Fly ashes react with Ca(OH)2 and water to form calcium-silicate-hydrates. This consumption of Ca(OH)2 may increase the carbonation rate of fly ash concretes and mortars. On the other hand, the pore size distribution will be changed toward smaller pores due to the long term pozzolanic reaction of the fly ash so that the diffusion of CO2 will be diminished. Investigations have been carried out with mortars made of different types of cements and different fly ashes (different granulometry) from bituminous coal by variation of water-(cement + fly ash) ratio and the time of curing in water. The depth of carbonation dc, the carbonation rate vc, and the compressive strength fc have been determined during storage in the laboratory (20 C and 65 percent relative humidity) over several years. The test results have been compared with results of other investigations. There is an approximately linear relationship between vc and 1 / û fc28 (fc28 = compressive strength at 28 days). Related to the same compressive strength, there is no significant difference between concretes and mortars with and without fly ash as long as cements with normal CaO content are used, the cement content c is not too low ( > about 250 kg/m3) and the fly ash content f is not too high (f/c > about 0.3). In the case of concretes and mortars with low CaO contents, an increasing fly ash content leads to an increase in carbonation rate. Fly ashes with different granulometry have not influenced the carbonation behavior marked. However, the time of moist curing had a very strong influence.

Discusses natural and artificial carbonation of mortar and concrete. The theoretical analyses and experimental results show that in both cases the mechanism of carbonation of mortar or concrete is the same. They are comparable when CO2 diffuses in the gas phase, the carbonation coefficient equation is Q = a1(2C1 / KP)«. The experimental results also indicate that the pores with radii over 320 A have a great effect on the diffusion coefficient, and the following relation holds: ln a1 = 105.66Ec - 0.877 where Ec is the volume of these pores divided by the total volume of the system. The results point out that there are some active sites on pore walls where CaCO3 first nucleates and the Ca++ near the pore moves toward these sites as the CaCO3 crystals grow.

It has been stated that some structures failed due to the use of alumina cement. This failure was connected with the physical and chemical changes of concrete. The reason for this effect has been studied on the samples prepared from a concrete structure that collapsed suddenly after 30 years of use without any symptoms of defects. Various methods of examination were used, e.g., chemical and thermal analysis, X-ray diffractometry, scanning electron microscopy, besides mechanical tests. The failure was attributed to a combination of two main factors. First, the hydrated alumina cement was converted and then carbonated so that gibbsite and calcite, which have slight binding properties, were formed. The highly converted and carbonated concrete lost considerable strength and could not sustain the stress in the construction.

Expansion of the concrete at Center Hill Dam during its 38-year service life resulted in binding of the spillway gates and closing of the expansion joints in the bridge spans in 1983. An extensive laboratory and field investigation identified an alkali-carbonate rock reaction as the cause of expansion. Shortening of the spillway gates and bridge spans in 1985 remedied the operational deficiencies.

Different amounts of fly ash, silica fume, or ground granulated iron blast furnace slag have been used with portland cement in an effort to prevent excessive expansion due to alkali-silica reaction or sulfate attack or both. The test methods used to establish optimum proportions were ASTM C 441 and ASTM C 1012. Once the optimum proportions were known, concrete mixtures were made and tested.

Alkali-aggregate reaction is essentially a reaction between the hydroxyl ions in the pore solution of a mortar or concrete and the siliceous (or other alkali-susceptible) minerals in the aggregate. Study of the pore solution chemistry of mature cement pastes, mortars, and concretes has been possible in recent years through the development of pore solution expression techniques. This paper reviews the progress that has been made in explaining the phenomena associated with alkali-aggregate reactions in terms of pore solution composition. In particular, consideration is given to the effects of alkali level and water content of the concrete on the severity of reaction, the role of alkalis in pulverized fuel ashes, granulated blast furnace slags, and other cement replacement materials in determining their effectiveness in preventing damage and the contribution to pore solution alkalinity made by salt contamination of aggregates and deicing salts.

