justified. The effective width method could not reliably cover the issue with retaining its original margin of safety.