Researchers' Opinion

  • Eric Giannini. University of Alabama
  • There is a disconnect between the knowledge that has been generated by recent, and even not-so-recent, research on AAR, and implementation in engineering practice. In the United States, the construction specifications used by many state transportation agencies still use the accelerated mortar bar test (AMBT / ASTM C1260) or even the older mortar bar test (ASTM C227) to test for aggregate reactivity. Only a few have adopted the concrete prism test (CPT / ASTM C1293). The common sentiment is that the CPT takes too long. Exposure site testing is rarer still, but new sites have been established in the states of Oregon and Wyoming, in addition to the large exposure sites in Texas. Given that most aggregate quarries and pits will be operated for many years, it would be wise to require testing of materials that are expected to be exploited in future years using longer duration tests such as the CPT and outdoor exposure specimens. This would allow assessment of the reactivity and potential effective mitigation measures for these materials well in advance of their use in construction.
    Two important research needs going forward include the need to effectively assess the potential for and impact of internal alkali release from aggregates and SCMs, and a reassessment of accepted mitigation measures for ASR. The first need is of greatest importance for long-life structures, where alkalis released from aggregates or high-alkali SCMs may be sufficient to trigger expansive ASR in concrete otherwise thought to contain a sufficiently low alkali loading to prevent deleterious reactions. The RILEM AAR-8 method is a work in progress on this front. The second need ties in with the first, and the recent development of cracking of outdoor exposure specimens long thought to contain effective mitigation measures (sufficient SCMs or lithium) raises concern about what is truly needed for concrete to be resistant to ASR over the full 50+ year service life of a structure.

  • Isabel Fernandes. University of Lisbon
  • There are a number of well-known silica phases (opal, tridymite, chalcedony, cristobalite, Si-rich volcanic glass) considered to be alkali-reactive and for which the expansion tests nowadays accepted have given good and reliable results. In addition, the reactivity of cryptocrystalline quartz (e.g. in chert) is recognized both by petrography and by expansion tests. The main problem still to be solved, at least in certain areas of the world, is related with the identification of the characteristics of microcrystalline and strained quartz which originate slow/late ASR.
    Some studies have been carried out in order to find possible correlation between the intensity of deformation on rocks of granitic composition (e.g. granitic-gneiss and fault zones) and their reactivity.
    The studies developed so far revealed the importance of a detailed geological mapping of the quarries as the facies can vary in mineralogical composition and, most important, in texture. Petrographic assessment has shown that fault zones and variable strain degree originate higher content of microcrystalline quartz. The occurrence of subgraining, stretched crystals and ribbon quartz have been found to enhance the potential reactivity of the rocks. These features can occur side-by-side with medium to coarse grained innocuous granitic rocks and the different facies are mixed in the stock piles for concrete manufacture, in variable contents.
    Although the petrographic characterization is accepted as the first step in the assessment of the aggregates regarding ASR, the establishment of a quantified frontier between innocuous and reactive aggregates is lacking. It represents a subject of major interest that deserves to be studied in depth and may also contribute to the decrease in the subjectivity intrinsic to the petrographic studies.
    Regarding the laboratory expansion tests, the accelerated mortar bar tests have given false negative results for several slow reactive rocks. Different results have been found for concrete prism tests, namely for the accelerated concrete prism at 60ºC. The reason for this different performance is not yet well explained. It might be related with the destruction of the microcrystalline quartz during the crushing and sieving to obtain the granulometry sizes for the tests or it might have an unknown reason.
    Some suggestions have been made regarding the limits of expansion to be accepted for these rocks (e.g. use lower values of expansion) and/or the duration of the tests, namely in what regards the accelerated mortar bar test. However, this might be a useless change as the main advantage of the mortar bar test is to be a short term test.
    For the reasons above, slow/late reactive rocks represent a challenge for ASR researchers which needs further studies and clarifications.

