Uncertainty quantification in the vibration-based damage assessment of a reinforced concrete beam

Finite element (FE) model updating is a technique that is commonly used for structural damage identification and localization. In FE model updating, the objective is to adjust parameters of a FE model such that its output corresponds better with experimental observations of the structural behaviour. In a structural mechanics context, the experimental data in many cases consists of modal characteristics of several natural modes. When FE model updating is applied for damage assessment, it is often assumed that the damage can be identified as a decrease in structural stiffness; using the available observations, some (substructure) stiffness parameters are updated to detect, localise and quantify structural damage. The FE model updating process involves solving an inverse problem and is subject to measurement and modelling errors, which give rise to errors/uncertainties in the predictions that are made by the FE model. The objective of this research is to quantify the effects of errors/uncertainties on the results of the damage assessment. To this end, two alternative approaches are employed: a non-probabilistic interval-based approach and a probabilistic Bayesian approach. Both methods are elaborated and applied to the damage assessment of a reinforced concrete beam.