- Email: [email protected]
Properties and durability of coarse igneous rock aggregates and concretes
Construction and Building Materials 126 (2016) 119–129
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Properties and durability of coarse igneous rock aggregates and concretes Wojciech Piasta a,⇑, Jacek Góra b, Tadeusz Turkiewicz c a
Faculty of Civil Engineering and Architecture, Kielce University of Technology, Tysiaclecia PP 7, 25-314 Kielce, Poland Faculty of Civil Engineering and Architecture, Lublin University of Technology, Nadbystrzycka 40, 20-618 Lublin, Poland c The Institute of Technical Sciences and Aviation, The State School of Higher Education in Chelm, Pocztowa 54, 22-100 Chelm, Poland b
h i g h l i g h t s Crushing strength is helpful at assessing frost durability of igneous rock aggregate. Igneous rock aggregate with weathered minerals is freeze-thaw and alkali sensitive. Strained quartz was the alkali reactive phase in granodiorite; chalcedony in basalt. Alkali nepheline basalt showed best properties and durability.
a r t i c l e
i n f o
Article history: Received 8 July 2016 Received in revised form 5 September 2016 Accepted 8 September 2016
Keywords: Concrete Coarse aggregate Igneous rock Alkali-silica reaction Freeze-thaw resistance Aggregate crushing value
a b s t r a c t Direct tests allowed determining properties of coarse igneous rock aggregates in comparison with gravel along with freeze-thaw resistance and susceptibility to alkali-aggregate reaction. The relationship between the properties and durability of the aggregates and concretes was evaluated taking into account the results from microstructural analyses. Granodiorite, basalt and gravel were sensitive to alkali-silica reaction due to the presence of strained quartz, chalcedony and opal, respectively. The resistance of the aggregate to freezing and thawing was demonstrated to agree with the values of crushing resistance and the lowest contents of pores with diameters unsafe in terms of the freeze-thaw resistance. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction Concrete structures exposed to severe environment require the applied concrete to be not only of good mechanical properties but also of high durability. In moderate and subarctic climate zones and wet conditions, the durability of concrete is strongly dependent on the resistance of coarse aggregate to freezing and thawing. In wet environments, the durability of structures is also related to the aggregate sensitivity to detrimental internal physical and chemical processes as for instance the alkaline pore solution present in the cement paste [1]. The structure of igneous rock capillary porosity is different from that of the majority of sedimentary rocks. The system and structure of pore spaces in igneous rocks are more random and ⇑ Corresponding author. E-mail addresses: [email protected] (W. Piasta), [email protected] (J. Góra), [email protected] (T. Turkiewicz). http://dx.doi.org/10.1016/j.conbuildmat.2016.09.022 0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.
attributed to their genesis. Occasionally, minerals with capillary porosity may occur locally due to their partial weathering. A characteristic arrangement of microcracks or a tendency towards their formation due to the microstructure and mineral composition may be important factors for the properties of many igneous rocks, and thus in part, aggregates. In addition to the pores, primary microcracks may contribute to the flow of water, as may the aggregate grains mechanically induced during the aggregate production process. The formation of new microcracks in the grains has been confirmed in the studies which demonstrated an increase in water absorption and reduction in bulk density of aggregates even up to about 10% relative to the rock raw material [2–4]. Uncontrolled microcracks form when the rocks, after being crushed during extraction, are fragmented several times in a production process until an adequate particle fraction is obtained. The authors of this paper suggest that analyses and evaluations predicting the freeze-thaw durability of igneous rock aggregate should also allow for its resistance to cracking under mechanical
120
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
load. This parameter is dependent on the mineral content, mineral microstructure and microcracks in the aggregate, not only on the pore size distribution and porosity, which usually, though not in the case of these rocks, may play a dominant role. Ukrainian igneous rock aggregates have been readily exported and used in some European countries in the past few years. To evaluate their suitability, our detailed studies were performed of the properties and durability of these aggregates, including their application to concrete. The outcome of the studies should indentify the igneous rock aggregates that are suitable for the use at a broader level, also in the structures exposed to severe environments.
2. Literature survey Limiting the considerations about the freeze-thaw resistance of igneous rock aggregate to the porosity-dependent degree to which the water fills the pores or to water absorption level, even when the pore size distribution is taken into account, may be insufficient, as the effect of the rock matrix strength, dependent on its mineral composition, mineral structure and, in part, the microcracks, is neglected. Cyclic freezing and thawing makes the occurring stresses exceed the strength limit of the rock material leading to grain cracking – the final effect of freeze-thaw damage. This fact is used by standard testing methods for the resistance of aggregate to freezing and thawing [5,6]. Martinez-Martinez et al. [7] observed rock microstructures and reported that during the cycles of freezing and thawing, localized damage to the structure is initiated ‘‘as inter-particle-mickrocracking”. Then the propagation of this local mickrocracking is stopped for a number of cycles. During the stable period, isolated microcracks appear from where new ones nucleate and grow. When a critical threshold is exceeded, microcracks turn into cracks and grow rapidly, causing ultimate failure of rock after a few cycles [7]. Huge mineral and structural diversity of rocks hinders proposing an universal scheme and rules of cracking during freeze-thaw deterioration. There are opinions that freeze-thaw durability of rocks is not related to their origin, composition or mineral content, crystal size or water absorption level [8]. The deterioration of concrete in the structure may be caused not only by external actions (loads, temperature and chemical attack) but also by internal physical and chemical processes, in which the share of the environment is negligible. These deleterious processes include alkali-aggregate reaction (AAR) due to the influence of alkali present in the concrete pore solution on certain phases in the aggregate. The reaction occurs when the air relative humidity exceeds 80%. The cement clinker is the main alkali source, but they can also come from the aggregate, for example, from granite [9], or from the external environment. The alkali-silica reaction (ASR) product is a calcium-alkali-silicate-hydrate (C-N(K)-S-H) gel or an alkali-silicate-hydrate (N(K)-S-H) [10] gel, which is precipitated in the aggregate-cement paste interfacial transition zone, paste microcracks or aggregate microcracks. By absorbing water and swelling, the gel induces internal stresses that cause expansion and micro- and macrocracking in the paste, which gradually propagates in the concrete across the entire structural element. From the structural standpoint, AAR and microcracks, which extend into cracks and increase the transport of mass in concrete, can be a serious problem, as they open access for air and water or an aggressive agent to the steel reinforcement thus accelerating rebar corrosion substantially. Concrete mechanical properties worsen with progressing AAR. Ahmed et al. [11] and Giaccio et al. [12] reported a significant drop in the tensile strength of concrete containing reactive aggregate in the early stage of ASR, even though no noticeable reduction in compressive strength of concrete was observed at that time. At that ASR stage, elastic strains increase
in concrete along with gel formation and first microcrack formation, reducing the static modulus of elasticity of concrete [11–13]. AAR was observed in various elements of bridges and marine structures. The reaction was first identified by Stanton [14] in the concrete element of a bridge. Lukschová et al. [15] reported concrete defects in 13 bridges due to the swelling of the ASR gel product. The bridge concrete was produced with coarse basalt, granite and diorite aggregates. The gel formed as a result of the reaction with the grains of quartzite, monomineral quartz or greywacke present in fine aggregates. The basalt, granite and diorite aggregates did not react with alkali [15]. Shayan and Lancucki [16] reported ASR of the apparently non-reactive granite aggregate slowly progressing in the concrete of the bridge. Fifteen years after construction first microcracks appeared with alkali silicate gel formed in the cracks around the granite aggregate particles detected after 36 years. Sibbik and Page [17] also observed a slow deterioration of concrete with granite aggregate. Ponce and Batic [18] think that many granite and granodiorite aggregates can be classified as aggregates reacting very slowly and showing no signs of ASR in standard tests. This slow reaction can be associated with the presence of cryptocrystalline or strained quartz in granite. Basalt as the extrusive rock containing no quartz or minor quantities of quartz is recommended as the aggregate nonreactive to alkalis. However, some basalts can be reactive. Katayama et al. [19] showed that when the SiO2 content in the chemical composition of basalt exceeds 50%, it might be suspected to be potentially reactive to alkalis, nearly as reactive as andesite aggregate. Tiecher et al. [20] found that the amount of silica in volcanic rocks used to produce aggregates was the critical quantity for predicting their reactivity to alkalis. The reactivity is attributed to amorphous volcanic glass present in the interstices between the grains of the matrix. This material is crypto-microcrystalline and made up of quartz, silica, apatite and hematite [20]. ASR may affect various sedimentary but also plutonic and volcanic rocks including granite and basalt, which are apparently nonreactive. But in the case of igneous rocks, ASR proceeds very slowly due to their tightness and the special forms of silica. Therefore the aggregates from this rock material require reliable and long-term tests, as the majority of the aggregates react slowly. 3. Materials and methods 3.1. Aim and scope of the experiment The aim of this study was to investigate and evaluate the properties of coarse crushed aggregates derived from igneous rocks, plutonic and volcanic, used for constructing concrete structures exposed to severe environments. The same investigations were conducted with the postglacial gravel for comparison. The evaluation included the use of direct test methods for aggregate characteristics and for the properties of concretes formed with these aggregates. In addition to mechanical properties of the concrete and aggregate, the properties responsible for the durability of structures in severe climatic conditions were also studied. To explain the differences between the performance of individual aggregate types and concretes, the results of the tests and microstructural analyses are presented in the paper. 3.2. Materials The tests were performed on five types of crushed aggregate derived from igneous rocks and, for comparison, one type of gravel aggregate, along with six concretes made with these aggregates. Three aggregate types came from western Ukraine – basalt from Iwaniczi (Buk), granite from Vyrivs‘kyi karier (GRuk) and
121
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
granodiorite from Klesov (GDuk) – and two from south-western Poland – basalt from Winna Gora near Piotrowice (Bpl) and granite from Siedlimowice near Strzegom (GRpl), plus natural postglacial gravel from Sokolka near Suwalki (Gpl) in north-eastern Poland. Rinsed natural quartz sand from Sokolka was used in the concretes. The gravity of the sand was 2.64 kg/dm3, and the content of mineral dust (less than 0.063 mm in diameter) was 0.43%. The particle size gradation of the fine and coarse aggregates is summarised in Table 1. The concretes with all aggregate types were made with Portland cement CEM I 42,5 R. Chemical composition of the cement clinker, tested in accordance with PN-EN 196-2 [21], is presented in Table 2. Physical properties of the cement are shown in Table 3. In order to ensure reliable comparison of the effects of aggregate type on the properties of concrete, the principle of maintaining the same volume of aggregate in 1 m3 of the mixture was adopted. Thus, the contents of cement, water and sand were unchanged in all mixtures. All concrete mixtures had the same w/c ratio equal to 0.55. Table 4 shows the composition of the mixtures. No chemical additions or mineral additives were used in the mixtures. The indicator of the concrete mixture consistency, the slump was 9 ± 3 cm, which corresponds to S2 S3 classes of slump [22].
