CC BY
89Magazine of Civil Engineering. 2022. 113(5). Article No. 11308
Magazine of Civil Engineering issn
2712-8172
journal homepage: http://engstroy.spbstu.ru/
Research article UDC 624.69-033 DOI: 10.34910/MCE.113.8
Metamorphic rocks for manufactured sand and coarse aggregate for concrete and mortar
B. Arulmolya B , K.M.C. Konthesinghaa, S.M.A. Nanayakkarab , H.M.R. Premasiric
a Department of Civil Engineering, University of Sri Jayewardenepura, Ratmalana, Western Province, Sri Lanka
b Department of Civil Engineering, University of Moratuwa, Moratuwa, Western Province, Sri Lanka
c Department of Earth Resources Engineering, University of Moratuwa, Moratuwa, Western Province, Sri Lanka
M branavanarulmoly@sjp.ac.lk
Keywords: aggregates, concrete, mortar, construction industry, porosity, water absorption, compressive strength, tensile strength, mechanical properties, durability
Abstract. The present study determined the most applicable high-grade metamorphic rocks as the sources for manufactured sand and coarse aggregates production to apply in the construction industry. As the first stage of this study, representative samples of the most common ten types of metamorphic rock were collected and tested for mineralogical and chemical compositions to select the favorable rock types in concrete and mortar production. XRD diffraction patterns and quantitative analysis results of XRF were referred for identifying the mineral and chemical contents respectively. Secondly, among the favorable rock types, based on the geology, the area contribution by each rock was calculated and few abundantly available rocks were considered for further investigation. Finally, the compatibility of rocks was checked by analyzing the physical, mechanical, and durability characteristics. As the physical parameters, the specific gravity, water absorption, loose and packing density, and water content were checked. Mechanical properties of rocks were determined by investigating compressive and tensile strength, impact resistance, abrasion resistance, and crushing value. The durability of rocks against weather fluctuations was probed through the slake durability indexes. Results from the first stage revealed that Charnockite, Hornblende-Gneiss, Intrusive Charnockite, and Granitic-Gneiss were favorable rocks rich with concrete and mortar friendly minerals such as albite, k-feldspar, and quartz. Among the above four rocks, Charnockite contributed around 40% and Hornblende-Gneiss covered approximately 9.5% of the area, which were then considered for the next stage of the investigation. Charnockite and Hornblende-Gneiss showed similar porosities and water absorption in the range of 0.25-0.26%. Each mechanical property of both rocks complied with the requirements provided by the standards; moreover, they manifested excellent durability performance against cyclic wetting and drying as well.
Acknowledgment: The authors are grateful to the INSEE Cement, Sri Lanka for providing supports to conduct the Chemical and Mineralogical analysis of rocks.
Citation: Arulmoly, B., Konthesingha, K.M.C., Nanayakkara, S.M.A., Premasiri, H.M.R. Metamorphic rocks for manufactured sand and coarse aggregate for concrete and mortar. Magazine of Civil Engineering. 2022. 113(5). Article No. 11308. DOI: 10.34910/MCE.113.8
1. Introduction
The implementation of a proper alternative for river sand and natural coarse aggregate is the major problem that has risen now among the construction industries in most of the developing countries including
© Arulmoly, B., Konthesingha, K.M.C., Nanayakkara, S.M.A., Premasiri, H.M.R., 2022. Published by Peter the Great St. Petersburg Polytechnic University
Sri Lanka. River sand is currently being used as the main fine aggregate in most construction activities [1]. Many restrictions have been imposed on river sand mining due to the adverse impact on the environment and ecological systems. Excessive excavations near the riverbanks caused serious environmental threats not only in Sri Lanka but also in many countries that are still extracting river sand for infrastructure developments [2]. The most identified impacts on the environment are riverbank collapses, loss of vegetation on the bank of rivers, saline water intrusion, river buffer zone encroachment, erosion of riverbank land areas, and deterioration of river water quality and groundwater quality [2-5]. Presently, the construction companies are trying to introduce manufactured sand as an alternative to river sand for the construction activities to immediately mitigate the issues with river sand mining [1, 6-8]. This leads to an increased number of quarries that are operated to supply the coarse aggregates as well as manufactured sand for both concrete and mortar works [9]. The source to produce manufactured sand is 'high-grade parent rock', which is a solid material with the collection of one or more mineralogical and chemical compositions [10-12]. Therefore, the quality of aggregates produced directly depends on the compositions of parent rocks, which may give positive and negative impacts on the performance of concrete and mortar [7, 8, 13].
Deliberation to manufactured sand production is crucial, as the implications of finely crushed minerals in concrete and mortar are higher than coarse aggregates. It was identified that still, the question remains among the local construction industries, mining industries, and quarries on selecting appropriate rock types for manufactured sand production due to the inclusions of different minerals and chemicals present in the rocks as well as the performance of the rocks [14]. This problem is serious because the potentially deleterious mineralogical and chemical compositions in parent rocks may affect the long and short-term durability of concrete and mortar. Also, when the rock-derived materials used in concrete and mortar are not capable to resist higher stresses or withstand the climatic changes, may fail the structural elements. Therefore, it is a must to investigate the chemical, physical, mechanical, and durability properties of highgrade metamorphic rocks before utilizing them as the sources for aggregate production in the construction work.
This study is focused on determining suitable metamorphic rock types in Sri Lanka and their characteristics for use as the raw materials for producing manufactured sand. Therefore, the outcomes of this research are expected as the guide for selecting appropriate high-grade parent rocks for producing manufactured sand to enhance a durable and sustainable built environment.
Earth's crust is mainly composed of three major rock classes such as igneous rocks, sedimentary rocks, and metamorphic rocks [10, 11]. Based on the simplified geology of Sri Lanka represented by Fig. 1, the crust part (nine-tenths) is mainly underlain by 'Precambrian' age rocks which are subdivided into three major groups based on tectonic setting and the age of rocks: Wanni Complex (WC), Vijayan Complex (VC) and Highland Complex (HC) [12, 15, 16]. The age of each above rock complex is also given in the figure as 'Ma', which defines the mega annum or millions of years. Sedimentary limestone with Miocene age (22.5 to 5 million years ago) can be identified in the North-Western area (coastal belt from Madurankuli to Killinochi). Northern province areas contain limestone facies and the North-Eastern area (coastal belt from Paranthan to Mullaithivu) is underlain by sandstone [17]. Small clusters of mudstone and sandstone of Jurassic age (201 million years) rocks were identified in Tabbowa, Andigama, and Pallama areas having hard sandstone and clay interbedded with limestone and mudstone [18]. Precambrian rocks such as quartzite, charnockite, granulite grade pelitic gneisses, limestone, biotite gneisses are found in HC. Similarly, granitic-gneiss, biotite gneiss, and calc-gneiss rocks are identified in VC while WC is with prominent rock types [15-17]. It is identified that mainly high-grade metamorphic rocks are distributed within this island, which is very strong and hard shows the possibility for manufacturing good quality concrete and mortar aggregates. Some parts of this country are covered with weathered rocks which are not deemed for aggregate production [19].
Some studies proved the linkage between Sri Lankan geology with the global geology, which leads to the depth of this research work to an extent. Prame [20] has explained the lithological similarity of HC with the Trivandrum block of the South Indian granulite belt. Kroner [21] and Pinna et al. [22] proposed that HC geology highly coincides Madagascar-Tanzania and Mozambique while VC rocks have similarities with East and West Gondwana. Sandiford (1986, cited in [20]) compared the lithological affinity with the Rayner Complex of Enderby Land, Antarctica. Yoshida et al. (1992, cited in [20]) correlated Ongul/Skallen groups, Yamato-Belgica Complex, and Prince Olav Complex of Antarctica with HC, VC, and WC respectively.
Limited studies have been carried out so far on the effect of mineral and chemical compositions on the quality of fine aggregates for concrete and mortar production and some of them are included in this section. Dahanayake and Jayasena [23] carried out a study on some of the mineral compounds, texture, and grain direction of gneiss, amphibolite, and some intrusive rocks in VC. A microscopical mineralogical backscattered electron analysis was done by Malaviarachchi and Takasu [24] on some pelitic and intermediate to mafic granulites from HC to check the metamorphic grade based on the mineral
assemblages. A key investigation was carried out by Jayawardena & Dissanayake [25] on the selection of suitable rock types for quarry dust production where the sample collections were done only in the Central and Northwestern provinces of Sri Lanka.
