Assessment of the corrosion degree of beams subjected to accelerated corrosion based on indirect data

Abstract

In the view of circular construction, there should also be a focus on the assessment of existing structures. Their remaining strength and capacity should be determined, taking into account the degradation over their life-time. Often, only indirect measurement data is available to assess the level of degradation. Previous studies have illustrated how strain data from proof-loading and modal data from vibration tests can be used to estimate the corrosion level of reinforced concrete structures. However, most of these studies are based on virtual measurement data, and the use of real-case data will often be accompanied with additional challenges. To get a better insight in the challenges of real-case data, an experimental campaign has been performed in the Magnel-Vandepitte laboratory at Ghent University. Different real-size reinforced concrete beams have been subjected to accelerated corrosion. They have been subjected to 4-point bending tests and vibration tests. Afterwards, the actual corrosion degree was determined based on the mass loss of the reinforcement. The influence of the corrosion degree on the measurement data has been investigated and a Bayesian inference framework has been adopted to estimate the corrosion degree of these beams from the test results. From these analyses, important conclusions can be drawn towards the use of Bayesian inference for assessment of the corrosion degree on real-case structures.