Bidirectional Probabilistic Bridge Load Rating Model

This paper introduces a bi-directional model, which applies to bridge load ratings based on the methodology of the following codes: (1) allowable stress; (2) load factor; and (3) load and resistance factor design. The prognostic component calculates bridge load ratings probabilistically considering random variables representative of loads and material properties. The diagnostic component helps identify possible deficiencies that result in an undesirable bridge load rating. Monte Carlo simulation and Bayesian Networks are tools employed in the model. The model is applied to a prestressed concrete beam in an example with conclusions related to its applicability as a decision-making tool in areas related to bridge posting, inspection methods and rehabilitation options.