Petrographic studies were followed for the collected samples and some important conclusions were derived: quartz, feldspar, hornblende, hypersthene, and a few biotite mica contents were identified in charnockite, hornblende-gneiss, and granitic-gneiss rocks from the collection areas which were mainly used for quarry dust production in the country. A recommendation of using quartzite for manufacturing quarry dust was also suggested by the authors as it is usually free from mica content. Mica and clay are considered the most harmful minerals in the rocks. Mica drastically lowers the workability of mixes due to flaky particles and clay significantly reduces the strength of concrete and mortar due to its fine form. Tugrul et al. [26] investigated the effects of feldspar, mica, and clay minerals on compressive strength of mortar where replacement of river sand with 20 % of mica and 4 % of clay minerals such as illite and montmorillonite reduced 50-60 % compressive strength at 28 days. A detailed study was accomplished by Leemann and Holzer [27] on various grain size inclusions of mica minerals such as muscovite, biotite, and chlorite in the mortar and muscovite in concrete. Diminished workability, compressive and flexural strengths, and increased void content were observed with higher larger grain-sized mica inclusions in mortar and concrete. From the investigation executed by Norvell et al. [28], greater water demand and drying shrinkage and dwindled compressive strength of concrete was noticed when the aggregates had higher contents of montmorillonite than illite and kaolinite.
Coarse aggregate is also a manufactured primary aggregate that is obtained from several industrial processes. The quality of rock-derived coarse aggregates mainly depends on the properties of the original rock which are used for manufacturing [29]. Therefore, coarse aggregates should have high strength, better particle shape and gradation, good wear resistance and resistance to climatic changes, etc. [30]. van Wyk and Croucamp [13] checked the suitability of quartzite, tillite, sandstone, and shale rocks for concrete based
Figure 1. Simplified geological map of Sri Lanka [17].
on strength and durability parameters. Authors found that required concrete compressive strength was achieved with each rock aggregates but tillite and shale showed unsatisfactory performance with durability. Quartzite revealed the best durability among the selected rocks from slake durability test. Loorents and Johansson [31] studied the quality of granitoid rock aggregate based on free mica grains. Authors identified that the finer particles were abundant with free mica at each fraction while medium and coarse rocks showed high mica content at the fraction of 125-250 ^m. Ugolini et al. [32] analyzed mineralogical, physical, and chemical properties of sandstone and siltstone fragments in the soil. The authors noticed that there was a distinct mineralogical difference between sandstone and siltstone while the physical and chemical properties were marginally similar. Scott and Rollinson [33] investigated the mineralogy and surface textures of granite, granodiorite, lava, pyroclastic, gabbro, dolerite, limestone, dolostone, gritstone, greywackes, anorthosite rocks. Johansson et al. [34] researched the quality of mica-schist rock by conducting the micro-Deval test, Los-Angeles abrasion test, and Nordic test on drilled samples. Because the results were within the required limits, the authors suggested the selected rock to produce good quality aggregate.
To achieve the aim of this study, a number of tasks were solved:
a) investigation on identifying the high-grade metamorphic rocks available based on the geology and linkage with global geology based on the literature review;
b) research of the mineralogical and chemical compositions of high-grade metamorphic rocks and identifying the most suitable rock types for concrete and mortar;
c) study of the availability of selected rock types based on the geology;
d) analyzation of the most determining physical, mechanical, and durability properties of the suitable and abundantly available rocks and compliance with standards.
2. Methods
2.1. Materials
Concerning the geology of Sri Lanka [17], for the preliminary investigation on chemical and mineralogical composition identifications, ten rock types were selected considering the metamorphic class. Fig. 1 shows the locations of sample collection for the first stage of this research. Here the weathered rocks such as alluvial, fluvial, and residual deposits which are covering the Northern and North-Western parts of the country were not selected as those are not usually suitable for aggregate. 'ExpertGPS 7.03 Map' software was used to locate the quarries with the global North (N)-East (E) coordinate system as mentioned in Table 1.
Table 1. Locations of sample collection for Stage 1.
Coordinates
code Rock type Complex Quarry Province North (N) East (E)
1 Charnockite HC Unawatuna Southern 6.01913° 80.25134°
2 Calc-Gneiss HC Walapane Central 7.09548° 80.85943°
3 Hornblende-Gneiss HC Pahala-Bomiriya Western 6.93210° 80.00483°
4 Cordierite-Garnet-Gneiss HC Batawala Western 6.88119° 80.05676°
5 Quartzite HC Mabima Western 6.96388° 79.97711°
6 Schist HC Magala Western 6.29783° 80.12590°
7 Marble HC Kosgoda Western 6.33665° 80.03380°
8 Intrusive Charnockite VC Unnichai Eastern 7.63300° 81.53725°
9 Granitic-Gneiss VC Damana Eastern 7.20869° 81.61647°
10 Biotite-Gneiss VC Komari Eastern 6.99742° 81.86762°
Approximately 12" * 12" rubble sample was collected from each quarry as a representative of the corresponding rock type. Sample collection was done in two complexes as represented by Fig. 1: Highland Complex (HC) (covering the Central, Sabaragamuwa, and Western provinces and some parts of Eastern province) and Vijayan Complex (VC) (covering Eastern, Uva, and Southern provinces). Fig. 2 illustrates the digital images of the collected ten rock rubble samples for the identification of the mineralogical and chemical composition.
To investigate the physical parameters of the rock types which were filtered from the first and second stages of this research, various sizes of rock samples were collected from the corresponding quarries as mentioned here. For the determination of relative density (specific gravity) of rocks, a nominal maximum size of 19 mm rock samples was used. A nominal maximum size of 12.5 mm rock samples was utilized for identifying the loose and packing densities of rocks. The absorbed water and surface moisture of rocks were tested using the maximum size of 9.5 mm samples. Uniaxial compressive strength and tensile strength were determined based on the representative rock samples within the range of 55 mm to 70 mm. The aggregate impact value test was conducted using the rock specimens passing 12.5 mm sieve and retained on 9.5 mm sieve according to gradation criteria provided in ASTM C136. The same quantities of rock specimens retained on 9.5 mm and 6.3 mm sieves were collected for investigating the abrasion. For testing the crushing strength, rock specimens passing 12.5 mm sieve and retained on 9.5 mm sieve were utilized. The response of rocks to wetting and drying conditions was checked using representative rock specimens within the range of 50 mm to 65 mm.
Figure 2. Digital images of rocks listed in Table 1 (a - Charnockite; b - Calc-Gneiss; c - Hornblende-Gneiss; d - Cordierite-Garnet-Gneiss; e - Quartzite; f - Schist; g - Marble; h - Intrusive-Charnockite;
I - Granitic-Gneiss; j - Biotite-Gneiss).
2.2. Experiments and test methods
This study was carried out to suggest the most applicable metamorphic rock types for producing construction aggregates under three stages: Stage 1 (filtering the selected ten rock types based on the constitution of potentially harmful minerals), Stage 2 (determining the most suitable rock types among the filtered rocks from Stage 1, according to the availability) and finally Stage 3 (checking the physical, mechanical and durability characteristics of the selected rocks from Stage 2). Fig. 3 illustrates a flowchart for the methodology of this study and Table 2 lists down the experiments, standards referred to, the number of trials used for the experiments, and equations used for calculating the parameters in this study.
Figure 3. Methodology flowchart. Table 2. Experiments, standards referred and equations used in this study.