3.3. Methods 3.3.1. Test methods for aggregate Each size fraction was tested for crushing strength, water absorption (wt.), bulk gravity and specific gravity as the basis for the determination of the total porosity. The other tests focused on the resistance to freezing and thawing, the potential alkali reactivity and the microstructure of the aggregates. The crushing strength of the aggregate was determined using the aggregate crushing value (ACV) to PN-B-06714-40 [23], in compliance with the British Standard BS 812 [24], but with certain modifications. The aggregate sample placed in a steel cylinder 150 mm nominal internal diameter was loaded with a force of 200 kN. The aggregate crushing value is the ratio of the mass of fines (less than 1/4 of the lower sieve of a given size fraction) formed by the crushing process to the total mass of the sample expressed as a percentage. The ACV can be used for the classification of aggregate and evaluation of its suitability for given concretes. The test is applicable for 4 8 mm, 8 16 mm and 16 31.5 mm size fractions. Two methods were used to study the water absorption of each fraction sample. In the first method the absorption was determined at the atmospheric pressure [25]. Clean and dry aggregates were
Table 1 Sieve analysis of the fine and coarse aggregates, %. Sieve size, mm
Sand
31.5 16 8 4 2 1 0.5 0.25 0.125
Coarse aggregate
100 100 100 100 98 78 46 12 2
Bpl
Buk
GRpl
GRuk
GDuk
Gpl
100 95 51 9 1 – – – –
100 100 48 11 2 – – – –
100 97 51 15 3 – – – –
100 90 49 13 3 – – – –
100 100 50 11 1 – – – –
100 100 51 12 2 – – – –
Table 2 Chemical composition and alkali content in cement clinker, %. Compound
CaO
SiO2
Al2O3
Fe2O3
MgO
Na2O
K2O
Na2Oe
Content
63.97
20.71
4.35
3.46
1.48
0.26
0.50
0.66
Table 3 Physical properties of cement.
**
Le Chatelier, mm
Spec. surface, cm2/g
Spec. gravity, kg/dm3
Initial setting time, min.
Heat of hydr, J/g**
2-day compr. strength, MPa
28-day compr. strength MPa
0.9
3790
3.08
160
306
29.1
54.3
Heat of hydration measured after 41 h with the use of semiadiabatic method.
Table 4 Composition of concrete mixtures. Concrete components
Designation of concrete according to aggregate type used C-Bpl
Cement, kg/m3 Aggregate 2 8 mm, kg/m3 Aggregate 8 16 mm, kg/m3 Sand, kg/m3 Water, dm3/m3
350 644 644 681 193
C-Buk
C-GRpl
C-GRuk
C-GDuk
C-Gpl
626 626
569 569
580 580
576 576
576 576
122
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
Table 5 Test results for coarse aggregates. Properties
Fraction (mm–mm)
Bulk specific gravity, kg/dm3 Specific gravity, kg/dm3 Total porosity, % Crushing value, %
– – – (4–8) (8–16) (4–8) (8–16) (4–8) (8–16) (4–8) (8–16) (4–8) (8–16)
Absorption under atmospheric pressure, % Absorption under vacuum, % Freeze-thaw resistance to PN-B-06714-19, % Freeze-thaw resistance to EN-1367-1, %
Coarse aggregate Bpl
Buk
GRpl
GRuk
GDuk
Gpl
2.96 3.02 1.99 4.3 4.9 0.44 0.32 0,51 0.45 0.2 0.25 0.1 0.15
2.84 2.99 5.02 3.9 8.5 1.45 1.31 1.57 1.48 3.6 4.0 0.6 0.7
2.62 2.67 1.87 14.9 15.3 0.47 0.32 0.60 0.52 1.4 3.7 0.3 0.45
2.64 2.67 1.12 5.9 9.8 0.31 0.18 0.39 0.31 2.0 2.7 0.2 0.4
2.63 2.68 1.87 6.2 5.8 0.34 0.26 0.47 0.38 0.8 1.2 0.1 0.3
2.63 2.71 2.95 7.5 10.6 0.98 0.76 1.09 1.05 3.6 4.5 1.6 2.0
4.6 0.38 0.48 0.23 0.13
weighed and submerged in water up to 1/2 of the height of the smallest particle. After three hours, the water level was raised to 2/3 of the largest particle height and after another three hours the water level was not lower than 2 cm above the largest particle. Forty-eight hours from the start of saturation, the aggregate was surface dried and weighed. The second procedure involved water saturation under vacuum. Dry aggregates were stored under the pressure lowered to 4 kPa for 1 h and then saturated in water for 3 days. Table 5 summarises the absorption results from both test methods. The specific gravity was measured with the use of a pycnometer after the fragmentation of the aggregate to sizes less than 0.08 mm. The air contained between the grains of the powdered material was removed by placing the pycnometer in the vacuum chamber and by lowering the pressure to 2.33 kPa. The apparent specific gravity was determined to EN 1097-6 [26]. The total porosity was calculated based on the values of apparent and specific gravities. The resistance to freezing and thawing of the aggregate was investigated by subjecting each size fraction to 80 freeze-thaw cycles after optimal water saturation, as during the absorption test after the vacuum saturation. The number of freeze-thaw cycles was increased relative to the standard recommendations in order to obtain larger differences in the mass of grains crushed. After freeze-thaw cycles, the damage to the aggregate grains was assessed. For that purpose, the relative quantity of grains crushed was determined through the cycles of freezing and thawing – in accordance with PN-B-06714-19 [5] – crushed to the size less than the lower sieve and – in accordance to EN-1367-1 [6] – crushed to the size less than 1/2 of the lower sieve for the given size fraction. The potential alkali reactivity of the aggregates was determined using the ASTM C–1260 [27] method with six mortar bars 25 25 250 mm of each aggregate type. To explain the relationships and changes observed during the durability tests, the aggregates and mortars were subjected to microstructural analysis. The ordinary petrographic examination was carried out on the samples from the thickest aggregate grains. The standard thin sections were examined with a petrographic microscope in the polarized light and fluorescence modes. The crystalline phases in the aggregates were identified by the XRD analysis. The AutoPore IV 9500 mercury porosimeter was used to determine the pore size distribution. Alkali-aggregate reaction products were identified through the environmental SEM with EDXA examination of mortar microstructures (operating voltage: 20 kV; working distance 9–12 mm).
6.6 1.37 1.52 3.85 0.65
15.1 0.40 0.56 2.55 0.38
8.0 0.24 0.35 2.40 0.30
6.0 0.30 0.43 1.00 0.20
9.3 0.85 1.07 4.10 1.85
3.3.2. Test methods for concrete Twelve cubic specimens 100 mm were made from each concrete mix for compressive strength testing and six beams 100 100 500 mm for bending tests. The bending tests were carried out after 365 days of curing; the compressive tests were carried out after 28 and 365 days. The specimens were saturated in water up to 1/2 of their height after being placed on a grate 10 mm above the bottom of the bathtub. At 24 h the water level was raised to reach 10 mm above the specimen surfaces. After consecutive 24 h, they were removed, wiped dry and weighed. The specimens remained immersed in water until no increase in mass was recorded. After that, they were dried to a constant mass, weighed and the mass absorption was determined. To determine the resistance of concretes made with various aggregate types to freezing and thawing, the change in mass and compressive strength were studied on 10 cm cubic specimens. Six specimens for each concrete were formed and six were sawed from long beams 10 10 cm in cross-section. The cubes cut out from the beams showed the cut aggregate grains (not covered in cement paste) on two walls. Lack of air entraining modification allowed exposing and emphasizing the freeze-thaw action on the aggregate grains. The resistance to freezing and thawing was determined after saturating the specimens in water, as in absorption tests. The specimens were subjected to freezing at 20 °C for 4 h, and then to thawing in water for 4 h and heating to +20 °C. The parameters measured included mass loss and reduction in compressive strength relative to the reference specimens stored in water at +20 °C ± 2 °C for the entire testing period.
4. Test results 4.1. Test results of commonly measured aggregate properties Table 5 compiles the values of bulk specific gravity, specific gravity, total porosity, crushing value and water absorption of coarse aggregates. Two values are given if the tests were conducted on two aggregate fractions, with their weighted arithmetic mean shown next to them. Analysis of the test results of aggregates indicates relatively low bulk specific gravity of basalt Buk, which confirms the highest porosity of all aggregates studied and the highest water absorption. Granite GRpl has the highest crushing value, on average twice as high as that of the other aggregates. The results for the vacuum saturated aggregates show that basalt Buk has the highest water absorption, higher than that of
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
gravel Gpl despite the fact that most grains were derived from sedimentary rocks of capillary porosity. Aggregates GRuk and Gduk show the lowest levels of water absorption. The absorption results for atmospheric pressure and vacuumsaturated aggregates are in agreement, which confirms the same relations between absorption values of individual aggregates. The highest absorption level at atmospheric pressure saturation is also seen in basalt Buk and gravel Gpl. Absorptions of most of the other aggregates were in the range 0.3%–0.5%. With regard to both fractions, better freeze-thaw resistance was recorded for aggregate Bpl, while Buk was crushed most of all igneous rock aggregates during the cycles of freezing and thawing.
4.2. Results of the mercury porosimetry test of aggregates The differential distribution of pore sizes shown in Figs. 1 and 2 indicates that the most remarkable pore size distribution curves represent basalt Buk, granite GRpl and basalt Bpl. Basalt Buk has the largest number of pores with diameters smaller than 0.4 lm. Granite GRpl has a considerable number of pores with diameters in the range 0.1–1 lm, whereas in basalt Bpl, pore diameters over 0.1 mm and below 10 nm are most common. Similar pore size
123
distribution is seen in granite GRuk and granodiorite GDuk, the diameters being mostly from 0.005 to 0.2 lm, with fewer pores coarser than 0.1 mm in diameter. 4.3. Test results for aggregate microstructure 4.3.1. XRD analysis X-ray diffraction revealed the presence of the following crystalline minerals in the aggregates. Basalt Bpl: labradorite (plagioclase), pyroxene – in the form of augite (magnesium iron silicate), nepheline (sodium and potassium aluminosilicate) silica-undersaturated alkali feldspathoid, Basalt Buk: labradorite, augite, nontronite – clay mineral (isomorphic smectite with montmorillonite), Granite GRpl: quartz, microcline – potassium alkali feldspar (similar to orthoclase), albite (plagioclase), biotite, Granite GRuk: quartz, microcline, albite, biotite, Granodiorite GRDuk: quartz, microcline, albite, biotite, muscovite, Gravel Gpl: quartz, microcline, albite, calcite, dolomite, illite – clay mineral. 4.3.2. Polarizing petrographic microscope Additional petrographic analysis with the use of a polarized microscope revealed more details about mineral characteristics of Buk and GDuk aggregates, significant due to the highest expansion of mortars made with them observed in Ukrainian aggregates during the AAR tests. Basalt aggregate Buk is composed of plagioclases and pyroxenes as most important constituents, with a minor number of the minerals as a residue of partial weathering (Fig. 3). The rock does not contain any igneous amorphous glass, which initially filled the spaces between the plagioclases and pyroxenes and was finally replaced by microcrystalline smectite mineral and iron hydroxides (probably goethite). These minerals make up about 10% of the basalt volume. The analysis reveals some loose accumulations of silica in the form of fine spherical chalcedony grains 0.02– 0.03 mm in size, which is an important observation for the AAR test results. The accumulations of fine chalcedony grains are dispersed in the secondary material (among iron hydroxides and smectites) and do not form any larger conglomerates (Fig. 4).
