Condition assessment and damage modelling of corroding reinforced concrete structures through Bayesian coupled learning models

Condition assessment and lifetime prediction of corroding reinforced concrete structures is a major engineering challenge. In practice, inspection and condition assessment are often based on (qualitative) visual inspections, occasionally complemented by quantitative measurements. Data from local measurements provides information on concrete properties, lay-out of reinforcement or specific damage phenomena such as reinforcement corrosion. The major challenge and reason why complementary non-destructive and minor-destructive testing (NDT/MDT) has limited application in practice is related to the local nature of the information and the difficulties to integrate such information in a general structural assessment with a consistent treatment of all uncertainties, despite its potential benefit. For integration in structural performance predictions, local information on material properties and damage needs to be translated, taking into account spatial variability and estimation uncertainties. Therefore, a major challenge that will be tackled in this research is the transformation of such complex, spatially distributed damage symptoms, detected by means of local measurements, towards degradation modelling and practically applicable structural models. This PhD research is part of a larger project with focus on developing a life-cycle Methodology for the Assessment of existing Concrete Structures (lifeMACS).