Optimizing on-site testing for post-tensioned concrete bridges

The decreasing structural reliability of key infrastructure due to aging and degradation represents a significant challenge for engineers, infrastructure managers, and society in general. The concealed nature of tendons in post-tensioned concrete bridges makes the assessment of their structural health a crucial point of several national bridge inspection guidelines. This assessment is commonly performed by means of direct and/or indirect tendon testing. A widely spread approach to the design and quantification of the number of tests is the American FHWA method, also adopted by the Italian guidelines. Unfortunately, in light of its frequentist nature and the consequent negligence of the prior knowledge on the health state of the structure, this one can lead to test campaigns with redundant and non-optimized allocation of resources and time. The present article proposes a novel Bayesian perspective of the FHWA test quantification methodology that accounts for: (i) the prior knowledge in an explicit manner by means of Bayes' theorem; (ii) the test uncertainties, as in the case of Nondestructive Testing. By means of a demonstrative application, this article displays how this novel methodology leads to an optimized number of tests whilst guaranteeing the required structural reliability and user safety levels.