The performance of Irish aggregates in the 'ultra-accelerated test' for alkali-aggregate reactivity with particular reference to chert content

Irish aggregates have a satisfactory history of service regarding alkali-aggregate reaction despite containing significant amounts of chert and use for a period in conjunction with cements of high alkali level. An investigation was conducted of the performance in the draft RILEM ‘ultra-accelerated mortar-bar test’ of 23 individual aggregates and 14 combinations found in Irish practice. An X-ray diffraction study was also carried out. Definitive expansion limits have yet to be finalised but a value of 0.15% after 14 days immersion has been suggested as the innocuous/deleterious threshold and 0.25% expansion is suggested as the lower limit for reactive aggregates. Despite their satisfactory history of use, almost half of the aggregates tested individually and one third of the combinations exceeded the innocuous/deleterious threshold value. However, the lower limit for reactive aggregates was reached by only four individual aggregates and by one of the combinations. The X-ray diffraction study revealed the presence of well-crystallised quartz, confirming the likelihood that the aggregates would not be reactive. The highest expansions were recorded in samples from argillaceous limestones and those containing small amounts of greywacke and chalcedonic chert. As expected, there was no correlation between chert content and degree of expansion. It is concluded that most Irish cherts are unreactive; that the draft mortar-bar test returns pessimistic results and is therefore a suitable screening test but could not be used universally for definitive assessment unless national expansion limits could be agreed in the place of use; and that X-ray diffractometry may be used in conjunction with the rapid screening test to assist early judgement on likely reactivity.