Property (Stage No) Experiment Standard referred Trials Equations for determining numerical values
Chemical (1) X-Ray diffraction (XRD) X-Ray fluorescence (XRF) - 1 1 -
Bulk SSD specific gravity ( Gssd ) 2 GSSD = MSSD ! (M SSD - MSSDW ) [i]
Physical (3) Apparent specific gravity ( Gap ) Water absorption ( W% ) ASTM C127 2 2 GAP = MD l(MD - MSSD,W ) w = (mssd -md ) /md X100 [ii] [iii]
Loose ( Pi )& packing density ( Pk ) ASTM C29 2 pk,pl = MSSD ! V [iv]
Water content ( w% ) ASTM D2216 3 w = (Mw /MD ) x 100 [v]
Mechanica l (3) Uniaxial compressive strength ( fc ) Tensile strength ( f ) ASTM D5731 10 fc = Is (50) x factorC ft = 80% x Is (50) [vi] [vii]
Aggregate impact value ( AIV ) EN 1097-2 2 aiv % = (md / m 236) x100 [viii]
Los-Angeles abrasion ( LAA ) ASTM C131 2 LAA% = (M236 / md ) x100 [ix]
Aggregate Crushing Value ( ACV ) EN 1097-2 2 acv % = (md / m236) x100 [x]
Durability (3) Slake durability index ( Id (2) ) ASTM D4644 2 Id(2)% = (MD,2/ MD,j) x100 [xi]
Note: Mssd is mass of SSD sample Mw is mass of moist sample (soaked in water for 24 hours)
Md is mass of oven-dried sample M2 36 is mass of sample passing 2.36 mm sieve after the
test
Mssd W's mass of SSD sample in Md 1 is mass of oven-dried sample after 1st cycle of
water rotation
v is volume of measure Md 2 is mass of oven-dried sample after 2nd cycle of
rotation
2.2.1 Chemical properties
Firstly, a quantitative analysis was carried out to determine the chemical properties of the selected ten rock types. Initially, each collected rubble sample was crushed using a laboratory-scale 'Barmac' crusher to get a powdered form (particle size < 30 ^m) and oven-dried at 105 ± 5 °C for 48 h. XRD test was executed for investigating the mineralogical compositions of rocks using 'Rigaku Ultima IV X-ray Diffractometer'. Crushed powdered granular samples were tested from 2-theta (20) as small as 2° to 90° diffraction angle. XRF test was conducted to analyze the chemical compositions of the collected rock types. 'Spectro - XEPOS XRF Spectrometer' was used for this analysis using the pellet samples prepared from the rubble specimens. Output results such as spectral data and intensity of minerals present in the rocks and quantitative contents of chemical compositions were received.
2.2.2 Physical properties
Surface characteristics were visually observed using 10X magnification digital image process. The composed behavior of various minerals and chemicals in rocks influence the surface textures and porosity. The uniform composition of rocks (made up of uniform-sized minerals) enables higher porosity than rocks
with different-sized minerals. The total porosity of a rock (n) was calculated theoretically using Equation 1 suggested by the studies [35, 36] based on bulk density (Pb ) and particle density (Pd ).
n = (1 - Pb/Pd )x100. (1)
Bulk saturated surface dry (SSD) specific gravity (OSSD), apparent specific gravity (Gap ) and
water absorption (W %) of rocks were analyzed using buoyancy balance with wire mesh basket conforming to ASTM C127. The corresponding equations for calculating the above parameters are mentioned in Table 2 [i-iii]. For the above tests, rock specimens were soaked in water for 24 hours to essentially fill the pores before the testing. Rock usually contains permeable and impermeable voids which determine the rate of water absorption. This influences the quality of rocks for construction works or is suitable for aggregate production. The density of rocks was determined in both loose and compacted form, which are also some of the important properties of rocks, determine the degree of compaction and self-weight of the
structures that resulted from it. After the required calibration of measure, the loose density (p/) and
compacted or packing density (pk) were investigated using 'shoveling' and 'rodding' procedures respectively as described in ASTM C29. Both absorbed water and surface moisture in rock-derived aggregates influence the drying shrinkage of concrete and mortar. Here, the water content of rocks (w%) was determined by mass using 'Method B' as stated in ASTM D2216 and based on the equation provided in Table 2 [v].
2.2.3 Mechanical properties
Determination of uniaxial compressive strength (fc ), tensile strength (ft) and failure patterns of
rocks were investigated using the 'Point Load Index' test method according to ASTM D5731. Strengths
were obtained by performing an irregular lump test to identify the point load strength index (ls (50)) of the
irregular rock samples. Ten rock specimens were selected from each rock type and initially, the preliminary specimen size requirements were checked as per the above standard. The load was applied to the specimens using platen-to-platen end condition attached to the point load tester and the strengths were calculated using the standard charts and the equations provided in Table 2 [vi-vii]. Additionally, the common failure patterns were also observed with the fractured rock specimens, and finally, the water content of the fractured specimens was determined individually. Fig. 4 shows the point load tester used for this analysis with the support conditions.
platen-to-platen end condition
Figure 4. Point load tester.
Rock-derived aggregates should have enough toughness to withstand sudden shocks or impacts. Therefore, the rock sources for aggregates must have sufficient toughness to resist the impacts. The rock impact resistance capacity (AIV) was tested using an aggregate impact tester conforming to EN 1097-2. After enough impacts, mass losses were calculated using the weight of the specimen retained on a 2.36 mm sieve. The equation provided in Table 2 [viii] was used to calculate the AIV of each selected rock in percentage.
When rock aggregates are used in heavy-duty structures, there is a possibility of failure of particles due to the higher abrasion. Therefore, investigating the abrasion resistance of parent rocks may also give additional support to the selection of materials. Los Angeles abrasion (LAA) test was performed here for investigating the abrasion resistance of rocks according to ASTM C131. After the test, mass losses were determined based on the number of fines passing 2.36 mm sieve, and LAA % was determined according to the equation mentioned in Table 2 [ix]. Under the standardized conditions as per EN 1097-2, the crushing test was performed on rock specimens by applying a gradual compressive load using a universal testing machine. The crushing strength of aggregates (ACV) also partially determines the strength of constructions when the aggregates are only exposed to higher stresses. Once the test was carried out, the ACV was calculated in percentage using the weight of crushed particles passing through a 2.36 mm sieve based on the formulae given in Table 2 [x].
2.2.4 Durability properties
The durability of rock is affected mainly due to climatic changes. The cyclic process of wetting and drying can develop permanent strain in rocks, which may result in a volume increase. Due to the changes in temperature and moisture content, the expansion and contraction coefficients of rocks are varied. In this study, the durability of rocks was investigated quantitatively using slake durability test according to ASTM D4644. Rock specimens were undergone for two cycles of wetting and drying, and the mass losses were obtained after the specimens were oven-dried at 105 ± 5 °C for 48 hours. The equation given in Table 2 [xi]
was used for determining the slake durability index (ld (2)) after the two complete cycles of wetting and
drying.
3. Results and Discussion
3.1. Stage 1: Identification of chemical and mineralogical compositions
Initially, complete mineralogical testing was carried out for each collected ten rock types. Table 3 represents the quantities (%) of the most common minerals present in the rock types. Based on the results, it was identified that except Calc-Gneiss, Intrusive Charnockite, and Granitic-Gneiss rocks, all others were abundant with quartz. Also, considerable intensities of feldspar minerals such as albite (plagioclase feldspar - NaAlSi3O8), anorthite (plagioclase feldspar - CaAlSi3O8, k-feldspar (orthoclase feldspar - KAlSi3O8) were noticed in the collected samples. The constitutions of calcite (CaCO3), dolomite (CaMg(CO3)2), pyrite (FeS2), and biotite minerals were very low compared to other available minerals. Here, special attention should be given to illite and biotite minerals, which are from non-expanding clay and mica mineral groups respectively may significantly give adverse effects on the performance of concrete and mortar.
