Validation of a New Limit in Mortar Bars for Spanish Granitic Aggregates

The study of structures with alkali-silica reaction problem carried out in CEDEX (Spain) has identified up to seven dams affected by slow expansive processes produced by granitic rocks. Although the reactivity of this aggregates has been demonstrated in real works, the standard tests are not capable of detecting them with the limits proposed in the accelerated mortar bar test (AMBT). To avoid this problem, a new limit is recommended in Spain for the AMBT, when there is a possibility that the aggregate is slow-reactive. This communication presents the study of a dam made with granitic aggregates and unevenly affected by an alkali-silica reaction: no signs of damage to the concrete appear on its right bank versus intense cracking on its left bank. The detailed study of the aggregates used in different parts of the dam (RILEM petrography and AMBT) has made it possible to verify that the damaged areas correspond to the use of slow-reacting aggregates. In contrast, the aggregates are classified as innocuous in both methods where no damage is apparent. Therefore, this study has made possible to verify, within the same structure, the validity of the proposed limit for slow aggregates in the AMBT, and to corroborate how microfractured and deformed quartz produces expansion in the slow-late reaction. Finally, the analysis of the construction process of the dam and the observed distribution of damage has made it possible to demonstrate how important is to evaluate the reactivity of aggregates as quarry faces progress.