Fig. 1. Total pore volume versus pore diameter.
Fig. 2. Logarithmic differential distribution of pore sizes.
Fig. 3. Basalt Buk. Bladed plagioclases (grey), isometric pyroxenes (white) and accumulations of clay minerals and iron hydroxides (brown). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
124
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
Fig. 4. Basalt Buk. Chalcedony in sparse accumulations are seen in the centre of smectite pseudomorphs. Smectite mineral and iron hydroxides between plagioclase blades (grey) and pyroxenes. Around the pyroxenes, opaque rims of iron hydroxides are seen.
Fig. 6. Biotite blades and clinozoisite-epidote minerals in the medium to finegrained granodiorite.
Granodiorite GDuk is composed of plagioclases, alkali feldspars, quartz, biotite and muscovite. Quartz is the only form of silica present, fine and coarse-grained in part, with irregular, wavy and jagged edges. These irregular edges of the larger particles suggest that in some cases these may be strained quartz grains (Figs. 5 and 6). The rock shows only a minor degree of secondary transformation. The changes are observed in the form of a weak sericitic alteration of plagioclases and presence of fine-grained accumulations of biotite blades and epidote group minerals. In addition to these constituents, the fine-grained variant contains sparse grains of opaque minerals (probably pyrite) and single particles of muscovite and garnet up to 0.5 mm in size (Fig. 6). 4.4. AAR results The ultrafast test [27] of linear strains in mortars revealed that expansion of the mortar with granodiorite aggregates was 0.13% (Fig. 7) after 14 days of soaking in a 1 N solution of NaOH at 80°C, exceeding the expansion limit, 0.1%, for potentially reactive aggregate. At 28 days, the expansion strain increased to 0.21%. Note that in the gravel-based mortar, expansion was 0.09% at
Fig. 5. Irregular wavy edges – strained quartz grains in a medium to fine-grained granodiorite.
Fig. 7. Results of expansion strain for mortars stored in a 1 N NaOH solution at 80 °C.
14 days, being at the boundary of strain level typical of mortars with potentially reactive aggregate. At 28 days the value rose to 0.15%. The mortars with Ukrainian and Polish basalts showed different behaviours in the NaOH solution. The mortar with basalt aggregate derived from Poland’s quarries was not subject to any expansion, which is characteristic of alkali basalts. The mortar with Ukrainian basalt aggregates showed the expansion of 0.07% at 28 days, which may seem interesting in terms of the causes of these varying behaviours of two aggregates derived from the same type of rock. After the test for Potential Alkali Silica Reactivity of Aggregates in the NaOH solution conducted, SEM and EDXA were used to study the microstructure of mortars. Fig. 8a shows a photograph of a relatively smooth surface of a gel which formed in the mortar with granodiorite aggregate. Microcracks are running across this relatively smooth surface. The chemical composition at point A, Fig. 8b, suggests that most likely it is calcium sodium silicate hydrate. Fig. 9a shows an image of the gel identified on the fracture surface of the specimen made with gravel aggregate and soaked in a
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
125
Fig. 8. Microstructure of mortar with granodiorite aggregate in scanning electron microscope: a) a layer of calcium sodium silicate b) chemical composition at point A by EDXA.
Fig. 9. Results of microstructural analysis of the mortar with gravel aggregate under a scanning electron microscope: a) calcium sodium silicate gel with cracking b) chemical composition at point B by EDXA.
1 N NaOH solution at 80 °C for 28 days. There is a ‘‘spider web” of microcracks on its surface. Some of these microcracks have large widths. The chemical composition determined on the surface at point B in Fig. 9b suggests that it is a calcium sodium silicate gel hydrate. 4.5. Test results of the strength of concretes Table 6 compiles compressive strengths f28d cm,cube at 28 days and f365d cm,cube at 365 days of curing of cubic specimens with a side length of 100 mm, cylindrical specimens f28d cm,cyl at 28 days and, determined
based on these results, compressive strength classes of the concretes. The table also shows flexural strengths f365d ct,flex of concretes at 365 days of curing. The results from the test for compressive strength f28d cm.cyl of concretes showed differences of 2 strength classes. The differences in flexural strength results f365d ct.flex of individual concretes were similar considering that according to PN-EN 1992-1-1:2008 [28] the difference between two consecutive classes is 0.3 MPa. In both cases best results were obtained for the concrete made with basalt Bpl. The lowest value of compressive strength and the lowest class were recorded for the concrete made with granodiorite GDuk. As
126
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
for the flexural strength f365d ct.flex, the lowest values were recorded for the concrete with granite GRpl and the Ukrainian granodiorite, at least 2 classes lower, when referred to the concrete strength classes, than the concrete with the Polish basalt aggregate. 4.6. Test results for freeze-thaw resistance and water absorption of concretes After 150 cycles of freezing and thawing, the decrease in mass of the moulded specimens of concrete C-GDuk with granodiorite (Fig. 10) was 7% (Table 7). The specimens were cracked locally. The mean strength was only 55%. Two of the remaining concretes were different from others: C-Gpl and C-Buk. The mass of the gravel concrete increased by about 3%, with the strength drops of nearly 40%. The mass of the concrete made with the Ukrainian basalt hardly changed (Fig. 10), with a minor only drop in strength (Table 7). Other concretes had similar low freeze-thaw resistance after 150 freezing and thawing cycles.
A mass gain of 2%–3% in the sawed specimens made from concretes with aggregates GDuk and Gpl for a few dozen cycles of freezing and thawing (Fig. 11) and then rapid mass loss, strength decline, cracking and gradual failure of the specimens confirm a typical pattern of frost related damage of low resistance concrete. The change in mass of the sawed specimens made of concretes C-GRuk, C-Buk and C-Bpl suggests better resistance to freezing and thawing. However, after 150 freezethaw cycles all concretes showed a considerable drop in strength (Table 7), and the majority of specimens were seriously cracked (Fig. 12). Higher strengths of concretes containing basalt aggregates and better condition of the specimens indicate their better resistance. The results (Table 8) show that the water absorption levels of the concretes with igneous rock aggregates were similar and ranged from 5.2 to 5.4%, except for aggregate Buk, for which the value was nearly 15% higher. Higher absorption was also recorded for the gravel concrete, but it was only 5–6% higher.
Table 6 Mean results of compressive strength and flexural strength and strength class of the concrete. Properties
Compr. strength f28d cm,cube, MPa Compr. strength f365d cm,cube, MPa Flexural strength f365d ct,flex, MPa Compr. strength f28d cm,cyl, MPa Strength class to EN 206-1
Concrete C-Bpl
C-Buk
C-GRpl
C-GRuk
C-GDuk
C-Gpl
47.1 51.5 5.20 44.2 C40/50
46.0 50.4 4.96 40.8 C35/45
45.1 51.1 4.95 39.7 C35/45
43.4 47.3 4.38 40.1 C35/45
43.4 48.5 4.60 37.9 C30/37
42.0 50.1 5.02 38.0 C30/37
Fig. 10. Average mass change of moulded specimens subjected to freezing and thawing.
Table 7 Mean compressive strengths after freeze-thaw cycles. Properties
* **
Concrete C-Bpl
C-Buk
C-GRpl
C-GRuk
C-GDuk
C-Gpl
Moulded specimens (m) Compr. strength after 150 cycles, % Reference strength, MPa
80 51.9
93 52.9
73 51.5
77 50.9
55 50.9
62 49.5
Sawed specimens (s) Compr. strength after 150 cycles, % Reference strength, MPa
67 49.6
75 45.9
60 50.6
70 48.3
13* 46.2
55** 48.1
After 120 cycles. After 80 cycles.
127
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
Fig. 11. Average mass change of sawed specimens subjected to freezing and thawing.
Fig. 12. Images of the sawed cubes of concretes after 150 cycles of freezing and thawing: a) C-GRuk b) C-GRpl c) C-GDuk d) C-Gpl.
Table 8 Water absorption by weight of concretes. Properties
Water absorption by weight, %
Concrete C-Bpl
C-Buk
C-GRpl
C-GRuk
C-GDuk
C-Gpl
5.2
6.1
5.2
5.4
5.3
5.6
5. Analysis and discussion 5.1. Freeze thaw resistance of aggregates and concretes Fig. 13 shows the results for three characteristics of aggregate fraction 8–16 mm: resistance to freezing and thawing to [5],
vacuum saturation absorption and the crushing value (ACV) to [23]. The freeze-thaw resistance of aggregates (Bpl and GDuk) was found to be highest when absorption and crushing value were low. It has to be noted that the values of ACV and freeze-thaw resistance presented the best agreement. Low freeze-thaw resistance (>3% crushed grains) was observed in Buk (4.0% crushed
128
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
Fig. 13. Dependence of the resistance to freezing and thawing on absorption levels and crushing value.
