Bayesian framework for non-destructive post-fire assessment of reinforced concrete beams using the incremental neutral axis position

Post-fire assessment in concrete structures is a complex challenge. This study addresses this challenge, by incorporating an innovative technique that employs Fiber Bragg Gratings (FBGs) to measure strains at the top and the bottom of a beam and analyse them into a Bayesian inference framework. The FBGs allow to determine the position of the incremental neutral axis under bending deformation. The change of position of the incremental neutral axis relates to the degradation of concrete stiffness induced by fire, providing a key indicator of the structural condition of the whole member. The Bayesian methodology allows for a systematic handling of uncertainties, integrating prior knowledge with new data to improve the assessment's accuracy. By combining FBG-based strain sensing and advanced concrete modelling within a Bayesian framework, a novel approach is proposed to tackle the high uncertainties of post-fire assessments and deliver more reliable predictions than existing techniques. This offers a structured framework for interpreting the measured data and predicting the structural health of fire-affected concrete structures. To enable Bayesian inference, a numerical model is developed to calculate the incremental neutral axis position during and after fire. The model evaluates all the strain components, both reversible and irreversible and aggregates them to calculate the neutral axis position. The model's capabilities are validated using experimental data. The application of this methodology to a demonstration case shows its potential. The results show that employing the assessment technique can provide information on both the fire exposure, material properties of the member and its residual capacity. It highlights the feasibility of using FBG-based measurements for the post-fire assessment of concrete structures and underscores the value of Bayesian methods in managing the uncertainties inherent in such evaluations.