Reliability Based Design Optimization of Load and Rating Models for Bridge Structures in Michigan

The main objective of this study is to develop optimal live load models for design and rating of bridges using reliability-based design optimization (RBDO) methodology, such that target reliability levels for bridge girders subjected to Michigan traffic loads can be consistently met. Traffic data from 20 high-fidelity weigh-in-motion (WIM) sites collected over a two-year period across Michigan will be used for statistical analysis. From the filtered data, load effects are generated for a series of hypothetical bridges considering spans from 20 to 200 ft. and girder spacings from 6 to 12 ft. Simple moments and shears, for both single lane and two lane live load effects, are considered.;Based on the load effect data generated from the WIM vehicle configurations, load effects are probabilistically projected to 5 years (for rating) and 75 years (for design) to obtain estimates for the maximum load effect statistics. An extreme type I projection will be considered.;Optimal design and rating models are developed with a reliability-based optimization process using a genetic algorithm such that discrepancies in bridge structure reliability index are minimized.