Development of models for assessing corroded steel girder web panels in bridges
Neglecting the case of cable-stayed bridges, bridge steel girders are mainly subject to bending and shear, not so much to axial forces. What is more, these girders are usually built-up cross-sections with a very large web slenderness ratio, often exceeding 100. In this case, one major aspect of structural safety is the local buckling resistance under shear loading of the web. Moreover, it has been shown that the primary enemy to sustainability of steel structures is corrosion which arises from the material’s unstable thermodynamic nature. Bearing this in mind, given the increasing scarcity of natural resources and the high economic cost of corrosion, understanding how to properly consider it in the assessment of existing structures so as to extend their service life is of greatest importance.
Within this framework, this work proposes two models for accounting for a partial uniform thickness reduction in the determination of the remaining peak shear buckling capacity of steel girder web panels. The first is a finite element (FE) model developed in ABAQUS. This model is shown to yield results close to test data and was subsequently used to conduct an extensive numerical study on the effect of corrosion on the shear capacity of partially corroded steel plates. Findings indicate that corrosion may reduce the capacity by up to 50% depending on the case and that it also affects the overall plate shear buckling behavior. The second model, in the form of equations, was developed on the basis of a rigorous statistical analysis of the previous numerical results.
Making use of an existing weathering steel bridge located in Switzerland, the performance of the two proposed models is compared with that of two other simpler methods. Based on this comparison, it is observed that the proposed equations provide an additional degree of refinement with respect to the simpler methods while remaining more cost-effective than the FE model. This additional refinement was shown to be potentially sufficient for proving that an existing structural element still has sufficient carrying capacity. Furthermore, the corresponding method was proven more advantageous with respect to the simpler methods in case of highly heterogeneous thicknesses over the cross-section height.
2024-01-19
Lausanne
EPFL