Table 3. Mineral constituents of rocks (%).
e e ti * * * * ar e ti *
Rock type it ilc la im ol ol z tr a *e it e ti ry e it lb p s dl el hi tr o e ti it oi
C D uQ P < 4— ^ n < iB
Charnockite * 1.97 0.23 27.95 1.66 0.01 18.03 27.15 4.36 0.00
Calc-Gneiss 0.00 0.46 2.51 13.43 1.06 1 10.63 9.39 50.33 0.09
Hornblende-Gneiss * 0.62 0.22 49.48 4.23 0.11 10.44 10.68 20.30 0.00
Cordierite-Garnet-Gneiss 0.00 0.18 68.33 9.50 0.15 4.92 3.08 12.53 0.05
Quartzite 1.64 0.26 32.59 4.77 0.08 16.48 10.35 20.32 0.12
Schist 0.44 0.60 30.44 4.73 0.46 14.35 12.41 30.37 0.00
Marble 2.45 1.43 40.11 2.85 0.15 15.99 24.07 8.33 0.00
Intrusive Charnockite * 0.35 0.09 18.06 1.57 0.12 27.90 28.63 19.02 0.00
Granitic-Gneiss * 2.06 0.82 I 21.49 3.07 0.06 36.66 31.13 2.86 0.00
Biotite-Gneiss 0.76 1.11 31.67 3.39 0.06 34.77 12.03 9.27 1 0.35
1st richest rock * Most harmful minerals identified
2nd richest rock * Minerals give positive effects
3rd richest rock * Filtered rocks for Stage 2
A quick chemical/elemental analysis was carried out on the ten rock types and the chemical constituents are listed in Table 4. Rocks selected for this study can be categorized as silicate rocks, which are mainly composed of SiO2 (more than 50 % of total mineral content). Oxides of Al (AhO3) and Fe (Fe2O3) were also identified as abundant compositions in the samples, where the maximum values were noticed in Quartzite, Calc-Gneiss, and Schist rocks. It was noted that the constituents of sulfur trioxide (SO3) and chloride (Cl-) can be negligible in each sample which is the main source, give adverse effects to the function of concrete and mortar. Higher contents of SO3 may affect the dimensional stability, where Zayed et al. [37] found that aggregates contained SO3 beyond 3 %, significantly increased the drying shrinkage of cement. More Cl- content in aggregates gives harmful effects to the corrosion of reinforcements.
Alkali metal oxides such as Na2O and K2O and alkaline-earth metal oxides like CaO and MgO were also identified in the rocks with smaller quantities. Fig. 5 exhibits the diffraction patterns of the ten rock types resulted from the XRD test. The graphs were plotted with the counts against diffracted angle 2-theta (20). The peak intensities mentioned in Table 3 are also marked with the abbreviations of minerals. Various minerals present in the rock-derived aggregates can influence the strength, workability, and durability properties of concrete and mortar. When parent rocks are extracted and crushed for manufactured sand and coarse aggregate production, it enables the reactive form of minerals present in the outputs.
Table 4. Chemical compositions of rocks (%)
Rock type SiO2 Al2O3 Fe2O3 CaO MgO SO3 Na2O K2O Cl-
Charnockite 72.01 7.83 2.09 3.95 0.25 0.00 2.08 2.54 0.00
Calc-Gneiss 54.17 12.10 12.38 8.75 2.76 0.00 1.51 0.67 0.02
Hornblende-Gneiss 73.59 7.59 4.83 3.07 1.02 0.00 1.75 1.33 0.00
Cordierite-Garnet-Gneiss 79.14 5.70 7.05 0.00 0.73 0.00 1.03 1.09 0.00
Quartzite 68.38 10.60 7.01 3.83 0.85 0.00 2.26 1.32 0.00
Schist 66.30 10.56 8.41 4.29 0.83 0.00 2.77 1.18 0.00
Marble 70.51 8.09 1.34 4.43 0.22 0.00 2.05 2.34 0.00
Intrusive Charnockite 70.55 9.23 5.42 3.20 0.24 0.00 2.81 2.43 0.00
Granitic-Gneiss 68.86 9.55 6.89 3.21 0.38 0.00 3.02 2.52 0.00
Biotite-Gneiss 69.61 8.12 2.56 5.49 0.35 0.00 2.28 2.32 0.00
As discussed above, most of the selected rocks revealed higher content of quartz which is a crucial mineral for aggregates as it is always inert. At normal conditions quartz is a less reactive and more controllable tool for concrete and mortar and even at high temperatures, it is more stable with a very strong Si-O bond. In some cases, there is a possibility of the formation of silica gels (which attract water, swells, and form cracks) in concrete as the result of the reaction between very fine crystalline quartz and alkaline substances (typically Ca(OH)2) [38]. Feldspar minerals were the second richest minerals identified in the rocks. Albite can fuse and act as a cementing medium at high temperatures which helps to bind the
materials. K-feldspar mineral usually has a greater hardness, which results in higher strength of cement-based mixes [35]. Anorthite is the least stable feldspar mineral but can encourage less hydration rate during the acceleratory period of cement [40]. Therefore, considering the durability, rocks such as Calc-Gneiss, Quartzite, and Schist should be neglected as the sources for aggregate production.
Figure 5. X-Ray diffraction patterns of rocks.
The availability of carbonate minerals is restricted in the collected samples. Two minerals of the carbonate mineral class were discerned: calcite and dolomite. Increased calcite demands the water requirement for the mixes and decreases free calcium hydroxide and bindings of aluminate phases, which help form stable hydration products [41, 42]. It is better to avoid marble for aggregate production as it showed a considerable percentage of calcite minerals. Dolomite mineral aggregates contribute high compressive strength and resistance against abrasion; however, dolomite contents are seemed identical in each rock type. Pyrite (FeS2) is a sulfide mineral, gives harmful effects to concrete when manufactured sand or aggregates contained it. Pyrite starts to break down and forms sulfuric acid (from the oxidation process) when water and air get into the concrete through small cracks and pores. This enables the expansion of concrete by cracking and allowing more water and air into it [43, 44]. Based on Table 3, Calc-Gneiss rock revealed considerable pyrite content and should be avoided for aggregate production.
Calc-Gneiss and Cordierite-Garnet-Gneiss rocks were identified as the two main sources of illite, which is a clay toxic mineral that directly influences the compressive strength and workability of concrete and mortar. Illite is an intermediate level between smectite (clay mineral) and muscovite (mica mineral), which significantly increases the water requirement in the mixes. It also reduces the strength of concrete and mortar by diminishing the binding capacity between cement paste and aggregates when it acts as a coating [28].
Determination of mica minerals should be highlighted here as they are the most damaging minerals in the aggregates. According to ASTM C294, the mica class contains muscovite, biotite, and chlorite minerals, but based on the XRD patterns, biotite was the only mica mineral identified in the selected rocks. Increased mica minerals in the mixes lead to low workability and considerably lessen the strength and durability with the growth of voids. Among the ten rock types, Calc-Gneiss, Cordierite-Garnet-Gneiss, Quartzite, and Biotite-Gneiss rocks were sparsely detected with biotite; hence, the quarry operators should be vigilant in selecting these rocks for manufacturing aggregates.
Consequently, Charnockite, Hornblende-Gneiss, Intrusive Charnockite, and Granitic-Gneiss were deemed as the most suitable rock types for manufactured sand and coarse aggregate production as they were not highly detected with the potentially harmful minerals for concrete and mortar.
3.2. Stage 2: Filtering the rocks based on availability
Due to the rapid development of infrastructures, the construction activities have been escalated in most of the developing countries including Sri Lanka. It was also identified that; Sri Lanka is now running with the lack of availability of natural resources to fulfill the increased demand for aggregates in construction works. Therefore, proposing alternatives for natural aggregates should also comply with the possibility of continuous usage in the construction works.
As discussed in the previous section, Charnockite, Hornblende-Gneiss, Intrusive Charnockite, and Granitic-Gneiss were concluded as the acceptable rocks for producing aggregates among the ten rock types. To satisfy the persistent construction activities with rock-derived aggregates, this section focuses on the availability of the above four applicable rocks based on a detailed geological map of Sri Lanka. Fig. 6 elucidates the approximate area covered by each rock type mentioned in km2.
Charnockite Hornblende- intrusive Granitic-Gneiss
Gneiss Charnockite
Figure 6. Area covered by the rocks selected from Stage 1.
Charnockite and Hornblende-Gneiss rocks are available in HC of the country. It was identified that Charnockite rock covers the most part, which is around more than 60 % of the total area of HC. This contributes around 40 % of the area of Sri Lanka. Around 15 % area of HC is contributed by Hornblende-
Gneiss rock which is around 9.5 % of the country. Intrusive Charnockite rock was observed in few areas of VC, wherefrom Fig. 6, it was noticed that the rock covers only 351.65 km2. Granitic-Gneiss rock contributes around 25 % of the area of VC, which is around 6 % of the country's terra. Accordingly, Charnockite and Hornblende-Gneiss rocks were designated as the long-term available sources for construction aggregates production and selected for further investigation under Stage 3.