  • Leandro Sanchez. University of Ottawa
  • Civil infrastructures are very critical for society, connecting the nation’s businesses, communities and people, driving the economy and improving the quality of life of human beings. It is known that several critical aging infrastructures are reaching their expected service life and thus actions should be taken in order to either manage/guarantee their performance over their last few years in service or expand their expected life span. Moreover, several aging infrastructures present clear signs of distress due to numerous damage mechanisms (e.g. alkali-aggregate reaction, delayed ettringite formation, sulfate attack, steel corrosion, freezing and thawing cycles, etc.), which might very likely decrease their performance down to unacceptable thresholds. In this regard, one of the biggest challenges in dealing with critical aging/deteriorating concrete infrastructures is to identify the cause of distress, to establish the correlation between the reductions in mechanical properties, physical integrity, durability and performance of the affected material, to evaluate possible structural implications and also their potential for further deterioration. Those are critical steps in the selection of management actions on the above structures.
    In this context, I believe it is very important to improve the state of art of various techniques used to either appraise the actual damage degree or the potential for further distress of aging infrastructures affected by alkali-aggregate reaction (AAR), which is one of the most harmful deterioration process that affects concrete structures worldwide, and other correlated mechanisms such as delayed ettringite formation (DEF) freezing and thawing (FT), etc.

  • Nicole Pagan Hasparyk. ELETROBRAS Furnas; Arnaldo Battagin. ABCP. BRAZIL
  • ABNT NBR 15577 is the Brazilian standard that prescribes the evaluation of potential reactivity of aggregates as well as the preventive measures to avoid damage in concrete structures due to ASR. An initial approach on the risk of occurrence of ASR is made. Once defined the level of risk that is acceptable the next step is the adoption of appropriate preventive measures, including laboratory tests.
    Despite its limitation, AMBT, one of test methods recommended, has shown good results for majority of rocks tested in Brazil. It has the advantage to be a rapid test method compared to CPT, that takes a long time and in the majority is not used due to the fact it can compromise the schedule and progress of construction works. However, sometimes failures occur in AMBT test results, especially for some slow reactive granitic rocks.
    In order to contribute to prevent “false negative results” and specify a national expansion threshold for AMBT, some Brazilian researchers grouped several data from this test (more than 500 samples analyzed) in order to evaluate the behavior of general Brazilian rocks and their slow reactive character. These studies have shown that the limit below of 0.10% expansion at 16 days to classify an aggregate as non-reactive or more than 0.20% to classify as reactive is not adequate to define the reactivity of our national rocks. Due to their late ASR, in majority, extending the test up to 30 days has been shown an effective measure in the order to recognize slow rate reactive aggregates.
  • In fact, the standard limit was changed and an expansion value of below 0.19 % at 30days was adopted in AMBT national standard. This change has allowed more confident results although it seems conservative. Finally, it is important to emphasize that CPT test results are mandatory for the decision of use of crushed rocks aggregates.


  • Stéphane Multon. INSA Toulouse
  • The prediction of future behaviour of structures damaged by ASR has to be established on a better understanding of all the phenomena occurring during the concrete degradation. Currently, the analysis is mainly performed from expansion tests on laboratory specimens. But, using expansions test on specimens to represent expansion of massive structures is still in question: expansions on structures appear to be often longer and larger and these differences cannot be only explained by temperature and moisture effects. New experimental and modeling works on the impacts of moisture and alkali transfers and the effect of sizes (aggregate and specimens sizes) on expansion should be performed to understand all these scale effects which lead to obtain smaller expansions on smaller specimens. Development of modeling combining physicochemical and mechanics are still necessary to explain the differences between expansion measured on specimens and on structures. The quantification of the different scale effects on ASR expansion is important to analyze the results of expansion test performed in laboratory and should be very useful to future diagnosis of ASR-damaged structures.


  • Victor Saouma. University of Colorado. USA
  • Despite the multitude of expansion tests across the world, there is not a centralized data repository accessible to researchers.
    To address such a need, a pilot project has recently been launched( It will enable researchers to upload their data, and for others to search the data base.