Contribution à l'étude de l'expansion résiduelle des bétons atteints de réactivité alcalis-silice

Abstract

L'objectif de cette thèse est d'apporter des éléments nouveaux contribuant au développement d'une méthode de prédiction de l'expansion résiduelle du béton atteint de réactivité alcalis-silice (RAS). L'originalité de ce travail réside principalement dans l'apport de résultats expérimentaux sur la variation de la concentration en alcalis en solution dans divers types de bétons atteints ou non de RAS. Quatre mélanges réactifs et un mélange témoin ont été étudiés. Trois granulats canadiens ont été utilisés: Ie calcaire Spratt et Ie grès de Potsdam (réactifs) et Ie calcaire Limeridge (non réactif). Des éprouvettes de béton ont été confectionnées a partir de ces mélanges enrichis en alcalis et elles ont etc entreposées a 38°C en air humide (>90% H.R.) afin d'accélérer la réaction selon la norme canadienne CSA A23.2-14A. Des carottes de forage provenant d'un barrage hydro électrique atteint de RAS ont également été conservées dans ces conditions. Des mesures d'allongement et de masse ont été prises a intervalle régulier. Plusieurs études ont été réalisées au cours de l'expansion de ces éprouvettes: examen pétrographique quantitatif, analyse de la solution interstitielle extraite des bétons, dissolution de la silice réactive en solution alcaline concentrée, caractérisation de l'auréole de réaction.

Abstract: The aim of this thesis is to bring about some new elements that may contribute to the development of a method to predict the future expansion of structures affected by alkali-silica reaction (ASR). The main contribution of this work is to report experimental results on the variations of the alkali concentration in the pore solution of various concrete specimens affected or not by ASR. Five mixtures (four reactive and one non reactive) were tested Three Canadian aggregates were used: the Spratt limestone, the Potsdam sandstone (both reactive), and the Limeridge limestone (non reactive). Several concrete prisms were made from these alkali-enriched mixtures and were stored at 38 [Celcius degrees] in humid air (>90% H.R.) in order to accelerate the reaction (Canadian standard CSA A23.2-14A). Some cores drilled from a large hydroelectric dam suffering from ASR were also kept in these conditions. Expansion and mass measurements were taken at regular interval on these specimens. Several investigations have been performed on the tested specimens: quantitative petrographic examination, analysis of the extracted pore solution, dissolution of the reactive silica in NaOH solution, characterization of the reaction rim. The main conclusions are: (1) A relation is shown between the expansion of reactive prisms and their water gain. (2) The pore solution alkalinity of concrete submitted to the accelerated test CSA A23.2-14A decreases with time. (3) According to an alkali mass balance performed in this thesis, no more than 30% of the total alkalis would be in solution after 12 weeks. (4) It seems more realistic to consider the alkali critical limit stopping ASR in terms of concentration rather than in terms of alkali content. (5) The results obtained from this thesis confirmed that a certain part of the alkali ions in concrete pore solution that had become fixed by drying are not subsequently extracted after re-wetting: the alkali concentration [Na+K] was reduced from 34% to 61% by the drying and re-wetting treatments. (6) A direct correlation is shown between the dissolution rate of silica and the temperature when reactive sandstone aggregates are kept in concentrated NaOH solution. (7) The reaction rim surrounding reactive sandstone particles is mainly composed of precipitated silica."--Résumé abrégé par UMI.