Service life prediction of reinforced concrete structures subjected to corrosion: a comparative study
Kuleni Fekadu Yadeta,Sudath C. Siriwardane,Tesfaye Alemu Mohammed
DOI: https://doi.org/10.1108/ijsi-12-2022-0149
2023-05-09
International Journal of Structural Integrity
Abstract:Purpose Reliable estimations of the extent of corrosion and time required to reach specific safety limits are crucial for assessing the reliability of aging reinforced concrete (RC) bridges. Engineers and decision-makers can use these figures to plan suitable inspection and maintenance operations. Design/methodology/approach Analytical, empirical and numerical approaches for estimating the service life of corroded RC structures were presented and compared. The concrete cover cracking times, which were predicted by the previously proposed analytical models, were compared with the experimentally obtained cracking times to identify the model/s for RC bridges. The shortcomings and limitations of the existing models are discussed. Findings The empirical models typically depend on the rate of corrosion, diameter of steel reinforcement and concrete cover depth and based on basic mathematical formula. In contrast, the analytical and numerical models contain the strength and stiffness properties of concrete as well as type of corrosion products and incorporate more complex mechanical factors. Four existing analytical models were analyzed and their performance was evaluated against existing experimental data in literature. All the considered analytical models were assumed thick-walled cylinder models. The maximum difference between observed cracking time from different test data and calculated cracking time using the developed models is 36.5%. The cracking times extend with increase in concrete cover and decrease with corrosion current density. The development of service life prediction models that considers factors such as heterogeneity of concrete, non-uniform corrosion along rebar, rust production rate and a more accurate representation of the corrosion accommodating region are some of the areas for further research. Research limitations/implications Outcome of this paper partially bridge the gap between theory and practice, as it is the basis to estimate the serviceability of corrosion-affected RC structures and to propose maintenance and repair strategies for the structures. For structural design and evaluation, the crack-width criterion is the greatest practical importance, and structural engineers, operators and asset managers should pay close attention to it. Additionally, repair costs for corrosion-induced serviceability failures, particularly concrete cracking and spalling, are significantly higher than those for strength failures. Therefore, to optimize the maintenance cost of RC structures, it is essential to precisely forecast the serviceability of corrosion-affected concrete structures. The lifespan of RC structures may be extended by timely repairs. This helps stake holders to manage the resources. Practical implications In order to improve modeling of corrosion-induced cracking, important areas for future research were identified. Heterogeneity properties of concrete, concept of porous zone (accommodation effect of pores should be quantified), actual corrosion morphology (non-uniform corrosion along the length of rebar), interaction between sustain load and corrosions were not considered in existing models. Therefore, this work suggested for further researches should consider them as input and develop models which have best prediction capacity. Social implications This work has positive impact on society and will not affect the quality of life. Predicting service life of structures is necessary for maintenance and repair strategy plans. Optimizing maintenance strategy is used to extend asset life, reduce asset failures, minimize repair cost, and improve health and safety for society. Originality/value The degree of accuracy and applicability of the existing service life prediction models used for RC were assessed by comparing the predicted cracking times with the experimentally obtained times reported in the literature. The shortcomings of the models were identified and areas where further research is required are recommended.