Reliability-Based Calibration of the Code for Steel Girder Bridges

Abstract

Accurate evaluation and design of bridge structures can maximize the performance-to-cost ratio, which is one of the most critical aspects in infrastructure maintenance and expansion. The author of this dissertation presents the importance of steel girder bridges and their proportion in the national infrastructure, reviews methods used in calibration of bridge design code, and develops new statistical parameters and material models for structural A709 steel used in bridges. Material models for A992 steel and normal weight concrete were established from available test results. Material models served as inputs for simulation of resistance distribution of steel girders using the Monte-Carlo method. Resistance analysis was performed for noncomposite, and composite steel girders including rolled I-shaped sections, plate girders, and box girders. For high-performance steel sections, an analytical model was developed that uses nonlinear constitutive material models and accurately captures flexural moment carrying capacity. From the obtained simulations updated statistical parameters were developed for rolled sections. New statistical parameters of resistance were derived for plate girders and steel box girders. Using calculated resistance models, the author performed reliability analysis and assessed safety levels for steel girder bridges designed in accordance with the current design specification. Finally, a reliability-based calibration of resistance factors was performed, and appropriate recommendations were suggested for implementation in the code.