grains) and GRpl (3.7% crushed grains), but then the ACV of granite GRpl and absorption of Buk were the worst. The results show that in the case of igneous rocks, crushing strength is a more practical indicator of resistance to freezing and thawing than absorption is. The correlation between freeze-thaw resistance and absorption is also questioned by Rusin et al. [8]. In the case of the lowest freeze-thaw resistance of aggregates, there is a good correlation between their low absorption and high crushing strength. The best (Bpl) and worst (Buk) physical properties and freezethaw resistance of basalt aggregates are related to their mineral composition and microstructure. Basalt Bpl is an alkali rock with no trace of weathering and strong rock matrix (ACV = 4.6%). The differential distribution of pore sizes in basalt Bpl (Fig. 2) confirms the presence of two dominant pore groups – coarse pores with diameters larger than 0.1 mm, regarded as safe [29] and fine pores with diameters ranging from 5 nm to 10 nm. The high porosity of basalt Buk is associated with the presence of the minerals (Fig. 3) typical for sedimentary rocks, which indicates partial weathering. In addition to chalcedony, this basalt contains smectite (nontronite) with a laminar structure and ability to absorb water, which increases absorption levels. As a result of transformations, basalt Buk contains a large group of pores that are regarded unsafe in terms of freeze-thaw resistance [29], with diameters between 20 nm to 400 nm (Fig. 2). This group of pores is a major factor in high absorption and poorest freeze-thaw resistance of the basalt, despite its relatively strong matrix. In granite GRpl, the presence of dominant group of pores about 200 nm in diameter explains its low freeze-thaw resistance. The lowest freeze-thaw resistance of gravel Gpl is related to the origin of grains present in its composition. Most grains of the gravel resulted from the erosion of sedimentary rocks (e.g., limestones, dolomite limestones, sandstones with silica and carbonate cements containing small amounts of clay minerals) with capillary porosity, which usually causes higher absorption of the material. Analysing the freeze-thaw resistance of aggregates and looking for the influencing factors and their significance, it is important to remember that the deterioration (crushing) of aggregate grains is caused by the stress from increasing pressure due to water freezing in their pores and microcracks. The measurement of the material crushed by the cycles of freezing is required by standards for the evaluation of the freeze-thaw resistance of aggregates [5,6]. This requirement, met by the authors of this study, makes the freeze-thaw resistance evaluation dependent on the condition and mechanical properties of aggregates, while the strength characteristics can be estimated by assessing its crushing strength [23,24]. Analysis of the freeze-thaw resistance of aggregate only in the light of its absorption levels and porosity, even accounting for pore
geometries, can be difficult and insufficient. This approach ignores the effect of the rock matrix strength, which is dependent on its mineral composition and structures of the minerals. Therefore, for practical reasons it seems justified to facilitate the diagnosis by taking into account the strength of the grains represented by the ACV. Good behaviour of the concrete with aggregate Buk during the cycles of freezing can be surprising considering a relatively low freeze-thaw resistance of this aggregate. Two issues require attention. – (1) Basalt Buk contains a lot of pores, including coarse ones (Fig. 1) not fully filled with water in the freezing concrete, especially that dry aggregates were used in the concrete mix. These partially filled pores can take part of the water and reduce pressure in the aggregate grains and in the paste surrounding them, therefore protecting them against damage as air entrainment does. – (2) The dry absorbent aggregate in the surrounding paste reduces its w/c ratio, which has a critical influence on the frost protection of concrete. The average freeze-thaw resistance of concretes made with the other aggregate types was similar and resulted from the high w/c ratio and lack of air entrainment. The freeze-thaw resistance of all concretes in the moulded specimens was considerably higher than that in the sawed specimens which had aggregate grains cut and exposed on two sides. 5.2. Alkali-aggregate reactions The SEM images (Fig. 8a) and EDXA results (Fig. 8b) suggest that the large expansion of the mortar bars made with granodiorite aggregate was caused by the formation and swelling of the calcium sodium silicate gel. The XRD analysis and petrographic test results (Fig. 6) show that the aggregate contains mainly quartz, alkali feldspar, plagioclases and a little of biotite and muscovite. This gel is probably a product of the reaction between sodium and calcium hydroxides and quartz present in the granodiorite in the form of strained quartz (Fig. 5). Just as in the case of some granites [17,18], ASR of these granodiorite aggregate can be assumed to be associated with the presence of strained quartz grains or cryptocrystalline quartz which are thermodynamically unstable and have deformation-related high free energy in their crystalline networks. Although the expansion of the mortar with aggregate Buk (Fig. 7) in the NaOH solution is less than 0.1%, it differs from that of the mortar with basalt aggregate Bpl, which does not deform in any way for 28 days. This results from the fact that basalt aggregate Bpl is derived from an alkali basalt rock with a minor content of volcanic glass, shows no trace of weathering and contains silicaundersaturated minerals such as nepheline (alkali feldspathoid) identified with XRD analysis. The expansion of the mortar with basalt Buk was probably related to the presence of small aggregations of chalcedony (Fig. 4) built from cryptocrystalline silica. As commonly known, chalcedony reacts easily with sodium and calcium hydroxides. Alkali silica reaction of this basalt aggregate might be also partially related to the reactions between calcium and alkalis with nontronite (smectite isomorphic with montmorillonite) (Fig. 4). Because the basalt matrix is very strong and amounts of chalcedony and nontronite are small, the kinetics of the reactions and expansion are limited considerably. Reactions of the paste with smectites were also observed by Batic et al. [30], who focused on ASR of basalt containing montmorillonite. Accelerated expansion of the paste with gravel aggregate in the NaOH solution resulted from the formation and swelling of the calcium sodium silicate gel, as observed under scanning electron microscope on the specimen fracture surface (Fig. 9a and b). The formation of the gel was related to the presence in the gravel [31] of few grains of quartz-pyroxene slate with opal cement.
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
129
6. Conclusions
References
1. In addition to pore size distribution, the crushing strength together with water absorption can be a good and practical indicator of the resistance of crushed aggregate derived from igneous rocks to freezing and thawing. Absorption alone cannot be an independent indicator of freeze-thaw resistance of aggregate. For example, the highest freeze-thaw resistance of Polish basalt aggregate as well as Ukrainian granodiorite and granite aggregates corresponded to their high crushing strength, low water absorption and the lowest content of pores with diameters unsafe in terms of the aggregate freeze-thaw resistance. 2. The occurrence of any quantity of secondary minerals formed as a result of weathering in an igneous rock can increase porosity, leading to the weakening of strong rock matrix thus reducing its freeze-thaw resistance. In the case under investigation, low freeze-thaw resistance of the Ukrainian basalt matched its porosity, high absorption and moderately high crushing strength (ACV = 8.5% for 8–16 mm fraction). The results of microstructural analysis of this aggregate indicate that it was related to the presence of secondary minerals, such as smectites, chalcedony and other minerals formed during the weathering process. 3. The test results of mortar expansion in a NaOH solution indicated that granodiorite aggregate and gravel might be susceptible to ASR (potentially reactive). Sodium and calcium silicate gel was found to form in the mortars. In granodiorite, strained quartz is probably the reactive phase, and in the gravel – few grains containing opal cement or those with clay-limestone cements. The use of the Ukrainian basalt aggregate in concrete exposed to harsh environmental conditions is risky because of the detected presence of chalcedony and smectite group of minerals (nontronite). 4. Three out of six aggregate types studied – plutonic granodiorite, volcanic basalt and sedimentary sandstone (grains in the gravel) – were found to be potentially sensitive to ASR due to the presence of strained quartz, chalcedony and opal, respectively. It is thus absolutely necessary to subject each and every aggregate – also that derived from igneous rocks, with very good physical and mechanical properties – to tests for potential alkali reactivity. 5. Only two out of six aggregate types and concretes studied, Polish basalt and Ukrainian granite, were found to be versatile and suitable for a wide range of use, including the structures exposed to severe environment.
[1] T. Katayama, D. Bragg, Alkali-aggregate reaction combined with freeze/thaw in Newfoundland. Canada Petrography using EMPA, Proc. 10th ICAAR, Melbourne Australia, 1996, pp. 243–250. [2] T. Gawenda, Z. Naziemiec, T. Tumidajski, D. Saramak, Ways of optimization of grain composition and shape of grains in aggregates in products of jaw _ crushers, Górnictwo i Geoinzynieria 3/1 (2006) 109–124 (30). [3] J. Bromowicz, Changes of rock material properties during production process of crushed aggregates, Surowce i Maszyny Budowlane 4 (2007) 14–17. [4] Z. Naziemiec, Influence of Treatment Technology on Quality of Obtained Aggregate, 3, Scientific Works of Institute of Glass, Ceramics Refractory and Construction Materials, Warszawa-Opole, 2009. pp. 182–193. [5] PN-B-06714-19:1978, Mineral aggregates. Testings. Determination of resistance to frost by the direct method. [6] EN-1367-1:2007, Tests for thermal and weathering properties of aggregates. Determination of resistance to freezing and thawing. [7] J. Martínez-Martínez, D. Benavente, M. Gomez-Heras, L. Marco-Castaño, M.A. García-del-Cura, Non-linear decay of building stones during freeze–thaw weathering processes, Constr. Build. Mater. 38 (2013) 443–454. [8] Z. Rusin, P. S´wiercz, Z. Owsiak, Effect of microstructure on frost durability of rock in the context of diagnostic needs. 7th Scientific-Technical Conference Material Problems in Civil Engineering (MATBUD’2015), Proc. Eng. 108 (2015) 177–184. [9] Z. Owsiak, Contribution of alkali from aggregate to pore solution of concrete, Cem. Wapno Beton 4 (2001) 151–153. [10] L. Bainga, T. Nawa, E. Iwatsuki, T. Awamura, ASR formation of reactive chert in conducting model experiments at highly alkaline and temperature conditions, Constr. Build. Mater. 95 (2015) 820–831. [11] T. Ahmed, E. Burley, S. Rigden, A. Abu-Tair, The effect of alkali reaction on the mechanical properties of concrete, Constr. Build. Mater. 17 (2002) 123–144. [12] G. Giaccio, R. Zerbino, J.M. Ponce, O.R. Batic, Mechanical behavior of concretes damaged by alkali-silica reaction, Cem. Concr. Res. 38 (2008) 993–1004. [13] I. Yurtdas, D. Chen, D.W. Hu, J.F. Shao, Influence of alkali silica reaction (ASR) on mechanical properties of mortar, Constr. Build. Mater. 47 (2013) 165–174. [14] T.E. Stanton, Expansion of concrete through reaction between cement andaggregates, Proc. Am. Soc. Civil Eng. 66 (1940) 1781–1811. [15] Š. Lukschová, R. Prˇikryl, Z. Pertold, Petrographic identification of alkali–silica reactive aggregates in concrete from 20th century bridges, Constr. Build. Mater. 23 (2009) 734–741. [16] A. Shayan, C.J. Lancucki, Alkali-aggregate reaction in the causeway bridge. Perth, Australia, Proc. 7th Int. Conf., AAR in Concrete, Ottawa, 1986. pp. 392– 397. [17] R.G. Sibbick, C.L. Page, Susceptibility of various UK aggregates to alkali– aggregate reaction. Proc. 9th Int. Conf., AAR in Concrete, London, 1992. pp. 980–987. [18] J.M. Ponce, O.R. Batic, Different manifestations of the alkali-silica reaction in concrete according to the reaction kinetics of the reactive aggregate, Cem. Concr. Res. 36 (2006) 1148–1156. [19] T. Katayama et al., The petrographic comparison of rocks from Japan and New Zealand-Potential reactivity related to interstitial glass and silica minerals. 8th Int. Conf., AAR in Concrete, London, 1989. pp. 537–541. [20] F. Tiecher, D.C.C. Dal Molin, M.E.B. Gomes, N.P. Hasparyk, P.J.M. Monteiro, Influence of mesostasis in volcanic rocks on the alkali-aggregate reaction, Cement Concr. Compos. 34 (2012) 1130–1140. [21] PN-EN 196-2:2013–11, Method of testing cement – Part 2: Chemical analysis of cement. [22] EN 206:2013, Concrete. Specification, performance, production and conformity. [23] PN-B-06714-40:1978, Mineral aggregates. Testings. Determination of crushing strength. [24] BS 812-110:1990, Testing aggregates. Part 110: Methods for determination of aggregate crushing value (ACV). [25] PN-B-06714-18, Mineral aggregates. Testings. Determination of absorption. [26] EN 1097-6:2013, Tests for mechanical and physical properties of aggregates. Determination of particle density and water absorption. [27] ASTM C1260 Test Method for Potential Alkali Reactivity of Aggregates (Mortar-Bar Method). [28] PN-EN 1992-1-1:2008 Eurocode 2: Design of concrete structures – Part 1–1: General rules and rules for buildings. [29] Z. Rusin, Technology of Frost-Resistant Concretes, Polski Cement, Kraków, 2002. [30] O. Batic, P. Maiza, J. Sota, Alkali silica reaction in basaltic rocks NBRI Method, Cem. Concr. Res. 24 (1994) 1317–1326. [31] Z. Owsiak, J. Zapała, P. Czapik, Sources of the gravel aggregate reaction with alkalis in concrete, Cem. Wapno Beton 3 (2012) 149–154.
Acknowledgements This study was supported by the Kielce University of Technology and Lublin University of Technology, within the research project ‘‘The influence of technological and operational factors on the technical properties of building materials and durability of structures.” Also we thank doctor Przemyslaw Czapik for his laboratory assistance with SEM studies of ASR products.
