Measurement-Based Evaluation of Non-Composite Steel Girder Bridges

The results from an experimental study into the behavior of non-composite slab-on-steel girder bridges are presented. A load rating technique was developed that uses site-specific measurements to estimate the capacity of a suspect bridge. Two bridge units were tested under increasing static loads beyond first yield of the girders with a flatbed trailer loaded with concrete barriers. Both units were originally four-span continuous but were altered to three spans to increase the maximum span length. The load tests were unique in that dump trucks were applied between trailer load cycles to monitor the effect of overloads on the bridge response. Measured deformations included strain, vertical deflection, and girder-slab slip.

Lateral load distribution was evaluated experimentally by expressing the bending moment in each girder section (derived from the measured strains) as a percentage of the total bridge moment. The strain derived bridge moment was lower than the statical bridge moment due in part to bearing restraint and deck stiffness. Measured distribution factors compared well with finite element analysis but were below empirical factors from AASHTO Standard (1996) and LRFD (1998) Specifications. There was more agreement with the LRFD-based values. Partial composite action was examined by comparing the measured section properties (i.e., moment of inertia, neutral axis position, section modulus) of the girders with those for a non-composite and full composite section. Due in part to the flexural participation of the curb, the exterior girder exhibited more composite action than interior girders. As a result, a higher percentage of the applied load was distributed to the exterior girder. The heavy trailer loads had no significant impact on the load distribution and partial composite action of the girders. More significant changes occurred between different transverse load positions of the dump truck.

The proposed experimental bridge rating process consists of a hierarchy of steps of increasing complexity and effort, some of which do not require load testing. The process terminates once a satisfactory rating is achieved. Procedures for using measured distribution factors and section moduli for the girders to improve the AASHTO Allowable Stress and Load Factor ratings are given.