Cross-frame stiffness modification factors for composite steel I-girder bridges

When analyzing steel I-girder bridges, the approach used to model cross-frames can significantly impact performance predictions for girder stability during construction and for cross-frame fatigue under in-service traffic loading. A common practice is to model cross-frame members as truss members subject to axial forces only. Recent research has shown that this approach can lead to erroneous predictions of cross-frame stiffness and cross-frame member forces. Actual cross-frames are typically constructed using single-angle members with gusset plate connections that introduce significant out-of-plane eccentricity and in-plane rotational restraint. These connection effects combined with the complex bending behavior of single-angles results in significant bending of the cross-frame members. This bending behavior can significantly change the axial stiffness of the cross-frame member and potentially introduce large errors in truss element models. The objective of this research is to study the behavior of cross-frames in steel I-girder bridges to better understand their stiffness and internal force distributions during in-service traffic loading on the completed bridge as well as during construction of the bridge. The research involves development of high-fidelity three-dimensional finite element models of steel I-girder bridge systems, with predicted cross-frame response validated using laboratory experimental data as well as data from field instrumentation of in-service bridges. The validated models are then used to conduct parametric finite element studies to examine a wide range of bridge and cross-frame geometries. Based on the results from the parametric studies, stiffness modification factor for truss element models is developed to improve the analysis of cross-frames in steel I-girder bridges.