Contents lists available at ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Properties and durability of coarse igneous rock aggregates and concretes Wojciech Piasta a,⇑, Jacek Góra b, Tadeusz Turkiewicz c a
Faculty of Civil Engineering and Architecture, Kielce University of Technology, Tysiaclecia PP 7, 25-314 Kielce, Poland Faculty of Civil Engineering and Architecture, Lublin University of Technology, Nadbystrzycka 40, 20-618 Lublin, Poland c The Institute of Technical Sciences and Aviation, The State School of Higher Education in Chelm, Pocztowa 54, 22-100 Chelm, Poland b
h i g h l i g h t s Crushing strength is helpful at assessing frost durability of igneous rock aggregate. Igneous rock aggregate with weathered minerals is freeze-thaw and alkali sensitive. Strained quartz was the alkali reactive phase in granodiorite; chalcedony in basalt. Alkali nepheline basalt showed best properties and durability.
a r t i c l e
i n f o
Article history: Received 8 July 2016 Received in revised form 5 September 2016 Accepted 8 September 2016
Keywords: Concrete Coarse aggregate Igneous rock Alkali-silica reaction Freeze-thaw resistance Aggregate crushing value
a b s t r a c t Direct tests allowed determining properties of coarse igneous rock aggregates in comparison with gravel along with freeze-thaw resistance and susceptibility to alkali-aggregate reaction. The relationship between the properties and durability of the aggregates and concretes was evaluated taking into account the results from microstructural analyses. Granodiorite, basalt and gravel were sensitive to alkali-silica reaction due to the presence of strained quartz, chalcedony and opal, respectively. The resistance of the aggregate to freezing and thawing was demonstrated to agree with the values of crushing resistance and the lowest contents of pores with diameters unsafe in terms of the freeze-thaw resistance. Ó 2016 Elsevier Ltd. All rights reserved.
1. Introduction Concrete structures exposed to severe environment require the applied concrete to be not only of good mechanical properties but also of high durability. In moderate and subarctic climate zones and wet conditions, the durability of concrete is strongly dependent on the resistance of coarse aggregate to freezing and thawing. In wet environments, the durability of structures is also related to the aggregate sensitivity to detrimental internal physical and chemical processes as for instance the alkaline pore solution present in the cement paste [1]. The structure of igneous rock capillary porosity is different from that of the majority of sedimentary rocks. The system and structure of pore spaces in igneous rocks are more random and ⇑ Corresponding author. E-mail addresses: [email protected] (W. Piasta), [email protected] (J. Góra), [email protected] (T. Turkiewicz). http://dx.doi.org/10.1016/j.conbuildmat.2016.09.022 0950-0618/Ó 2016 Elsevier Ltd. All rights reserved.
attributed to their genesis. Occasionally, minerals with capillary porosity may occur locally due to their partial weathering. A characteristic arrangement of microcracks or a tendency towards their formation due to the microstructure and mineral composition may be important factors for the properties of many igneous rocks, and thus in part, aggregates. In addition to the pores, primary microcracks may contribute to the flow of water, as may the aggregate grains mechanically induced during the aggregate production process. The formation of new microcracks in the grains has been confirmed in the studies which demonstrated an increase in water absorption and reduction in bulk density of aggregates even up to about 10% relative to the rock raw material [2–4]. Uncontrolled microcracks form when the rocks, after being crushed during extraction, are fragmented several times in a production process until an adequate particle fraction is obtained. The authors of this paper suggest that analyses and evaluations predicting the freeze-thaw durability of igneous rock aggregate should also allow for its resistance to cracking under mechanical
120
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
load. This parameter is dependent on the mineral content, mineral microstructure and microcracks in the aggregate, not only on the pore size distribution and porosity, which usually, though not in the case of these rocks, may play a dominant role. Ukrainian igneous rock aggregates have been readily exported and used in some European countries in the past few years. To evaluate their suitability, our detailed studies were performed of the properties and durability of these aggregates, including their application to concrete. The outcome of the studies should indentify the igneous rock aggregates that are suitable for the use at a broader level, also in the structures exposed to severe environments.
2. Literature survey Limiting the considerations about the freeze-thaw resistance of igneous rock aggregate to the porosity-dependent degree to which the water fills the pores or to water absorption level, even when the pore size distribution is taken into account, may be insufficient, as the effect of the rock matrix strength, dependent on its mineral composition, mineral structure and, in part, the microcracks, is neglected. Cyclic freezing and thawing makes the occurring stresses exceed the strength limit of the rock material leading to grain cracking – the final effect of freeze-thaw damage. This fact is used by standard testing methods for the resistance of aggregate to freezing and thawing [5,6]. Martinez-Martinez et al. [7] observed rock microstructures and reported that during the cycles of freezing and thawing, localized damage to the structure is initiated ‘‘as inter-particle-mickrocracking”. Then the propagation of this local mickrocracking is stopped for a number of cycles. During the stable period, isolated microcracks appear from where new ones nucleate and grow. When a critical threshold is exceeded, microcracks turn into cracks and grow rapidly, causing ultimate failure of rock after a few cycles [7]. Huge mineral and structural diversity of rocks hinders proposing an universal scheme and rules of cracking during freeze-thaw deterioration. There are opinions that freeze-thaw durability of rocks is not related to their origin, composition or mineral content, crystal size or water absorption level [8]. The deterioration of concrete in the structure may be caused not only by external actions (loads, temperature and chemical attack) but also by internal physical and chemical processes, in which the share of the environment is negligible. These deleterious processes include alkali-aggregate reaction (AAR) due to the influence of alkali present in the concrete pore solution on certain phases in the aggregate. The reaction occurs when the air relative humidity exceeds 80%. The cement clinker is the main alkali source, but they can also come from the aggregate, for example, from granite [9], or from the external environment. The alkali-silica reaction (ASR) product is a calcium-alkali-silicate-hydrate (C-N(K)-S-H) gel or an alkali-silicate-hydrate (N(K)-S-H) [10] gel, which is precipitated in the aggregate-cement paste interfacial transition zone, paste microcracks or aggregate microcracks. By absorbing water and swelling, the gel induces internal stresses that cause expansion and micro- and macrocracking in the paste, which gradually propagates in the concrete across the entire structural element. From the structural standpoint, AAR and microcracks, which extend into cracks and increase the transport of mass in concrete, can be a serious problem, as they open access for air and water or an aggressive agent to the steel reinforcement thus accelerating rebar corrosion substantially. Concrete mechanical properties worsen with progressing AAR. Ahmed et al. [11] and Giaccio et al. [12] reported a significant drop in the tensile strength of concrete containing reactive aggregate in the early stage of ASR, even though no noticeable reduction in compressive strength of concrete was observed at that time. At that ASR stage, elastic strains increase
in concrete along with gel formation and first microcrack formation, reducing the static modulus of elasticity of concrete [11–13]. AAR was observed in various elements of bridges and marine structures. The reaction was first identified by Stanton [14] in the concrete element of a bridge. Lukschová et al. [15] reported concrete defects in 13 bridges due to the swelling of the ASR gel product. The bridge concrete was produced with coarse basalt, granite and diorite aggregates. The gel formed as a result of the reaction with the grains of quartzite, monomineral quartz or greywacke present in fine aggregates. The basalt, granite and diorite aggregates did not react with alkali [15]. Shayan and Lancucki [16] reported ASR of the apparently non-reactive granite aggregate slowly progressing in the concrete of the bridge. Fifteen years after construction first microcracks appeared with alkali silicate gel formed in the cracks around the granite aggregate particles detected after 36 years. Sibbik and Page [17] also observed a slow deterioration of concrete with granite aggregate. Ponce and Batic [18] think that many granite and granodiorite aggregates can be classified as aggregates reacting very slowly and showing no signs of ASR in standard tests. This slow reaction can be associated with the presence of cryptocrystalline or strained quartz in granite. Basalt as the extrusive rock containing no quartz or minor quantities of quartz is recommended as the aggregate nonreactive to alkalis. However, some basalts can be reactive. Katayama et al. [19] showed that when the SiO2 content in the chemical composition of basalt exceeds 50%, it might be suspected to be potentially reactive to alkalis, nearly as reactive as andesite aggregate. Tiecher et al. [20] found that the amount of silica in volcanic rocks used to produce aggregates was the critical quantity for predicting their reactivity to alkalis. The reactivity is attributed to amorphous volcanic glass present in the interstices between the grains of the matrix. This material is crypto-microcrystalline and made up of quartz, silica, apatite and hematite [20]. ASR may affect various sedimentary but also plutonic and volcanic rocks including granite and basalt, which are apparently nonreactive. But in the case of igneous rocks, ASR proceeds very slowly due to their tightness and the special forms of silica. Therefore the aggregates from this rock material require reliable and long-term tests, as the majority of the aggregates react slowly. 3. Materials and methods 3.1. Aim and scope of the experiment The aim of this study was to investigate and evaluate the properties of coarse crushed aggregates derived from igneous rocks, plutonic and volcanic, used for constructing concrete structures exposed to severe environments. The same investigations were conducted with the postglacial gravel for comparison. The evaluation included the use of direct test methods for aggregate characteristics and for the properties of concretes formed with these aggregates. In addition to mechanical properties of the concrete and aggregate, the properties responsible for the durability of structures in severe climatic conditions were also studied. To explain the differences between the performance of individual aggregate types and concretes, the results of the tests and microstructural analyses are presented in the paper. 3.2. Materials The tests were performed on five types of crushed aggregate derived from igneous rocks and, for comparison, one type of gravel aggregate, along with six concretes made with these aggregates. Three aggregate types came from western Ukraine – basalt from Iwaniczi (Buk), granite from Vyrivs‘kyi karier (GRuk) and
121
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
granodiorite from Klesov (GDuk) – and two from south-western Poland – basalt from Winna Gora near Piotrowice (Bpl) and granite from Siedlimowice near Strzegom (GRpl), plus natural postglacial gravel from Sokolka near Suwalki (Gpl) in north-eastern Poland. Rinsed natural quartz sand from Sokolka was used in the concretes. The gravity of the sand was 2.64 kg/dm3, and the content of mineral dust (less than 0.063 mm in diameter) was 0.43%. The particle size gradation of the fine and coarse aggregates is summarised in Table 1. The concretes with all aggregate types were made with Portland cement CEM I 42,5 R. Chemical composition of the cement clinker, tested in accordance with PN-EN 196-2 [21], is presented in Table 2. Physical properties of the cement are shown in Table 3. In order to ensure reliable comparison of the effects of aggregate type on the properties of concrete, the principle of maintaining the same volume of aggregate in 1 m3 of the mixture was adopted. Thus, the contents of cement, water and sand were unchanged in all mixtures. All concrete mixtures had the same w/c ratio equal to 0.55. Table 4 shows the composition of the mixtures. No chemical additions or mineral additives were used in the mixtures. The indicator of the concrete mixture consistency, the slump was 9 ± 3 cm, which corresponds to S2 S3 classes of slump [22].