3.3. Stage 3: Checking the properties of most applicable and available rocks
Sections from here describe the analysis of physical, mechanical, and durability properties of the selected rocks from Stage 1 and Stage 2: Charnockite and Hornblende-Gneiss.
3.3.1 Surface texture, porosity, and other physical properties
Fig. 7(a) and 7(c) show the rubble samples of Hornblende-Gneiss and Charnockite collected to study the surface characteristics such as surface texture, crystallization, and surface pores which may influence the properties of concrete and mortar. Fig. 7(b) and 7(d) are the digital images of the representative surface of the above rock types used for this analysis. The surface texture of metamorphic rocks depends on the size, shape, and arrangement of the crystals, uniformity of composition, and degree of isotropy.
Figure 7. (a) Hornblende-Gneiss rubble sample; (b) digital surface image of Hornblende-Gneiss; (c) Charnockite rubble sample; (d) digital surface image of
Charnockite.
From the observations, it was found that Hornblende-Gneiss was configured with medium-sized (0.75-1 mm) grains and 'isometric' crystals which are euhedral cube-shaped with sharp angles, edges, and smooth flat facets. Charnockite was deposited with very coarse-sized (> 2 mm) grains and 'enantiomorphic' shaped crystals which are having optical characteristics. Charnockite crystals, grains, and matrices showed the continuity and non-fabric (uniform arrangement of grains) behavior while Hornblende-Gneiss crystals were formed as fabric (non-uniform arrangement). Minerals were formed as layers (uniform distribution) in Charnockite, but a non-uniform (random) composition of minerals was observed with Hornblende-Gneiss. The above characteristics enhanced Hornblende-Gneiss with rougher surface texture than Charnockite.
The porosity of rocks also depends on the arrangement of grains, composition, and rock type. The calculated porosity values using Equation 1 were 0.25 % for Hornblende-Gneiss and 0.26 % for Charnockite. Both rocks were formed by metamorphism and the slight difference of porosity was due to the variation in grain size, shape, and composition of crystals. Fig. 8 can be referred to for a better understanding of the distribution of surface pores in the above rock types. It was observed that the pores are created at the intersections of grains of the rock. As described above, because of the larger grans in Charnockite, increased pore spaces were noticed than Hornblende-Gneiss.
■W1 ■ ï-. -
■ ] ■ ■ *
Figure 8. Surface pore space structure (a) Hornblende-Gneiss; (b) Charnockite.
Some basic physical properties of the above rock types were determined and the average values of the properties are listed in Table 5. It was identified that Hornblende-Gneiss was slightly denser than Charnockite. Specific gravity at saturated surface dry (SSD) condition, apparent specific gravity, and bulk specific gravity were noticed higher for Hornblende-Gneiss than Charnockite. This can be directly correlated with the porosity values mentioned above where Hornblende-Gneiss revealed lower porosity than Charnockite which resulted in higher specific gravity. Charnockite manifested greater water absorption than Hornblende-Gneiss which was due to the increased pores. Because of the higher unit weight and lower porosity of Hornblende-Gneiss, greater packing density was observed than Charnockite. Hornblende-Gneiss proved an inflated moisture content than Charnockite. Because of the rougher surface texture of Hornblende-Gneiss, there is always a possibility for trapping the moisture particles between the euhedral cube-shaped crystals.
Table 5. Average physical properties of rocks
Properties Hornblende-Gneiss Charnockite
Pd (kg/m3) 2797 2697
gssd 2.804 2.704
Gap 2.816 2.716
W % 0.246 0.261
Pi (kg/m3) 1359 1258
Pk (kg/m3) 1553 1397
w% 0.305 0.267
3.3.2 Strength and other mechanical properties
Strength properties of Hornblende-Gneiss and Charnockite were tested at SSD conditions. An irregular lump test was performed on the non-uniform sized rock specimens used for this investigation by
calculating the equivalent core diameter ( De ) for all specimens to attain a common behavior against the
applied point loads. Hornblende-Gneiss performed well against the applied point loads where increased
uniaxial compressive strength ( fc ) and tensile strength ( ft ) were noticed than Charnockite. It was also
noted that both rocks were weak against the tensile loads as they are brittle in nature. Table 6 tabulates the dimension, failure load, calculated equivalent core diameter, point load index, and the strengths of irregular rock specimens.
Table 6 mentions the failure type of all selected rock specimens where most of the Hornblende-Gneiss specimens manifested axial failure (AF) while a considerable number of Charnockite samples were fractured due to the shear failure (SF). Fig. 9 represents few samples from both rock types and their failure planes. In both cases at the start of load applying, cracks were propagated perpendicular to the stress applied, which ultimately left flat surfaces at the breakpoint. The failure behavior of rocks can be directly influenced by the grain direction and crystallization where layer arrangement of minerals in Charnockite gave the possibility of breaking along the weaker mineral layer, which formed a failure shear plane. However, the absence of the above feature in Hornblende-Gneiss ended up in axial failure. The rock specimens were considered at SSD condition for determining uniaxial compressive strength and tensile strength. Impermeable and permeable pores in rock specimens usually contain water when they are in SSD
condition. Therefore, the water content of rock specimens played a major role in the uniaxial compressive strength and tensile strength. To investigate this behavior, the relationship between water content and uniaxial compressive strength was analyzed and represented by Fig. 10(a). Linear relationships were observed between water content (w%) and uniaxial compressive strength of both rock types (Hornblende-Gneiss: R2 = 0.9594 and Charnockite: R2 = 0.9151).
Table 6. Uniaxial compressive strength, tensile strength, and failure type of rocks
Rock type w (mm) d (mm) P (N) De (mm) F ls (50) (MPa) fc (MPa) ft (MPa) Failure type
1 37 17.5 13300 57.414 1.064 4.294 94.463 3.435 SF
2 35.5 23 17800 64.473 1.121 4.801 105.627 3.841 AF
U) tn <D c o 3 36 25 18500 67.689 1.146 4.627 101.801 3.702 AF
4 39 25 19600 70.453 1.167 4.608 101.367 3.686 SF
¿ T3 5 37 17.5 14800 57.414 1.064 4.778 105.116 3.822 AF
c 6 38 22.5 16700 62.406 1.105 4.738 104.231 3.790 AF
.Q c 7 45 27.5 18500 70.000 1.163 4.393 96.940 3.514 SF
o X 8 37.5 27.5 17700 42.457 1.182 4.389 96.560 3.511 AF
9 42.5 26 19400 71.848 1.177 4.424 97.329 3.539 AF
10 40 25 18500 66.742 1.139 4.729 104.048 3.784 AF
1 52 42 21300 105.445 1.399 2.680 64.323 2.144 SF
2 44 36 22500 89.800 1.301 3.631 87.153 2.905 AF
3 60 52.5 33400 126.635 1.519 3.164 75.939 2.531 SF
e v 4 62.5 53 39900 129.860 1.536 3.635 87.249 2.908 SF
c o 5 53 37 22500 99.916 1.366 3.078 73.862 2.462 SF
c. ro 6 45 32.5 25000 103.544 1.388 3.236 77.655 2.588 SF
O 7 64 56.5 26500 135.679 1.567 2.256 54.141 1.805 AF
8 55 51.5 26800 120.083 1.483 2.757 66.162 2.205 SF
9 48 36 23500 93.793 1.327 3.545 85.090 2.836 SF
10 57 32.5 21000 97.113 1.348 3.002 72.047 2.402 AF
w is sample width; d is sample depth; P is failure/fracture load
De is equivalent core diameter [= 4(w x d) / n] F is size correction factor [= (De / 50)045 ] ls (50) is corrected point load strength index [= F x
( p / D: )]
fc = ls (50) x factorC (for Hornblende-Gneiss = 22; Charnockite = 24) ft = 80% x ls (50)
Figure 9. Failure planes of typical rock specimens with respect to load applied (SF: Shear failure; AF: Axial failure).
110 100 -90 -
r 80
L
I 70 " 60 -
50 -
40
R* = 0,9594
R2 = 0,9151 "
• Hornblende-Gneiss a Charnockite
0.2
0,24
0,28
a)
0,32 w %
0,36
0,4
0,44
Figure 10 (a): Effect of water content on uniaxial compressive strength; (b): Young's modulus vs Corrected point load strength index.
