International Concrete Abstracts Portal

The volume stabilities of mortars have been monitored for periods up to 5 years. In addition to controls maintained at 20°C, prisms were subjected to high temperature curing regimes, typically 12 hours at 70, 80 or 9OO C. Curing temperature was found to be the dominant factor in determining whether significant (>O. 1%) expansion took place during subsequent water storage at 20°C. Most of the mortars expanded after curing at 90°C but very few at 80°C and none at 70°C. Analysis of the results for mortars cured at 90°C indicated that expansion increased with cement fineness and levels of alkali, C3 A, C3 S and MgO and showed a maximum at a cement SO3 content of -4%. The pessimum SO3 content increased with cement alkali content. The incorporation of ground limestone accelerated expansion but did not affect the ultimate value. Cement replacement levels of 30% or 40% with siliceous fly ash and 30% or 50% with ground granulated blastfurnace slag appeared to prevent expansion. A sample of Type IS cement gave only moderate expansion. The mortars that expanded after curing at 80°C had high alkali levels, through the incorporation of KOH or K2 SO4 in the mixing water. At 20°C no long-term expansions were observed, despite the inclusion of cements with total SO3 levels up to 5.6% and clinker SO3 levels up to 2.6%. The Type K cement expanded by -0.05% within the first 7 days and was then stable. Cements based on a high C3 A clinker gave early expansions (-56 days) which increased as the cement SO3 content increased and as the specific surface area decreased, although all expansions were <0. 1%.

Paste, mortar, and concrete cured at temperatures above certain limits may exhibit expansion and cracking during subsequent exposure to varying moist conditions. This phenomenon originally became known as delayed ettringite formation, DEF. DEF results in a typical microstructure which is demonstrated with examples from field samples and laboratory-made samples. The microstructure is compared with examples of internal sulfate attack in laboratory samples. These typical features include gaps around the aggregate where the paste shows an almost perfect replica of the individual aggregate surfaces. Expansion of the paste on a scale which is homogeneous relative to the aggregate would lead to such features. The chemistry in DEF is similar to that of sulfate attack. A mechanism involving hydrates of aluminates and possible unhydrated cement clinker particles is discussed.

The stability domain of stoichiometrically prepared ettringite and monosulfate hydrate in water at 30 and 1OOoC also in the presence of lime, alite, monocarboaluminate hydrate, or each other is reported. The formation conditions of both calcium sulfoaluminate hydrates in pure ordinary portland cement dealing mainly with the hydration behavior of the tricalcium aluminate and the tetracalcium aluminoferrite in the presence of gypsum, also of lime, alite, or each other at 300C is summarised. The type of hydrate formed in sodium -hydroxide, sulfate solution or as a result of alkali doping in the structure of the anhydrous phase is given. The ettringite/monophase formation and duration in the three stages of C3 A and C4 AF hydration (instantaneous reaction, dormant, and acceleration period) are described. The present review focused on the stability of the calcium sulfoaluminate hydrates is a summary of 13 publications.

Recent debate on delayed ettringite formation (DEF) as a form of internal sulfate attack - a distress mechanism for ambient- and steam-cured concrete - has motivated this paper, which examines relevant data from the literature and from the authors’ laboratory. DEF can be associated with distress in high temperature concrete, but not for ambient-curing in the absence of external SO3 sources, or an aggregate sulfate source. The two internal SO3 sources cited as potentially responsible -anhydrite in clinker and high SO3 concentrations in the silicate phases -are shown to be absent or fundamentally innocuous.

A series of mortars was studied, cured at 20, 80 or 90°C. The variables studied included sulfate level, alkali additions and slag additions. In parallel with measurements of dimensional changes, detailed study of the microstructural and microchemical changes was made by XRD and by SEM. One of the main findings of this study is that the composition of C-S-H gel around partially hydrated cement grains, analysed one day after heat curing, is significantly different between mortars which subsequently expand when water at 20°C and those mortars which do not expand. The details conserved in of this observation, its implications and expansion are discussed. limitations, and possible mechanisms of