Capteur chimique à fibres optiques pour la mesure des ions chlore dans le béton à un stade précoce

In the national and worldwide context, countless reinforced concrete structures are in an advanced state of deterioration. A principal cause of such degradation is chloride induced corrosion of reinforcement bars. This phenomenon is accentuated in countries where de-icing salts are used for road safety, as well as in maritime zones. To date, no non-destructive method reliably quantifies chloride content during the corrosion initiation phase. Measurement of such a parameter is important for development of a better understanding of the complexity of corrosion phenomena and more practically, for better management of existing structures. The principal objective of this study is to propose a method for non-destructive measurement in order to monitor continuously and in real time free chloride content in concrete pores. In this context, a chemical sensor that employs optical fibres was developed and tested. The sensor functions using the fluorescence of an indicator dye that is sensitive to chlorides. Through fluorescence spectroscopy, variations in the concentration of free chlorides are related to intensity fluctuations of fluorescence. In this type of sensor, the role reserved for optical fibres is limited to data transport, thereby enabling remote measurement. The use of optical fibres also provides an advantage compared with existing electric non-destructive detection systems due to superior electromagnetic stability. Theoretical and experimental studies calibrated and validated the sensor for implementation within mortar and concrete. Free chloride concentrations between 0 and 350 mM can be detected, this represents 0.05% and 0.6% of cement weight. The system stability is approximately 0.02% [Cl-]/year. Measurement accuracy is generally lower than ± 8%. Two experiments reproduced climatic variations in a controlled environment. The first test simulated a hot maritime climate and the second test simulated a cold continental climate. These tests confirmed that it is possible to determine with precision free chloride content. A constant temperature and relative humidity test showed a good concordance between results from the measurement system and an existing chloride transport model. Although measurements indicate faster chloride ingress compared with model predictions, test results and model predictions coincide after a few days. In conclusion, fluorescence spectroscopy with optical fibres offers an innovative means for early and non-destructive detection of free chloride content in concrete. Experiments and model predictions demonstrated complementary strengths and weaknesses. As a result, there is much potential for synergy in order to improve the science of corrosion process understanding and to plan appropriately for preventive action in practical situations.