Both relationships showed a decreasing trend from which it can be concluded that uniaxial compressive strength was linearly and negatively affected by water content. Based on the regression
equation derived by Leite and Ferland [45] as provided in Equation 2, Young's modulus (E) of rock specimens was calculated based on porosity (n) in this study.
E = 10.10 - 0.109 x n. (2)
Furthermore, the above regression was correlated with the corrected point load strength index (Is (50)) which were calculated using equivalent core diameter, size correction factor, and failure load in
this study as listed in Table 6. To achieve this, relationships between the E values determined using Equation 2 and Is (50) were plotted as shown in Fig. 10(b). Based on the results, it was observed that Equation 2 and the calculated values in this study were correlated with a high coefficient of determination values. Linear increasing relationships were noticed with E and Is (50) fo r Hornblende-Gneiss (with R2 = 0.9229) and for Charnockite (with R2 = 0.9649).
Multiple linear regression analysis was executed (at 95 % confidence interval) for uniaxial compressive strength of Hornblende-Gneiss (fc Hbl-Gn) (with adjusted R2 = 1.000) and Charnockite
( fc, Chkt) (with adjusted R2 = 1 000) as given in Equation 3(a) and 3(b) concerning porosity (n), water content (w), corrected point load strength index (Is (50)) and density (p).
fc, Hbl-Gn =-8.969 x10-13 + 2.084 x10-12 n-
- 3.268x10 w + 22ls (50) + 5.142x10-17p; fc, chkt = 3.100 x10-13 -1.100 x10-12 n -
- 1.900 x 10-13w + 24ls (50) + 6.000 x 10-17 p.
-17,
(3a)
(3b)
Simple regression was also done (at 95 % confidence interval) to determine the relationship between uniaxial compressive strength and young's modulus of rocks as represented by Equations 4(a) and 4(b). Hornblende-Gneiss (R2 = 0.9299) and Charnockite (R2 = 0.9649) revealed linear relationships.
E,
Hbl -Gn
= 7 x 10-5 f
17 r
c, Hbl - Gn
10.059;
(4a)
Echkt = 1x 10-4 fc, Chkt +10.062. (4b)
The results of other mechanical properties such as 'Aggregate Impact Value' (AIV), 'Los Angeles abrasion value' (LAA), and 'Aggregate Crushing Value' (ACV) of Hornblende-Gneiss and Charnockite are reported here. AIV, LAA, and ACV are the determinations of toughness, abrasion resistance, and crushing resistance of the rock specimens. Table 7 mentions the average AIV, LAA, and ACV of Hornblende-Gneiss and Charnockite, and the recommended values suggested corresponding standards. Lower AIV indicates the rock specimens are tougher or more resistant to impact loads. Based on the provisions given in BS EN 12620:2002, it was concluded that the aggregates produced from Hornblende-Gneiss and Charnockite are safe to be used for heavy-duty concrete floor constructions.
Table 7. Average AIV, LAA, and ACV of rocks
Properties Hornblende-Gneiss Charnockite Limiting values by standards
AIV (%) 22 23 < 25 % for heavy duty concrete floor finishes, BS EN 12620:2002
LAA (%) 60 61 < 30 % for stone matrix asphalt, AASHTO T96-02
ACV (%) 42 40 < 45 % for wearing surfaces, EN 1097-2
Rock abrasion characteristics are not very important in building constructions. However, in this study, LAA was determined to check the adequacy of rock specimens to degradation and disintegration when impact and abrasion loads are applied on the aggregates. It was found that the LAA of both rocks did not comply with the maximum limitations provided by AASHTO T96-02. ACV was determined under a gradually applied compressive load. Rock specimens with lower crushing values indicate the lower crushed fraction under the applied load and more economical performance. ACV of rocks acts a minor role in the constructions which are subjected to higher stresses. According to the test results, the ACV of the selected rocks satisfied the maximum limits and can be suggested for use in the construction of wearing surfaces.
3.3.3 Durability against wetting and drying
The durability of Hornblende-Gneiss and Charnockite rocks was estimated quantitatively after the two cycles of wetting and drying with abrasion. Slake durability index of both rocks were determined after
the second cycle as abbreviated by ' ld (2)'. The higher ld (2) value defines the higher resistance to cyclic
wetting and drying conditions. Results concluded that the corresponding average ld (2) values of
Hornblende-Gneiss and Charnockite rocks were 99.285 % and 99.445 % respectively. Fig. 11 illustrates the observations made during and after the experiment.
Figure 11. Slake durability test apparatus and transformation of rock specimens.
ASTM D4644 suggests three types of rock samples based on the visual observations after two complete cycles of wetting and drying: Type I (retained specimen remain virtually unchanged), Type II (retained specimen consist large and small fragments), and Type III (the retained specimen is exclusively small fragments). Because the retained Hornblende-Gneiss and Charnockite specimens remained virtually
unchanged, it was concluded that they conformed to Type I. According to the classifications of ld (2)
values provided by Franklin and Chandra [46], the rocks selected for this analysis were also termed as
'extremely high' durable rocks. Therefore, aggregates produced from Hornblende-Gneiss and Charnockite can be utilized for any building constructions without uncertainty.
According to the Franklin Rating System [47], Hornblende-Gneiss and Charnockite selected for this study were defined with a rating value (R) based on the average ld (2) and Is (50) values. The
interpolated R -value for Hornblende-Gneiss was observed as 8.26 with ld (2) of 99.285 % and Is (50)
of 4.578 MPa. The R -value for Charnockite was determined as 7.94 with ld (2) of 99.445 % and Is (50) of 3.098 MPa.
Various factors influence the durability of rocks, but the most dominant factors are the rock's mineral and chemical compositions. As observed in Section 4.1, Charnockite and Hornblende-Gneiss rocks are abundant with friendly minerals and chemicals. Anyhow, it was also noticed that there are few percentages of potentially harmful minerals present in those rocks. Therefore, the influence of those minerals on the durability of rocks was investigated and reported with the outcomes from multiple linear regression analysis and F -test. Here, the statistically significant effects of both friendly and harmful minerals on the slake durability indexes of rocks after the 1st cycle ld (1) and 2nd cycle ld (2) of wetting and drying were
examined using F -test results and the relationships were derived with the multiple linear regression analysis as shown in Table 8.
Table 8. Summary of F-test and multiple linear regression analysis for ld (1) and ld (2) with durable and harmful minerals.
F-test (at 95% confidence interval)
Predictor
df
* Statically significant?
Multiple regression analysis (Linear model) (at 95% confidence interval)
ld (1)
Quartz (Qz)
Albite (Ab)
K-feldspar (Kf)
Illite (Il) Pyrite (Py)
0.00015 Yes
0.00102 Yes
0.00021 Yes
0.00761 Yes
1.49302 Yes
0.00013 Yes
0.00093 Yes
0.00019 Yes
0.00694 Yes
ld (1) =98.3+0.003 Qz+0.17 Ab-0.06 Kf adj'
ld (1) =100.0-0.11 II+0.36 Py
R2 = 0.8745
adj. R2 = 0.9403
ld ( 2 ) Quartz (Qz) Albite (Ab) K-feldspar (Kf)
Illite (Il) Pyrite (Py)
ld (2) =92.1+0.01 Qz+1.17 Ab-0.51 Kf adj
1.36055
Yes
ld (2) =99.6-0.03 II-0.81 Py
R2 = 0.8321
adj. R2 = 0.9911
f
3
3
* Statistically significant result: "Yes": when f value < F critical value; "No": when f value > F critical value
df: Degree of freedom
As expected, from the F -test results, the durable and friendly minerals such as quartz, albite, and k-feldspar statistically and significantly affected both slake durability indexes. Also, the detected potentially harmful minerals such as illite and pyrite statistically and significantly affected the durability indexes after the 1st and 2nd cycle of wetting and drying. Here, biotite mineral was not considered as both Charnockite and Hornblende-Gneiss were not detected with considerable quantities. Therefore, based on the above statistical approaches, it was concluded the above five minerals revealed statistically significant effects on the durability of rocks. Consequently, analyzing the above minerals is mandatory when selecting the rock-derived aggregates for high durable applications.
