Research and guideline synthesis for horizontally curved steel I-girder bridges

Abstract

Horizontally curved steel I-girder bridges are commonly designed and constructed as economical options in congested areas despite their complex behavior. Girder responses during erection and deck placement are challenging to evaluate due to warping of the girder and cross-frame system, which induces fit-up concerns and causes locked-in forces. Determination of girder major-axis bending under live load is complicated by curvature, which often makes the standard line girder analysis that uses live load distribution factors invalid; quantification of lateral responses for horizontally curved steel I-girder bridges also generally relies on refined analysis. Additionally, global temperature variations and local thermal gradients on these bridges result in more complicated radial and lateral movements and stress distribution compared to straight bridges. Standard design and analysis simplifications generally do not apply to horizontally curved steel I-girder bridges—guidelines can be complicated to follow, which could discourage usage. Some (but limited) state transportation agencies in the U.S. have distinct requirements, preferences, and procedures for design and construction of these bridges that could benefit from a nationwide synthesis. This paper holistically reviews existing experimental and numerical research on the behavior of horizontally curved steel I-girder bridges in the past two decades and synthesizes current U.S. state practices for their design and analysis, which leads to observations and insights of research and application gaps. Aspects of this review include (1) research on horizontally curved steel I-girder bridge behavior from 2000 to 2024, considering construction (during erection and deck placement) and in-service (including vehicle and thermal) loading conditions, (2) the common refined methods of analysis adopted in practice and research, (3) the U.S. state preferences and guidelines for design and analysis, and (4) the research gaps (between existing work and practical needs) and disparities in various state specifications, as well as their implications.