The potential for alkali-reactivity of an aggregate and its actual reactivity as seen in practice can often be detected by a detailed petrographic examination where the standard chemical and physical tests provide ambiguous results.
French (this volume) comments that the standard mortar bar tests (ASTM-C227-71) may not always lead to accurate diagnosis of the likely performance of an aggregate. In practice, misunderstandings often arise over the application of the test to carbonate or impure carbonate aggregates or to other rocks lacking the specifically alkali-silica reactive minerals. Problems in interpretation also arise where the mortar bar results show progressive expansions that are close to or below the failure criteria suggested by ASTM C227-71. Silicate rocks showing slow expansion rates are now recognized, including Gillott's (1975) Type III reactions and marginal results of the mortar bar test cannot therefore be regarded as unimportant. The consideration of both these problems can be greatly assisted by petrographic examination which may allow proper diagnosis of the type of reaction that is likely to occur even though the rate may not be predictable.
Sometimes a petrographic examination identifies the principal cause of deterioration or failure of concrete. In one British example, unreinforced concrete bases were subject to cracking, and alkali-silica reactivity was established by petrographic analysis as the major mechanism involved (Fig. 1A and B). The coarse aggregate was an inert limestone, the cement was a high-alkali variety, and the sand was mostly quartz except for alkali-reactive chert which occurred in the coarsest sand fraction. The