4. Conclusions
This research was executed to implement the most appropriate high-grade metamorphic rocks available in Sri Lanka, for producing rock-derived outputs such as manufactured sand and coarse aggregate. All the rock types were selected from the high-grade metamorphic category having extreme
hardness and strength, which is expected as the ideal solution for construction aggregate manufacturing. It was identified that the demand for natural resources to fulfill the aggregate requirements in the construction activities has been escalated now. Therefore, the contractors are now trying to introduce both fine and coarse aggregates made from parent rocks in the construction works. The problem was noticed that the quarry operators who are producing manufactured sand and coarse aggregates are not paying much attention to selecting the suitable parent rocks. The investigation on analysis of the properties of highgrade rocks concludes the followings:
Initially, for sorting out the appropriate high-grade rocks, chemical and mineralogical tests were performed on the selected ten metamorphic rocks. Among the detected minerals, illite, pyrite, and biotite were considered as the most harmful constituents to the performance of concrete and mortar. Calc-Gneiss, Cordierite-Garnet-Gneiss, and Quartzite rocks showed a significant content of illite. Calc-Gneiss, Schist, and Marble rocks arrived with a considerable pyrite content. A noticeable content of biotite was observed in Calc-Gneiss, Quartzite, and Biotite-Gneiss rocks. Consequently, Charnockite, Hornblende-Gneiss, Intrusive Charnockite, and Granitic-Gneiss were concluded as the most suitable metamorphic rocks for producing aggregates for concrete and mortar.
To inspect the perpetuity of sources for aggregates to be used in constructions, the availability of the above selected four rock types was investigated using a detailed geological map of the country. Charnockite was identified as the most abundant rock type while, Hornblende-Gneiss revealed a considerable percentage of the total land area. Intrusive Charnockite and Granitic-Gneiss rocks underlie very small parts of the country's crust.
Physical, mechanical, and durability properties of the most applicable and plentifully available rocks such as Charnockite and Hornblende-Gneiss were examined. Hornblende-Gneiss manifested rougher surface texture than Charnockite. Other physical properties such as particle density, specific gravities, loose and packing densities, and water content were observed slightly higher with Hornblende-Gneiss than Charnockite. However, Charnockite showed higher water absorption than Hornblende-Gneiss due to the increased porosity.
Study on the strength properties resulted in good uniaxial compressive strength and tensile strength of Charnockite and Hornblende-Gneiss rocks. The AIV, LAA, and ACV of both rocks lied within the acceptable limits, which deduces the applicability of Charnockite and Hornblende-Gneiss derived aggregates in high-load bearing structures. Above both rocks manifested extreme performance against continuous wetting and drying cycles, which showed the pertinency of both rock-derived aggregates in high-durable applications.
As a consequence, high-grade metamorphic rocks such as Charnockite and Hornblende-Gneiss are deemed as the most appropriate and available rock types for producing manufactured sand and coarse aggregates to be used in concrete and mortar.
References
1. Branavan, A., Konthesingha, K.M.C. Fine Aggregate Usage in Concrete and Masonry Mortar by Local Construction Industries. Proceedings of 10th International Conference on Structural Engineering and Construction Management, Kandy. Sri Lanka, 2019. Pp. 106-113.
2. Pereira, K., Ratnayake, R. Water Integrity in Action: Curbing Illegal Sand Mining in Sri Lanka. Berlin, Germany. Water Integrity Network, 2013.
3. Padmalal, D., Maya, K. Sand Mining, Environmental Science and Engineering, 2014. Pp. 57-80.
4. Sing, M.C.P., Love, P.E.D., Tam, C.M. Review and exploration of river sand substitutes for concrete production in Asian countries, 2012. DOI: 10.1201/b13165-25
5. Gavriletea, M.D. Environmental Impacts of Sand Exploitation. Analysis of Sand Market. Sustainability. 2017. 9(1118). Pp. 1-26. DOI: 10.3390/su9071118
6. Branavan, A., Konthesingha, K.M.C., Nanayakkara, S.M.A., Premasiri, H.M.R. Optimizing Blending of Manufactured Sand with Offshore Sand Based on Physical and Virtue Characteristics. Journal of Materials Science Research and Reviews. 2020. 6 (3). Pp. 11-31. [Online]. URL: https://www.journaljmsrr.com/index.php/JMSRR/article/view/30156 (date of application: 08.10.2021)
7. Branavan, A., Konthesingha, C., Nanayakkara, A. Performance evaluation of cement mortar produced with manufactured sand and offshore sand as alternatives for river sand. Construction and Building Materials. 2021. 297. DOI: 10.1016/j.conbuildmat.2021.123784
8. Branavan, A., Konthesingha, C., Nanayakkara, A. Influence of Blended Fine Aggregates on the Performance of Lime - Cement Mortar - A Statistical Approach. Proceedings of 7th International Multidisciplinary Engineering Research Conference (MERCon 2021), IEEE Xplore, Colombo, Sri Lanka. 2021. DOI: 10.1109/MERCon52712.2021.9525765
9. Jayawardena, U., Dissanayake, D. Identification of the most suitable rock types for manufacture of quarry dust in Sri Lanka. Journal of National Science Foundation Sri Lanka. 2008. 36 (3). Pp. 215-218. [Online]. URL: https://jnsfsl.sljol.info/artic-les/10.4038/jnsfsr.v36i3.157/galley/298/download/ ( date of application: 05.05.2020).
10. Jayatileke, S. Geology and Mineral Resources of Sri Lanka. Vidurava. 2015. 30 (1). Pp. 7-14.
11. Piyadasa, A. Prehistory of Sri Lanka: the geographical and geological background of Sri Lanka. 2017. [Online]. URL: https://www.archaeology.lk/5186 (date of application: 14.04.2020).
12. Kroner, A., Cooray, P., Vitanage, P. Lithotectonic Subdivision of the Precambrian Basement in Sri Lanka. In: Summary of Research of the German-Sri Lankan Consortium. Lefkosia, 1991. Pp. 5-21. [Online]. URL: https://www.sdrp.org/(S(351jmbn-tvnsjt1aadkposzje))/reference/ReferencesPapers.aspx?ReferenceID=197417 (date of application: 15.08.2020).
13. van Wyk, P.R., Croucamp, L. Evaluation of rock types for concrete aggregate suitability for the construction of a gravimeter vault and access road at the Matjiesfontein Geodesy Observatory site near Matjiesfontein, South Africa. Journal of the South African Institution of Civil Engineering. 2014. 56 (2). Pp. 30-36. [Online]. URL: http://www.scielo.org.za/scielo.php?scri-pt=sci_arttext&pid=S1021 -20192014000200004 (date of application: 16.08.2020).
14. Branavan, A., Konthesingha, C. Pertinence of Alternative Fine Aggregates for Concrete and Mortar: A Brief Review on River Sand Substitutions. Australian Journal of Civil Engineering. 2021. DOI: 10.1080/14488353.2021.1971596 [Article in Press].
15. Vitanage, P. Tectonics and Mineralization in Sri Lanka. Geol. Soc. Finl. Bull. 1985. 57. Pp. 157-168. DOI: 10.17741/bgsf/57.1-2.013
16. Nawaratne, S.W. Feldspar and Vein Quartz Mineralization in Sri Lanka: A Possible Post Metamorphic Mid-Paleozoic Pegmatitic-Pneumatolytic Activity. Journal of Geological Society of Sri Lanka. 2009. 13. Pp. 83-96. [Online]. URL: http://www.investmentconsultantslk.com/industrial_reports_doc/feldspar%20&%20vein%20quartz%20in%20srilanka.pdf (date of application: 13.04.2020).
17. Geological Map of Sri Lanka, Colombo: Geological Survey Department, Geological Survey and Mines Bureau, Sri Lanka. 1980.
18. Cooray, P. The Geology of Sri Lanka - problems and perspectives. Journal of the Geological Society of Sri Lanka. 1989. 3. Pp. 1-14. [Online]. URL: http://ifs.nsf.ac.lk/bitstream/handle/1/375/PRJ_11_ab.pdf?sequence=2&isAllowed=y (date of application: 05.04.2020).
19. McConnell, D., Steer, D. The Good Earth: Introduction to Earth Science. McGraw-Hill; McGraw-Hill Higher Education. 2nd edn. London, NY, 2010.
