PSI - Issue 37

João Custódio et al. / Procedia Structural Integrity 37 (2022) 590–597

595

6

João Custódio et al. / Structural Integrity Procedia 00 (2019) 000 – 000

Nevertheless, it should be pointed out that, in the period assessed, the expansion did not appear to have stabilized for any of the evaluated aggregate samples. According to LNEC Specification E 461, in the case of Class II and Class III aggregates for concrete to be used in dams, a structure categorised onto risk category R3 – high risk ( e.g. , long service life or critical structures where the risk of deterioration from ASR damage is judged unacceptable) and onto service environment category E2 (concrete exposed to extraneous moisture), results in the level of prevention P3 (special level). In this level of prevention, P3, at least two of the preventive measures M1 (restricting pore solution alkalinity), M2 (ensuring the use of a non-reactive aggregate or aggregate combination), M3 (reducing moisture ingress to maintain the concrete in a sufficiently dry state to prevent deleterious expansion of the gel) or M4 (modifying the properties of the ASR gel so that it becomes non-expansive) must be applied. In the level of prevention P3, in the case of not being possible to use two of the measures M1 to M4, as is the present case, then, one measure shall be used but the concrete composition must be classified as having a very low potential for deleterious ASR (section §4.5 of LNEC Specification E 461). 4.2. Aggregate potential alkali-reactivity – Petrographic examination The macroscopic observation and the characterization of the rock(s) to be used as aggregate is being performed following the procedures currently used in petrography (identification the main and accessory constituents) complemented by specific information able to predict the behaviour of those minerals when subjected to aggressive solutions. Alkaline chemical solutions are able to enhance selective leaching of some constituents, namely silicon, sodium and potassium as the most relevant for the development of ARS. Besides mineral composition, information on the characteristics and size of crystals (particularly those considered potentially reactive), the presence of cracks, signs of deformation, existence and extent of alteration minerals that affect the primary phases of the rocks are examples of the aspects to be observed, in particular when the aggregates are granitic rocks (or similar). The several varieties of granitic rocks tested here (BHS, DHS, AHS) have great similarities concerning mineralogical composition. They are granular, leucocratic, grey coloured, medium grained size rocks. Quartz and potassium feldspars are the main mineral components (Fig. 2). They also include plagioclase in several amounts. As accessory minerals, biotite and muscovite (usually in lower amounts) can be found. In general terms they can be defined as biotite granites or biotite-muscovite granites. However, it should be mentioned that FHS aggregate includes great amounts of a charnockitic type granite, an unusual (melanocratic) rock not found in Portugal. Very often, the aggregate includes fragments of the same granite but expressing different levels of decay (as it is the case in BHS aggregate where “rose” and “grey” varieties are present) .

BHS (PPL view) DHS (PPL view) Fig. 2. The most representative minerals present in the granite varieties under study. Main components: Qz - quartz; Fk- potassium feldspar; P- plagioclase. Accessory minerals: B, Bcl - biotite, chloritized biotite; M - muscovite. All these varieties include mineral features that must be considered for their potential alkali-reactivity assessment, concerning not only silicon release but also alkalis release (Fig. 3). Despite de definition of the concept for “reactive forms of silica”, signs of deformation can be detected on different minerals besides quartz, quite often also in micas, as it is the case of DHS coarse aggregate. The presence of strained quartz identified by undulatory extinction, irregular EHS (XPL view)