Get started
Work

Integratie van lokale en globale monitoringstechnieken voor beoordeling van de structurele capaciteit van gecorrodeerde gewapende betonnen elementen

Year: 2024
Type: dissertation
Abstract: The abundance of ageing infrastructure and growing traffic volumes severely increases the risk of structural failure. Especially rebar corrosion in reinforced concrete (RC) structures poses global safety risks and economic burdens due to the hazardous effects of corrosion on the structural capacity. Hence, this dissertation addresses corrosion assessment in RC elements, an important topic for enhancing infrastructure management. The research topic is addressed through three main objectives. Firstly, a two-step experimental programme for multi-level structural health monitoring (SHM) is designed and performed, taking into account the gaps and challenges identified in the literature. In step one, RC beams are corroded in an accelerated way, while being monitored with local techniques such as acoustic emission (AE) sensing and crack measurements and global vibration-based monitoring (VBM). In step two, the beams are subjected to four-point bending while simultaneously being monitored. Secondly, the presented research further employs the experimental results to validate novel methodologies for estimating corrosion parameters in RC beams, such as the corrosion level (CL) and apparent concrete stiffness, which are both valuable for structural assessment. Thirdly, the deterministic methodologies for parameter estimation are integrated into a probabilistic Bayesian framework to update prior corrosion parameter predictions into enhanced posterior probability functions. The updated corrosion parameters are further employed within finite element models in order to simulate the beams' structural behaviour. The experimental results of local monitoring demonstrate that AE sensing can effectively detect and localise accelerated corrosion, even within one-day monitoring periods and before any surface cracks appear. Moreover, as investigated through a newly developed methodology, AE localisation provides information about the spatial variability of corrosion damage and thus accurately predicts the relative CL along an RC beam. Absolute values of the CL are afterwards obtained by performing selective crack measurements and employing well-established crack width-CL relations provided in the literature. Combined use of AE and crack measurements estimated the CL with a mean absolute error of 1.53%CL for beams reaching CLs up to 14%. Global VBM, on the other hand, appeared challenging for detection of stiffness losses after moderate corrosion damage (CL<10%) within the experimentally tested beams. Nevertheless, severe longitudinal corrosion cracks did cause changes in the beams' natural frequencies, mainly those of the torsional modes. Moreover, dynamic structural stiffness decreases were also noticed after mechanical loading by significant changes in all natural frequencies. In addition to VBM, a complementary, crack measurement-based approach has been developed which links crack maps to changes in apparent concrete stiffness, as validated by experimental testing. Implementation of the local test results within a Bayesian framework enables updating of prior corrosion parameter estimations such as the CL and apparent concrete stiffness. The prior CL estimations are based on predictions of the time to corrosion initiation and the corrosion rate, which are both prone to large uncertainties, and the prior structural stiffness decrease is assumed to be negligible. Bayesian updating through AE and crack measurements improves the accuracy and reduces the uncertainty of the prior results, as experimentally validated. The standard deviation of the CL estimation more than halves (5%CL to 2.15%CL) after updating and the mean absolute error of the best estimate reduces from 8.09%CL to 2.46%CL. The reliability of the apparent concrete stiffness prediction also significantly increases since the standard deviation decreases from 8GPa to 3.58GPa. The increased accuracy after Bayesian updating is validated by improved FE simulations of the beams' natural frequencies. Comparison of the load-deflection behaviour of the updated finite element models with the results of experimental loading highlights the effectiveness of combining AE and crack measurements to update corrosion parameters, as the ultimate capacity of the corroded RC elements can accurately be predicted. To conclude, the experimental local-global SHM results demonstrate the complementarity of AE and VBM, showing that AE is effective for detailed early-stage corrosion assessment, whereas VBM can detect structural stiffness losses after severe concrete cracking. Furthermore, AE and crack measurements allow for quantification of the CL and structural stiffness loss through newly developed deterministic approaches, which can be employed in a probabilistic framework to enhance the accuracy and reliability of the corrosion parameters and to predict the remaining capacity of RC elements. Hence, this research contributes significantly to the field of reinforcement corrosion assessment by offering a robust methodology for evaluating corrosion damage in RC structures with a coupled local-global monitoring approach.
Language: English
Open Access status: green
OpenAlex ID: W7048115251
Cites:
Cited by:
Related to: