Quantification of hidden safety reserves between different highway traffic load models based on probabilistic calibration analysis

This paper presents the initial findings from a research project initiated in Switzerland with the objective of developing a new traffic load model for highway bridges that quantifies safety reserves more uniformly. The various proposed new load models and their performance when compared to results from traffic simulations are presented. A particular attention is directed towards the full- probabilistic reliability analyses and the interpretation of the results, to demonstrate how these analyses facilitate a more uniform consideration of safety across various structures and failure modes. The framework used to identify the most optimal load model configuration is presented with a discussion of the different criteria leading to the final choice, including the accuracy and performance of the model, safety considerations and applicability for practitioners. The paper presents the results for single lane analysis and showcases how it will be possible to optimize the existing load model for improved performance.