3.3. Methods 3.3.1. Test methods for aggregate Each size fraction was tested for crushing strength, water absorption (wt.), bulk gravity and specific gravity as the basis for the determination of the total porosity. The other tests focused on the resistance to freezing and thawing, the potential alkali reactivity and the microstructure of the aggregates. The crushing strength of the aggregate was determined using the aggregate crushing value (ACV) to PN-B-06714-40 [23], in compliance with the British Standard BS 812 [24], but with certain modifications. The aggregate sample placed in a steel cylinder 150 mm nominal internal diameter was loaded with a force of 200 kN. The aggregate crushing value is the ratio of the mass of fines (less than 1/4 of the lower sieve of a given size fraction) formed by the crushing process to the total mass of the sample expressed as a percentage. The ACV can be used for the classification of aggregate and evaluation of its suitability for given concretes. The test is applicable for 4 8 mm, 8 16 mm and 16 31.5 mm size fractions. Two methods were used to study the water absorption of each fraction sample. In the first method the absorption was determined at the atmospheric pressure [25]. Clean and dry aggregates were
Table 1 Sieve analysis of the fine and coarse aggregates, %. Sieve size, mm
Sand
31.5 16 8 4 2 1 0.5 0.25 0.125
Coarse aggregate
100 100 100 100 98 78 46 12 2
Bpl
Buk
GRpl
GRuk
GDuk
Gpl
100 95 51 9 1 – – – –
100 100 48 11 2 – – – –
100 97 51 15 3 – – – –
100 90 49 13 3 – – – –
100 100 50 11 1 – – – –
100 100 51 12 2 – – – –
Table 2 Chemical composition and alkali content in cement clinker, %. Compound
CaO
SiO2
Al2O3
Fe2O3
MgO
Na2O
K2O
Na2Oe
Content
63.97
20.71
4.35
3.46
1.48
0.26
0.50
0.66
Table 3 Physical properties of cement.
**
Le Chatelier, mm
Spec. surface, cm2/g
Spec. gravity, kg/dm3
Initial setting time, min.
Heat of hydr, J/g**
2-day compr. strength, MPa
28-day compr. strength MPa
0.9
3790
3.08
160
306
29.1
54.3
Heat of hydration measured after 41 h with the use of semiadiabatic method.
Table 4 Composition of concrete mixtures. Concrete components
Designation of concrete according to aggregate type used C-Bpl
Cement, kg/m3 Aggregate 2 8 mm, kg/m3 Aggregate 8 16 mm, kg/m3 Sand, kg/m3 Water, dm3/m3
350 644 644 681 193
C-Buk
C-GRpl
C-GRuk
C-GDuk
C-Gpl
626 626
569 569
580 580
576 576
576 576
122
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
Table 5 Test results for coarse aggregates. Properties
Fraction (mm–mm)
Bulk specific gravity, kg/dm3 Specific gravity, kg/dm3 Total porosity, % Crushing value, %
– – – (4–8) (8–16) (4–8) (8–16) (4–8) (8–16) (4–8) (8–16) (4–8) (8–16)
Absorption under atmospheric pressure, % Absorption under vacuum, % Freeze-thaw resistance to PN-B-06714-19, % Freeze-thaw resistance to EN-1367-1, %
Coarse aggregate Bpl
Buk
GRpl
GRuk
GDuk
Gpl
2.96 3.02 1.99 4.3 4.9 0.44 0.32 0,51 0.45 0.2 0.25 0.1 0.15
2.84 2.99 5.02 3.9 8.5 1.45 1.31 1.57 1.48 3.6 4.0 0.6 0.7
2.62 2.67 1.87 14.9 15.3 0.47 0.32 0.60 0.52 1.4 3.7 0.3 0.45
2.64 2.67 1.12 5.9 9.8 0.31 0.18 0.39 0.31 2.0 2.7 0.2 0.4
2.63 2.68 1.87 6.2 5.8 0.34 0.26 0.47 0.38 0.8 1.2 0.1 0.3
2.63 2.71 2.95 7.5 10.6 0.98 0.76 1.09 1.05 3.6 4.5 1.6 2.0
4.6 0.38 0.48 0.23 0.13
weighed and submerged in water up to 1/2 of the height of the smallest particle. After three hours, the water level was raised to 2/3 of the largest particle height and after another three hours the water level was not lower than 2 cm above the largest particle. Forty-eight hours from the start of saturation, the aggregate was surface dried and weighed. The second procedure involved water saturation under vacuum. Dry aggregates were stored under the pressure lowered to 4 kPa for 1 h and then saturated in water for 3 days. Table 5 summarises the absorption results from both test methods. The specific gravity was measured with the use of a pycnometer after the fragmentation of the aggregate to sizes less than 0.08 mm. The air contained between the grains of the powdered material was removed by placing the pycnometer in the vacuum chamber and by lowering the pressure to 2.33 kPa. The apparent specific gravity was determined to EN 1097-6 [26]. The total porosity was calculated based on the values of apparent and specific gravities. The resistance to freezing and thawing of the aggregate was investigated by subjecting each size fraction to 80 freeze-thaw cycles after optimal water saturation, as during the absorption test after the vacuum saturation. The number of freeze-thaw cycles was increased relative to the standard recommendations in order to obtain larger differences in the mass of grains crushed. After freeze-thaw cycles, the damage to the aggregate grains was assessed. For that purpose, the relative quantity of grains crushed was determined through the cycles of freezing and thawing – in accordance with PN-B-06714-19 [5] – crushed to the size less than the lower sieve and – in accordance to EN-1367-1 [6] – crushed to the size less than 1/2 of the lower sieve for the given size fraction. The potential alkali reactivity of the aggregates was determined using the ASTM C–1260 [27] method with six mortar bars 25 25 250 mm of each aggregate type. To explain the relationships and changes observed during the durability tests, the aggregates and mortars were subjected to microstructural analysis. The ordinary petrographic examination was carried out on the samples from the thickest aggregate grains. The standard thin sections were examined with a petrographic microscope in the polarized light and fluorescence modes. The crystalline phases in the aggregates were identified by the XRD analysis. The AutoPore IV 9500 mercury porosimeter was used to determine the pore size distribution. Alkali-aggregate reaction products were identified through the environmental SEM with EDXA examination of mortar microstructures (operating voltage: 20 kV; working distance 9–12 mm).
6.6 1.37 1.52 3.85 0.65
15.1 0.40 0.56 2.55 0.38
8.0 0.24 0.35 2.40 0.30
6.0 0.30 0.43 1.00 0.20
9.3 0.85 1.07 4.10 1.85
3.3.2. Test methods for concrete Twelve cubic specimens 100 mm were made from each concrete mix for compressive strength testing and six beams 100 100 500 mm for bending tests. The bending tests were carried out after 365 days of curing; the compressive tests were carried out after 28 and 365 days. The specimens were saturated in water up to 1/2 of their height after being placed on a grate 10 mm above the bottom of the bathtub. At 24 h the water level was raised to reach 10 mm above the specimen surfaces. After consecutive 24 h, they were removed, wiped dry and weighed. The specimens remained immersed in water until no increase in mass was recorded. After that, they were dried to a constant mass, weighed and the mass absorption was determined. To determine the resistance of concretes made with various aggregate types to freezing and thawing, the change in mass and compressive strength were studied on 10 cm cubic specimens. Six specimens for each concrete were formed and six were sawed from long beams 10 10 cm in cross-section. The cubes cut out from the beams showed the cut aggregate grains (not covered in cement paste) on two walls. Lack of air entraining modification allowed exposing and emphasizing the freeze-thaw action on the aggregate grains. The resistance to freezing and thawing was determined after saturating the specimens in water, as in absorption tests. The specimens were subjected to freezing at 20 °C for 4 h, and then to thawing in water for 4 h and heating to +20 °C. The parameters measured included mass loss and reduction in compressive strength relative to the reference specimens stored in water at +20 °C ± 2 °C for the entire testing period.
4. Test results 4.1. Test results of commonly measured aggregate properties Table 5 compiles the values of bulk specific gravity, specific gravity, total porosity, crushing value and water absorption of coarse aggregates. Two values are given if the tests were conducted on two aggregate fractions, with their weighted arithmetic mean shown next to them. Analysis of the test results of aggregates indicates relatively low bulk specific gravity of basalt Buk, which confirms the highest porosity of all aggregates studied and the highest water absorption. Granite GRpl has the highest crushing value, on average twice as high as that of the other aggregates. The results for the vacuum saturated aggregates show that basalt Buk has the highest water absorption, higher than that of
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
gravel Gpl despite the fact that most grains were derived from sedimentary rocks of capillary porosity. Aggregates GRuk and Gduk show the lowest levels of water absorption. The absorption results for atmospheric pressure and vacuumsaturated aggregates are in agreement, which confirms the same relations between absorption values of individual aggregates. The highest absorption level at atmospheric pressure saturation is also seen in basalt Buk and gravel Gpl. Absorptions of most of the other aggregates were in the range 0.3%–0.5%. With regard to both fractions, better freeze-thaw resistance was recorded for aggregate Bpl, while Buk was crushed most of all igneous rock aggregates during the cycles of freezing and thawing.
4.2. Results of the mercury porosimetry test of aggregates The differential distribution of pore sizes shown in Figs. 1 and 2 indicates that the most remarkable pore size distribution curves represent basalt Buk, granite GRpl and basalt Bpl. Basalt Buk has the largest number of pores with diameters smaller than 0.4 lm. Granite GRpl has a considerable number of pores with diameters in the range 0.1–1 lm, whereas in basalt Bpl, pore diameters over 0.1 mm and below 10 nm are most common. Similar pore size
123
distribution is seen in granite GRuk and granodiorite GDuk, the diameters being mostly from 0.005 to 0.2 lm, with fewer pores coarser than 0.1 mm in diameter. 4.3. Test results for aggregate microstructure 4.3.1. XRD analysis X-ray diffraction revealed the presence of the following crystalline minerals in the aggregates. Basalt Bpl: labradorite (plagioclase), pyroxene – in the form of augite (magnesium iron silicate), nepheline (sodium and potassium aluminosilicate) silica-undersaturated alkali feldspathoid, Basalt Buk: labradorite, augite, nontronite – clay mineral (isomorphic smectite with montmorillonite), Granite GRpl: quartz, microcline – potassium alkali feldspar (similar to orthoclase), albite (plagioclase), biotite, Granite GRuk: quartz, microcline, albite, biotite, Granodiorite GRDuk: quartz, microcline, albite, biotite, muscovite, Gravel Gpl: quartz, microcline, albite, calcite, dolomite, illite – clay mineral. 4.3.2. Polarizing petrographic microscope Additional petrographic analysis with the use of a polarized microscope revealed more details about mineral characteristics of Buk and GDuk aggregates, significant due to the highest expansion of mortars made with them observed in Ukrainian aggregates during the AAR tests. Basalt aggregate Buk is composed of plagioclases and pyroxenes as most important constituents, with a minor number of the minerals as a residue of partial weathering (Fig. 3). The rock does not contain any igneous amorphous glass, which initially filled the spaces between the plagioclases and pyroxenes and was finally replaced by microcrystalline smectite mineral and iron hydroxides (probably goethite). These minerals make up about 10% of the basalt volume. The analysis reveals some loose accumulations of silica in the form of fine spherical chalcedony grains 0.02– 0.03 mm in size, which is an important observation for the AAR test results. The accumulations of fine chalcedony grains are dispersed in the secondary material (among iron hydroxides and smectites) and do not form any larger conglomerates (Fig. 4).
