Experimental and Numerical Investigation of Tapered Steel Girders with Flat and Sinusoidal Webs under Three-point Loading

Tapered steel I-girders with sinusoidal corrugated webs offer a promising alternative to conventional flat-web designs, yet no experimental study has addressed tapered configurations.

To address this gap, six tapered girders - three with flat webs and three with sinusoidal corrugated webs - were tested under three-point loading at mid-span depths of 400, 600, and 800 mm, combining experimental work with nonlinear finite element modelling.

The corrugated-web specimens carried 50%, 43%, and 13% more load than their flat-web counterparts at increasing depths. Larger strength gains at the shallower sections point to a pronounced dependence on the corrugation-to-depth ratio. The failure mechanisms were markedly different as well. Finite element models developed in ABAQUS captured the experimental ultimate loads and mid-span deflections well, with mean discrepancies of roughly 3% and 3.3%, respectively.

Comparison with available analytical predictions showed that the Nikoomanesh–Goudarzi formula gave reasonable estimates for the shallower specimens but overestimated capacity by 10% at 800 mm depth, while EN 1993-1-5 remained conservative for all specimens, with Pexp/PEN ratios between 1.80 and 2.05, suggesting that dedicated provisions for such members are needed.

Taken together, the findings indicate that sinusoidal corrugated webs can meaningfully improve both shear resistance and post-buckling ductility in tapered girders - especially when the corrugation-to-depth ratio exceeds about 15%. It is worth noting that neither AISC 360 nor EN 1993-1-5 currently covers this configuration. The tension field mechanisms observed here suggest that extending both codes to cover corrugated-web tapered members is justified, and the present dataset serves as the first experimental benchmark for such an extension.