Contribution of Stress Measurements in Modelling Structures Affected by Concrete Swelling

Some dams in the EDF parc exhibit irreversible displacements that can be attributed to internal swelling reactions. To evaluate the impact of this swelling, numerical models are implemented. These models allow for the reproduction of displacement evolution, estimation of stress levels, and prediction of long-term cracking. Calibration of these models involves the use of data obtained from monitoring and laboratory tests, which necessitates the formulation of certain assumptions, such as the areas prone to swelling and the magnitude of swelling. However, formulating these assumptions can be challenging due to the numerous parameters involved in chemical reactions. Factors such as the presence of water, temperature, concrete composition, and the temperature reached during concrete casting play a crucial role in these chemical reactions, and these influential factors are not always well-known within the structure. Additionally, parameters like creep and cracking directly impact the response of the modeled structure to external loading. All these factors influence the model's response and can lead to erroneous assumptions, resulting in an imperfect diagnosis of the structure, either by overestimating swelling in certain areas of the structure or by not considering it at all. To improve the validity of the assumptions made, one approach is to measure stress levels in situ and install sensors for continuous monitoring of the structure. These long-term measurements and observations allow for the validation or modification of the initial diagnosis, accordingly, ensuring better alignment with the actual behavior of the structure. This paper provides an example where stress measurements have enabled improved modelling.