Abstract
Of the three known types of alkali-aggregate reactions, the alkali-silica reaction is considered to be the most damaging to concrete.
The earliest study in Japan on the alkali-silica reaction was by Fujii (1950), and in the past few decades his work has been followed by a number of reports on this aggregate reaction. The present paper aims to review these studies and survey the current thinking concerning alkali-silica reactive aggregates and methods of testing them regarding their reactivity.
Most of these studies share the view that andesitic rock are generally more liable to the alkali-silica reaction than other types of rocks.
Iiyama (1987) and other authors attributed this type of reactivity to, among others, chart and sandstone. Furthermore, it is known that rocks of the same type can show varying degrees of alkali reactivity depending on their specific formative origins.
Therefore, aggregate rocks to be used for concrete must be thoroughly tested individually for their specific reactive properties.
Standard tests of alkali-aggregate reactivity currently used are the chemical method and the mortar bar method. However, these types of tests are not without problems such as the reliability of reactivity judgments made by means of them and/or the length of time that they will take before the test data are obtained. To deal with these problems, Tamura (1986) and Houzyo (1996) each proposed a special form of testing experiment devised to accelerate possible alkali-aggregate reactions and enable quicker obtainment of the test results.
The researchers' goal at present is to develop optimum testing methods which will not only yield the test results in the shortest time possible but also enhance the reliability of judgment.
