ABSTRACT

In structural safety, the bridges have been widely studied because of their importance in the transportation system and their high cost of maintenance and repair. In this work, the nonlinear finite element analysis (NLFEA) is integrated with the stochastic reliability approach to improve the safety assessment of concrete structures. The numerical simulations were performed in the software ATENA, using a concrete constitutive model based on plasticity theory and fracture mechanics. The probabilistic model was developed in MATLAB, and the random fields were generated using the software SARA. The reliability analysis of a post-tensioned concrete bridge under traffic loads was performed, using the Response Surface Method (RSM) and including the spatial variability of the concrete properties. The analysis studied the influence of the correlation length and the distribution of the properties values in the random fields. Also, the probability failure of the structure was estimated and compared with the fib Model Code 2010 recommendations. It can be concluded that through the methodology implemented, a more realistic simulation of the concrete properties was achieved, and the safety level assessment was improved.