Comparison of design and actual reliability indices of a reinforced concrete road bridge under the influence of traffic flow loads

Currently, the semiprobabilistic approach is employed in the practice of bridge design in Russia to account for uncertainties in loads and material properties through reliability coefficients. The semi-probabilistic approach is convenient to use, but it has several disadvantages. Firstly, the use of constant coefficients fails to consider the actual characteristics of traffic composition on specific roads. Secondly, assessments utilizing the semiprobabilistic approach provide only qualitative outcomes (for instance, confirming that a structure meets the conditions for the first group of limit states) and do not allow for a quantitative evaluation of the probability of structural failure. This study employs a full probabilistic approach to determine the reliability index (β) of a cross-sectional ribbed reinforced concrete span structure in a road bridge context. The loading conditions of the span structure were considered using both a model of design loads (taking into account its conversion to a full probabilistic supply) and a model of loads from actual traffic flow, for which data from weigh stations on Russian highways were analyzed and distributions of heavy vehicle masses were constructed. The SOFiSTiK software suite was used as a tool for conducting probabilistic analysis, utilizing the special RELY module for probabilistic calculations, which allows for computations using FORM and SORM methods (First-Order and Second-Order Reliability Method). The results of comparing reliability indices indicated that, for the specific case examined, the actual reliability index is 13 % lower than the design index and nearly 40 % lower than the target reliability index adopted according to recommendations from foreign standards. The algorithm developed during this research can be further applied to determine the reliability indicators of existing operational structures.