Fig. 1. Total pore volume versus pore diameter.
Fig. 2. Logarithmic differential distribution of pore sizes.
Fig. 3. Basalt Buk. Bladed plagioclases (grey), isometric pyroxenes (white) and accumulations of clay minerals and iron hydroxides (brown). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
124
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
Fig. 4. Basalt Buk. Chalcedony in sparse accumulations are seen in the centre of smectite pseudomorphs. Smectite mineral and iron hydroxides between plagioclase blades (grey) and pyroxenes. Around the pyroxenes, opaque rims of iron hydroxides are seen.
Fig. 6. Biotite blades and clinozoisite-epidote minerals in the medium to finegrained granodiorite.
Granodiorite GDuk is composed of plagioclases, alkali feldspars, quartz, biotite and muscovite. Quartz is the only form of silica present, fine and coarse-grained in part, with irregular, wavy and jagged edges. These irregular edges of the larger particles suggest that in some cases these may be strained quartz grains (Figs. 5 and 6). The rock shows only a minor degree of secondary transformation. The changes are observed in the form of a weak sericitic alteration of plagioclases and presence of fine-grained accumulations of biotite blades and epidote group minerals. In addition to these constituents, the fine-grained variant contains sparse grains of opaque minerals (probably pyrite) and single particles of muscovite and garnet up to 0.5 mm in size (Fig. 6). 4.4. AAR results The ultrafast test [27] of linear strains in mortars revealed that expansion of the mortar with granodiorite aggregates was 0.13% (Fig. 7) after 14 days of soaking in a 1 N solution of NaOH at 80°C, exceeding the expansion limit, 0.1%, for potentially reactive aggregate. At 28 days, the expansion strain increased to 0.21%. Note that in the gravel-based mortar, expansion was 0.09% at
Fig. 5. Irregular wavy edges – strained quartz grains in a medium to fine-grained granodiorite.
Fig. 7. Results of expansion strain for mortars stored in a 1 N NaOH solution at 80 °C.
14 days, being at the boundary of strain level typical of mortars with potentially reactive aggregate. At 28 days the value rose to 0.15%. The mortars with Ukrainian and Polish basalts showed different behaviours in the NaOH solution. The mortar with basalt aggregate derived from Poland’s quarries was not subject to any expansion, which is characteristic of alkali basalts. The mortar with Ukrainian basalt aggregates showed the expansion of 0.07% at 28 days, which may seem interesting in terms of the causes of these varying behaviours of two aggregates derived from the same type of rock. After the test for Potential Alkali Silica Reactivity of Aggregates in the NaOH solution conducted, SEM and EDXA were used to study the microstructure of mortars. Fig. 8a shows a photograph of a relatively smooth surface of a gel which formed in the mortar with granodiorite aggregate. Microcracks are running across this relatively smooth surface. The chemical composition at point A, Fig. 8b, suggests that most likely it is calcium sodium silicate hydrate. Fig. 9a shows an image of the gel identified on the fracture surface of the specimen made with gravel aggregate and soaked in a
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
125
Fig. 8. Microstructure of mortar with granodiorite aggregate in scanning electron microscope: a) a layer of calcium sodium silicate b) chemical composition at point A by EDXA.
Fig. 9. Results of microstructural analysis of the mortar with gravel aggregate under a scanning electron microscope: a) calcium sodium silicate gel with cracking b) chemical composition at point B by EDXA.
1 N NaOH solution at 80 °C for 28 days. There is a ‘‘spider web” of microcracks on its surface. Some of these microcracks have large widths. The chemical composition determined on the surface at point B in Fig. 9b suggests that it is a calcium sodium silicate gel hydrate. 4.5. Test results of the strength of concretes Table 6 compiles compressive strengths f28d cm,cube at 28 days and f365d cm,cube at 365 days of curing of cubic specimens with a side length of 100 mm, cylindrical specimens f28d cm,cyl at 28 days and, determined
based on these results, compressive strength classes of the concretes. The table also shows flexural strengths f365d ct,flex of concretes at 365 days of curing. The results from the test for compressive strength f28d cm.cyl of concretes showed differences of 2 strength classes. The differences in flexural strength results f365d ct.flex of individual concretes were similar considering that according to PN-EN 1992-1-1:2008 [28] the difference between two consecutive classes is 0.3 MPa. In both cases best results were obtained for the concrete made with basalt Bpl. The lowest value of compressive strength and the lowest class were recorded for the concrete made with granodiorite GDuk. As
126
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
for the flexural strength f365d ct.flex, the lowest values were recorded for the concrete with granite GRpl and the Ukrainian granodiorite, at least 2 classes lower, when referred to the concrete strength classes, than the concrete with the Polish basalt aggregate. 4.6. Test results for freeze-thaw resistance and water absorption of concretes After 150 cycles of freezing and thawing, the decrease in mass of the moulded specimens of concrete C-GDuk with granodiorite (Fig. 10) was 7% (Table 7). The specimens were cracked locally. The mean strength was only 55%. Two of the remaining concretes were different from others: C-Gpl and C-Buk. The mass of the gravel concrete increased by about 3%, with the strength drops of nearly 40%. The mass of the concrete made with the Ukrainian basalt hardly changed (Fig. 10), with a minor only drop in strength (Table 7). Other concretes had similar low freeze-thaw resistance after 150 freezing and thawing cycles.
A mass gain of 2%–3% in the sawed specimens made from concretes with aggregates GDuk and Gpl for a few dozen cycles of freezing and thawing (Fig. 11) and then rapid mass loss, strength decline, cracking and gradual failure of the specimens confirm a typical pattern of frost related damage of low resistance concrete. The change in mass of the sawed specimens made of concretes C-GRuk, C-Buk and C-Bpl suggests better resistance to freezing and thawing. However, after 150 freezethaw cycles all concretes showed a considerable drop in strength (Table 7), and the majority of specimens were seriously cracked (Fig. 12). Higher strengths of concretes containing basalt aggregates and better condition of the specimens indicate their better resistance. The results (Table 8) show that the water absorption levels of the concretes with igneous rock aggregates were similar and ranged from 5.2 to 5.4%, except for aggregate Buk, for which the value was nearly 15% higher. Higher absorption was also recorded for the gravel concrete, but it was only 5–6% higher.
Table 6 Mean results of compressive strength and flexural strength and strength class of the concrete. Properties
Compr. strength f28d cm,cube, MPa Compr. strength f365d cm,cube, MPa Flexural strength f365d ct,flex, MPa Compr. strength f28d cm,cyl, MPa Strength class to EN 206-1
Concrete C-Bpl
C-Buk
C-GRpl
C-GRuk
C-GDuk
C-Gpl
47.1 51.5 5.20 44.2 C40/50
46.0 50.4 4.96 40.8 C35/45
45.1 51.1 4.95 39.7 C35/45
43.4 47.3 4.38 40.1 C35/45
43.4 48.5 4.60 37.9 C30/37
42.0 50.1 5.02 38.0 C30/37
Fig. 10. Average mass change of moulded specimens subjected to freezing and thawing.
Table 7 Mean compressive strengths after freeze-thaw cycles. Properties
* **
Concrete C-Bpl
C-Buk
C-GRpl
C-GRuk
C-GDuk
C-Gpl
Moulded specimens (m) Compr. strength after 150 cycles, % Reference strength, MPa
80 51.9
93 52.9
73 51.5
77 50.9
55 50.9
62 49.5
Sawed specimens (s) Compr. strength after 150 cycles, % Reference strength, MPa
67 49.6
75 45.9
60 50.6
70 48.3
13* 46.2
55** 48.1
After 120 cycles. After 80 cycles.
127
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
Fig. 11. Average mass change of sawed specimens subjected to freezing and thawing.
Fig. 12. Images of the sawed cubes of concretes after 150 cycles of freezing and thawing: a) C-GRuk b) C-GRpl c) C-GDuk d) C-Gpl.
Table 8 Water absorption by weight of concretes. Properties
Water absorption by weight, %
Concrete C-Bpl
C-Buk
C-GRpl
C-GRuk
C-GDuk
C-Gpl
5.2
6.1
5.2
5.4
5.3
5.6
5. Analysis and discussion 5.1. Freeze thaw resistance of aggregates and concretes Fig. 13 shows the results for three characteristics of aggregate fraction 8–16 mm: resistance to freezing and thawing to [5],
vacuum saturation absorption and the crushing value (ACV) to [23]. The freeze-thaw resistance of aggregates (Bpl and GDuk) was found to be highest when absorption and crushing value were low. It has to be noted that the values of ACV and freeze-thaw resistance presented the best agreement. Low freeze-thaw resistance (>3% crushed grains) was observed in Buk (4.0% crushed
128
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
Fig. 13. Dependence of the resistance to freezing and thawing on absorption levels and crushing value.