The alkali reactivity of siliceous aggregates depends on the microstructural disorder of the aggregate, plus further factors that can increase the phenomenon. The use of infrared (IR) spectroscopy to quantify this disorder is suggested. A simple graphic elaboration of an IR spectrum allows one to determine a disorder coefficient Cd. This allows one to compare rapidly and with assurance the different degrees of disorder for a number of siliceous minerals. Using as reference some samples whose alkali reactivity has been determined in a different way, a Cd limit value has been fixed. For this Cd value, the disorder is such as to raise no doubt about the aggregate susceptibility to alkali-silica reaction. This method can also be applied to calcareous aggregates that include siliceous minerals; if these inclusions are more than 10 percent, it is possible to use the sample without exclusion of the calcareous matrix phase.

To use silica fume as a pozzolanic material for inhibiting alkali-silica expansion, the effects of various silica fumes on expansion of mortars containing Beltane opal were investigated. Different silica fumes were found to vary widely in their effect on the expansion of mortars. The properties of silica fume affecting alkali-silica expansion were explored. Pozzolanic activity of silica fumes was evaluated by measuring the amounts of calcium hydroxide consumed by pozzolanic reaction in silica fume-bearing cement pastes. The amounts of calcium hydroxide in portland cement-silica fume mixtures were determined by DSC-TG analysis. Pore solutions obtained from mortars containing three different silica fumes were also analyzed. The silica fume with the highest pozzolanic activity was the most effective in reducing alkali-silica expansion of mortars. However, it was found that reduction in expansion by the addition of silica fume was not necessarily in line with the amount of calcium hydroxide consumed as a whole. Although the concentrations of alkalis and OH- ions in the pore solutions in mortars were reduced to the same level by the addition of three different silica fumes, the reductions in expansions of the three silica fume-bearing mortars were greatly different from one another.

The effectiveness of the use of fly ash in concrete to reduce the damage to concrete due to alkali-aggregate reaction to acceptable levels was investigated. More than 1300 mortar bars were cast and tested according to the ASTM C 227 mortar bar test method, with 0, 17, 26, 34, 45, and 62 percent replacement of the volume of cement in the mixture with fly ash. The effect of silica fume in the mixture was compared to that of fly ash at 17, 34, and 45 percent cement replacement by volume. The variables studied included the type of aggregate, alkali content of the cement, type of pozzolan, percent of cement replacement, pH of mixing water, and blending of the cement with the fly ash. Both ASTM Class C and Class F fly ashes were investigated. The results indicate that the replacement of a portion of the volume of cement with an equivalent volume of fly ash tends to reduce the expansion caused by the reactivity between alkalies and reactive silica in the aggregate, provided the proper amount of cement is replaced. This amount appears to depend on the alkali content of the fly ash and on the chemical composition of the fly ash used, mainly calcium oxide content. For some high calcium fly ashes with more than 1.5 percent available alkali content, a "pessimum limit" was observed. Such a limit represents a percent replacement under which the addition of fly ash causes equal or greater expansions in the mortar bars than that of the mixture without fly ash, and above which the expansions are reduced. In some instances, more than one "pessimum limit" was detected for a given combination of materials.

The Bureau of Reclamation and Construction Technology Laboratories are conducting a joint program to study the effects of alkali-silica reactivity in concrete dams and to determine the remaining potential for further reactivity in the structures. The first phase of the study covers Coolidge Dam, near Globe, Arizona; Friant Dam, near Fresno, California; Matilija Dam, near Ventura, California; Parker Dan, near Lake Havasu City, Arizona; and Steward Mountain Dam, near Phoenix, Arizona. The three requirements for expansive alkali-silica reactivity are sufficient alkali, availability of moisture, and the presence of potentially reactive silica. The procedures used in this investigation include field measurements of the relative humidity of the concrete to determine if sufficient moisture is available to sustain a continued reaction, expansion measurements of cores immersed in water and in an NaOH solution, petrographic examination of the cores to identify reactive aggregate particles, and osmotic cell tests of aggregate particles to determine potential reactivity