Incorporating Grillage Model Derived Load Distribution Factors Into Ratings of Prestressed Concrete Bridges

Abstract

Characterization and evaluation of bridges is a laborious task. A large number of bridges in Oklahoma are potentially deficient for shear due to differences between the current codes and those used to design the bridges. A robust method is needed to evaluate the sufficiency of these bridges. One potential consideration for accurate evaluation of bridges is consideration of composite behaviour and resulting load distribution. Load distribution among the girders of a bridge’s superstructure is dependent on many parameters such as girder spacing, material properties, skew angle, stiffening lateral elements (diaphragms), etc. The American Association of State Highway and Transportation Officials (AASHTO) LRFD Bridge Design Specifications contain load distribution factor equations for most of the common bridge types. The methods presented in AASHTO are simple, empirical, conservative, and intended to be applicable for a large sample of bridges. The work presented in this thesis is part of on-going work on the shear behavior of prestressed concrete girder bridges and focuses on how to accurately model transverse load distribution relationships and determine accurate load distribution factors for use in bridge design and load rating. The analysis is primarily based on the 2D grillage modeling method using finite element analysis. Different configurations of superstructures are examined by varying parameters such as girder type, span length, deck thickness, girder spacing, and presence of diaphragms to determine the parameters most affecting load distribution. Results from this study are also compared to a 3D finite element plate model for specific cases to evaluate whether a 2D model sufficiently captures bridge behavior for taller girder sections. It was observed that girder spacing has the largest impact on the load distribution factor among all the parameters considers in this study. The impact of diaphragms was more evident in plate models, particularly, when the girder spacings were larger. It was found that AASHTO LRFD equations produce values for shear load distribution factor at least 1.9% to 22.5% larger than those from grillage models.