grains) and GRpl (3.7% crushed grains), but then the ACV of granite GRpl and absorption of Buk were the worst. The results show that in the case of igneous rocks, crushing strength is a more practical indicator of resistance to freezing and thawing than absorption is. The correlation between freeze-thaw resistance and absorption is also questioned by Rusin et al. [8]. In the case of the lowest freeze-thaw resistance of aggregates, there is a good correlation between their low absorption and high crushing strength. The best (Bpl) and worst (Buk) physical properties and freezethaw resistance of basalt aggregates are related to their mineral composition and microstructure. Basalt Bpl is an alkali rock with no trace of weathering and strong rock matrix (ACV = 4.6%). The differential distribution of pore sizes in basalt Bpl (Fig. 2) confirms the presence of two dominant pore groups – coarse pores with diameters larger than 0.1 mm, regarded as safe [29] and fine pores with diameters ranging from 5 nm to 10 nm. The high porosity of basalt Buk is associated with the presence of the minerals (Fig. 3) typical for sedimentary rocks, which indicates partial weathering. In addition to chalcedony, this basalt contains smectite (nontronite) with a laminar structure and ability to absorb water, which increases absorption levels. As a result of transformations, basalt Buk contains a large group of pores that are regarded unsafe in terms of freeze-thaw resistance [29], with diameters between 20 nm to 400 nm (Fig. 2). This group of pores is a major factor in high absorption and poorest freeze-thaw resistance of the basalt, despite its relatively strong matrix. In granite GRpl, the presence of dominant group of pores about 200 nm in diameter explains its low freeze-thaw resistance. The lowest freeze-thaw resistance of gravel Gpl is related to the origin of grains present in its composition. Most grains of the gravel resulted from the erosion of sedimentary rocks (e.g., limestones, dolomite limestones, sandstones with silica and carbonate cements containing small amounts of clay minerals) with capillary porosity, which usually causes higher absorption of the material. Analysing the freeze-thaw resistance of aggregates and looking for the influencing factors and their significance, it is important to remember that the deterioration (crushing) of aggregate grains is caused by the stress from increasing pressure due to water freezing in their pores and microcracks. The measurement of the material crushed by the cycles of freezing is required by standards for the evaluation of the freeze-thaw resistance of aggregates [5,6]. This requirement, met by the authors of this study, makes the freeze-thaw resistance evaluation dependent on the condition and mechanical properties of aggregates, while the strength characteristics can be estimated by assessing its crushing strength [23,24]. Analysis of the freeze-thaw resistance of aggregate only in the light of its absorption levels and porosity, even accounting for pore
geometries, can be difficult and insufficient. This approach ignores the effect of the rock matrix strength, which is dependent on its mineral composition and structures of the minerals. Therefore, for practical reasons it seems justified to facilitate the diagnosis by taking into account the strength of the grains represented by the ACV. Good behaviour of the concrete with aggregate Buk during the cycles of freezing can be surprising considering a relatively low freeze-thaw resistance of this aggregate. Two issues require attention. – (1) Basalt Buk contains a lot of pores, including coarse ones (Fig. 1) not fully filled with water in the freezing concrete, especially that dry aggregates were used in the concrete mix. These partially filled pores can take part of the water and reduce pressure in the aggregate grains and in the paste surrounding them, therefore protecting them against damage as air entrainment does. – (2) The dry absorbent aggregate in the surrounding paste reduces its w/c ratio, which has a critical influence on the frost protection of concrete. The average freeze-thaw resistance of concretes made with the other aggregate types was similar and resulted from the high w/c ratio and lack of air entrainment. The freeze-thaw resistance of all concretes in the moulded specimens was considerably higher than that in the sawed specimens which had aggregate grains cut and exposed on two sides. 5.2. Alkali-aggregate reactions The SEM images (Fig. 8a) and EDXA results (Fig. 8b) suggest that the large expansion of the mortar bars made with granodiorite aggregate was caused by the formation and swelling of the calcium sodium silicate gel. The XRD analysis and petrographic test results (Fig. 6) show that the aggregate contains mainly quartz, alkali feldspar, plagioclases and a little of biotite and muscovite. This gel is probably a product of the reaction between sodium and calcium hydroxides and quartz present in the granodiorite in the form of strained quartz (Fig. 5). Just as in the case of some granites [17,18], ASR of these granodiorite aggregate can be assumed to be associated with the presence of strained quartz grains or cryptocrystalline quartz which are thermodynamically unstable and have deformation-related high free energy in their crystalline networks. Although the expansion of the mortar with aggregate Buk (Fig. 7) in the NaOH solution is less than 0.1%, it differs from that of the mortar with basalt aggregate Bpl, which does not deform in any way for 28 days. This results from the fact that basalt aggregate Bpl is derived from an alkali basalt rock with a minor content of volcanic glass, shows no trace of weathering and contains silicaundersaturated minerals such as nepheline (alkali feldspathoid) identified with XRD analysis. The expansion of the mortar with basalt Buk was probably related to the presence of small aggregations of chalcedony (Fig. 4) built from cryptocrystalline silica. As commonly known, chalcedony reacts easily with sodium and calcium hydroxides. Alkali silica reaction of this basalt aggregate might be also partially related to the reactions between calcium and alkalis with nontronite (smectite isomorphic with montmorillonite) (Fig. 4). Because the basalt matrix is very strong and amounts of chalcedony and nontronite are small, the kinetics of the reactions and expansion are limited considerably. Reactions of the paste with smectites were also observed by Batic et al. [30], who focused on ASR of basalt containing montmorillonite. Accelerated expansion of the paste with gravel aggregate in the NaOH solution resulted from the formation and swelling of the calcium sodium silicate gel, as observed under scanning electron microscope on the specimen fracture surface (Fig. 9a and b). The formation of the gel was related to the presence in the gravel [31] of few grains of quartz-pyroxene slate with opal cement.
W. Piasta et al. / Construction and Building Materials 126 (2016) 119–129
129
6. Conclusions
References
1. In addition to pore size distribution, the crushing strength together with water absorption can be a good and practical indicator of the resistance of crushed aggregate derived from igneous rocks to freezing and thawing. Absorption alone cannot be an independent indicator of freeze-thaw resistance of aggregate. For example, the highest freeze-thaw resistance of Polish basalt aggregate as well as Ukrainian granodiorite and granite aggregates corresponded to their high crushing strength, low water absorption and the lowest content of pores with diameters unsafe in terms of the aggregate freeze-thaw resistance. 2. The occurrence of any quantity of secondary minerals formed as a result of weathering in an igneous rock can increase porosity, leading to the weakening of strong rock matrix thus reducing its freeze-thaw resistance. In the case under investigation, low freeze-thaw resistance of the Ukrainian basalt matched its porosity, high absorption and moderately high crushing strength (ACV = 8.5% for 8–16 mm fraction). The results of microstructural analysis of this aggregate indicate that it was related to the presence of secondary minerals, such as smectites, chalcedony and other minerals formed during the weathering process. 3. The test results of mortar expansion in a NaOH solution indicated that granodiorite aggregate and gravel might be susceptible to ASR (potentially reactive). Sodium and calcium silicate gel was found to form in the mortars. In granodiorite, strained quartz is probably the reactive phase, and in the gravel – few grains containing opal cement or those with clay-limestone cements. The use of the Ukrainian basalt aggregate in concrete exposed to harsh environmental conditions is risky because of the detected presence of chalcedony and smectite group of minerals (nontronite). 4. Three out of six aggregate types studied – plutonic granodiorite, volcanic basalt and sedimentary sandstone (grains in the gravel) – were found to be potentially sensitive to ASR due to the presence of strained quartz, chalcedony and opal, respectively. It is thus absolutely necessary to subject each and every aggregate – also that derived from igneous rocks, with very good physical and mechanical properties – to tests for potential alkali reactivity. 5. Only two out of six aggregate types and concretes studied, Polish basalt and Ukrainian granite, were found to be versatile and suitable for a wide range of use, including the structures exposed to severe environment.
[1] T. Katayama, D. Bragg, Alkali-aggregate reaction combined with freeze/thaw in Newfoundland. Canada Petrography using EMPA, Proc. 10th ICAAR, Melbourne Australia, 1996, pp. 243–250. [2] T. Gawenda, Z. Naziemiec, T. Tumidajski, D. Saramak, Ways of optimization of grain composition and shape of grains in aggregates in products of jaw _ crushers, Górnictwo i Geoinzynieria 3/1 (2006) 109–124 (30). [3] J. Bromowicz, Changes of rock material properties during production process of crushed aggregates, Surowce i Maszyny Budowlane 4 (2007) 14–17. [4] Z. Naziemiec, Influence of Treatment Technology on Quality of Obtained Aggregate, 3, Scientific Works of Institute of Glass, Ceramics Refractory and Construction Materials, Warszawa-Opole, 2009. pp. 182–193. [5] PN-B-06714-19:1978, Mineral aggregates. Testings. Determination of resistance to frost by the direct method. [6] EN-1367-1:2007, Tests for thermal and weathering properties of aggregates. Determination of resistance to freezing and thawing. [7] J. Martínez-Martínez, D. Benavente, M. Gomez-Heras, L. Marco-Castaño, M.A. García-del-Cura, Non-linear decay of building stones during freeze–thaw weathering processes, Constr. Build. Mater. 38 (2013) 443–454. [8] Z. Rusin, P. S´wiercz, Z. Owsiak, Effect of microstructure on frost durability of rock in the context of diagnostic needs. 7th Scientific-Technical Conference Material Problems in Civil Engineering (MATBUD’2015), Proc. Eng. 108 (2015) 177–184. [9] Z. Owsiak, Contribution of alkali from aggregate to pore solution of concrete, Cem. Wapno Beton 4 (2001) 151–153. [10] L. Bainga, T. Nawa, E. Iwatsuki, T. Awamura, ASR formation of reactive chert in conducting model experiments at highly alkaline and temperature conditions, Constr. Build. Mater. 95 (2015) 820–831. [11] T. Ahmed, E. Burley, S. Rigden, A. Abu-Tair, The effect of alkali reaction on the mechanical properties of concrete, Constr. Build. Mater. 17 (2002) 123–144. [12] G. Giaccio, R. Zerbino, J.M. Ponce, O.R. Batic, Mechanical behavior of concretes damaged by alkali-silica reaction, Cem. Concr. Res. 38 (2008) 993–1004. [13] I. Yurtdas, D. Chen, D.W. Hu, J.F. Shao, Influence of alkali silica reaction (ASR) on mechanical properties of mortar, Constr. Build. Mater. 47 (2013) 165–174. [14] T.E. Stanton, Expansion of concrete through reaction between cement andaggregates, Proc. Am. Soc. Civil Eng. 66 (1940) 1781–1811. [15] Š. Lukschová, R. Prˇikryl, Z. Pertold, Petrographic identification of alkali–silica reactive aggregates in concrete from 20th century bridges, Constr. Build. Mater. 23 (2009) 734–741. [16] A. Shayan, C.J. Lancucki, Alkali-aggregate reaction in the causeway bridge. Perth, Australia, Proc. 7th Int. Conf., AAR in Concrete, Ottawa, 1986. pp. 392– 397. [17] R.G. Sibbick, C.L. Page, Susceptibility of various UK aggregates to alkali– aggregate reaction. Proc. 9th Int. Conf., AAR in Concrete, London, 1992. pp. 980–987. [18] J.M. Ponce, O.R. Batic, Different manifestations of the alkali-silica reaction in concrete according to the reaction kinetics of the reactive aggregate, Cem. Concr. Res. 36 (2006) 1148–1156. [19] T. Katayama et al., The petrographic comparison of rocks from Japan and New Zealand-Potential reactivity related to interstitial glass and silica minerals. 8th Int. Conf., AAR in Concrete, London, 1989. pp. 537–541. [20] F. Tiecher, D.C.C. Dal Molin, M.E.B. Gomes, N.P. Hasparyk, P.J.M. Monteiro, Influence of mesostasis in volcanic rocks on the alkali-aggregate reaction, Cement Concr. Compos. 34 (2012) 1130–1140. [21] PN-EN 196-2:2013–11, Method of testing cement – Part 2: Chemical analysis of cement. [22] EN 206:2013, Concrete. Specification, performance, production and conformity. [23] PN-B-06714-40:1978, Mineral aggregates. Testings. Determination of crushing strength. [24] BS 812-110:1990, Testing aggregates. Part 110: Methods for determination of aggregate crushing value (ACV). [25] PN-B-06714-18, Mineral aggregates. Testings. Determination of absorption. [26] EN 1097-6:2013, Tests for mechanical and physical properties of aggregates. Determination of particle density and water absorption. [27] ASTM C1260 Test Method for Potential Alkali Reactivity of Aggregates (Mortar-Bar Method). [28] PN-EN 1992-1-1:2008 Eurocode 2: Design of concrete structures – Part 1–1: General rules and rules for buildings. [29] Z. Rusin, Technology of Frost-Resistant Concretes, Polski Cement, Kraków, 2002. [30] O. Batic, P. Maiza, J. Sota, Alkali silica reaction in basaltic rocks NBRI Method, Cem. Concr. Res. 24 (1994) 1317–1326. [31] Z. Owsiak, J. Zapała, P. Czapik, Sources of the gravel aggregate reaction with alkalis in concrete, Cem. Wapno Beton 3 (2012) 149–154.
Acknowledgements This study was supported by the Kielce University of Technology and Lublin University of Technology, within the research project ‘‘The influence of technological and operational factors on the technical properties of building materials and durability of structures.” Also we thank doctor Przemyslaw Czapik for his laboratory assistance with SEM studies of ASR products.