20. Prame, W. Crystalline Basement of Sri Lanka: An Overview based on Recent Advances. Journal of the Geological Society of Sri Lanka. 1995/96. 6. Pp. 37-49.
21. Kroner, A. African linkage of Precambrian Sri Lanka. Geologische Rundschau. 1991. 80 (2). Pp. 429-440. DOI: 10.1007/BF01829375
22. Pinna, P., Jourde, G., Calvez. J.Y., Mroz, J.M., Marques, J.M. The Mozambique Belt in northern Mozambique: Neo-proterozoic (1100-850 Ma) crustal growth and tectogenesis, and superimposed Pan-African (800-550 Ma) tectonism. Precambrian Research. 1993. 62(1-2). Pp. 1-59. DOI: 10.1016/0301-9268(93)90093-H
23. Dahanayake, K., Jayasena, H. General Geology and Petrology of some Precambrian Crystalline Rocks from the Vijayan Complex of Sri Lanka. Precambrian Research. 1983. 19 (3). Pp. 301-315. DOI: 10.1016/0301-9268(83)90018-9
24. Malaviarachchi, S.P., Takasu, A. Petrology of Metamorphic Rocks from the Highland and Kadugannawa Complexes, Sri Lanka. Journal of the Geological Society of Sri Lanka. 2011, 14, Pp. 103-122. DOI: 10.13140/2.1.5043.0401
25. Jayawardena, U., Dissanayake, D. Identification of the most suitable rock types for manufacture of quarry dust in Sri Lanka. Journal of National Science Foundation Sri Lanka. 2008. 36(3). Pp. 215-218. [Online]. URL: https://jnsfsl.sljol.info/artic-les/10.4038/jnsfsr.v36i3.157/galley/298/download/ (date of application: 04.08.2020).
26. Tugrul, A., Hasdemir, S., Yilmaz, M. The Effect of Feldspar, Mica and Clay Minerals on Compressive Strength of Mortar. Engineering Geology for Society and Territory. 2014. 5. Pp. 93-96. DOI: 10.1007/978-3-319-09048-1_18
27. Leemann, A., Holzer, L. Influence on mica on the properties of mortar and concrete. Proceedings of 8th EMABM. Athens, Greece, 2011, Pp. 199-204, [Online]. URL: https://www.researchgate.net/publication/304181219_Influence_of_mica_on_the_properties_of_mortar_and_concrete (date of application: 15.07.2020).
28. Norvell, J.K., Stewart, J.G., Juenger, M.C.G., Fowler, D.W. Influence of Clays and Clay-Sized Particles on Concrete Performance. Journal of Materials in Civil Engineering. 2007. 19 (12). Pp. 1053-1059. [Online]. URL: https://ascelibrary.org/doi/10.1061/%28ASCE%290899-1561%282007%2919%3A12%281053%29 (date of applicaion: 16.06.2020).
29. Grasso, J. Reasons for Choosing Quality Aggregate. Monroe Recycling & Aggregates. 2019. [Online]. URL: https://www.monroeaggregate.com/2019/12/18/reasons-for-choosing-quality-aggregate/ (date of application: 05.02.2020).
30. Mitchell, C. Construction aggregates: evaluation and specification. Proceedings of 3rd International Forum for Industrial Rocks and Mining. Fujairah, UAE, 2015. [Online]. URL: http://nora.nerc.ac.uk/id/eprint/510604/ (date of application: 16.05.2020).
31. Loorents, K.L., Johansson, E. Free mica grains in crushed rock aggregates. Bull Eng Geol Environ. 2007. 66. Pp. 441-447. DOI: 10.1007/s10064-007-0091 -4
32. Ugolini, F.C., Corti, G., Agnelli, A., Piccardi, F. Mineralogical, Physical and Chemical Properties of Rock Fragments in Soil, Soil Science. 1996. 161 (8). Pp. 521-542. DOI: 10.1097/00010694-199608000-00007
33. 33. Scott, P.W., Rollinson, G.K. Crushed Rock Aggregates: Their Mineralogy and Textures Using Automated Scanning Electron Microscopy. Proceedings of the 18th Extractive Industry Geology Conference 2014 and technical meeting, Oxford, UK, 2015. Pp. 49-68. [Online]. URL: https://www.researchgate.net/publication/307555555_CRUSHED_ROCK_AGGREGATES_THE-IR_MINERALOGY_AND_TEXTURES_USING_AUTOMATED_SCANNING_ELECTRON_MICROSCOPY (date of application: 24.09.2020).
34. Johansson, E., Miskovsky, K., Loorents, K.J. Estimation of Rock Aggregates Quality Using Analyses of Drill Cuttings. JMEPEG. 2009. 18. Pp. 299-304. DOI: 10.1007/s11665-008-9284-7
35. Rahmouni, A., Boulanouar, A., Boukalouch, M., Samaouali, A., Geraud, Y., Sebbani, J. Porosity, Permeability and Bulk Density of Rocks and their Relationships based on Laboratory Measurements. Romanian Journal of Materials. 2014. 44 (2). Pp. 147-152. [Online].
36. Yu, Y. An Experimental Method to Measure the Porosity from Cuttings: Evaluation and Error Analysis. Texas Tech University, Texas, 2013. [Online]. URL: https://ttu-ir.tdl.org/bitstream/handle/2346/48848/YU-THESIS.pdf?sequence=1 (date of application: 18.04.2020).
37. Zayed, A.M., Brown, K., Hanhan, A. Effects of Sulfur Trioxide Content on Concrete Structures Using Florida Materials. University of Florida, Tampa, 2014. [Online]. URL:
https://trid.trb.org/view/755078#:~:text=Concrete%20strength%20data%20indicate%20that,strength%20loss%20at%20180%20 days.&text=For%20cements%20moderate%20in%20tricalcium,tolerant%20to%20higher%20SO3%20content (date of application: 22.06.2020).
38. Bentz, D.P., Arnold, J., Boisclair, M.J., Jones, S.Z., Rothfeld, P., Stutzman, P.E., et al. Influence of Aggregate Characteristics on Concrete Performance. National Institute of Standards and Technology. Maryland, 2017. DOI: 10.6028/NIST.TN.1963
39. University of Minnesota. Mineral Web Pages, 2010. [Online]. URL: https://www.esci.umn.edu/courses/1001/minerals/potassium_feldspar.shtml (date of application: 12.02.2020).
40. Lagerblad, B., Kjellsen, K.O. Effect of Mineralogy of Fillers on the Cement Hydration. The Modelling of Microstructure and its Potential for Studying Transport Prperties and Durability. 1996. 304. Pp. 157-165. [Online]. URL: https://link.springer.com/chapter/10.1007%2F978-94-015-8646-7_7 (date of application: 19.06.2020).
41. Benachour, Y., Davy, C.A., Skoczylas, F., Houari, H. Effect of a high calcite filler addition upon microstructural, mechanical, shrinkage and transport properties of a mortar. Cement and Concrete Research. 2008. 38 (6). Pp. 727-736. DOI: 10.1016/j.cemconres.2008.02.007
42. Borziak, O.S., Plugin, A.A., Chepurna, S.M., Zavalniy, O.V., Dudin, O.A. The effect of added finely dispersed calcite on the corrosion resistance of cement compositions. iOp Conf. Series: Materials Science and Engineering. 2019. 708. DOI: 10.1088/1757-899X/708/1/012080
43. AGI, American Geosciences Institute, 2020, [Online]. URL: https://www.americangeosciences.org/critical-issues/faq/how-do-pyrite-and-pyrrhotite-damage-building-foundations (date of application: 13.07.2020).
44. Duchesne, J., Fournier, B. Deterioration of Concrete by the Oxidation of Sulphide Minerals in the Aggregate. Journal of Civil Engineering and Architecture. 2013. 7 (8). Pp. 922-931. DOI: 10.17265/1934-7359/2013.08.003
45. Leite, M.H., Ferland, F. Determination of unconfined compressive strength and Young's modulus of porous materials by indentation tests. Engineering Geology. 2001. 59 (3). Pp. 267-280. DOI: 10.1016/S00137952(00)00081-8
46. Franklin, J.R., Chandra, R. The Slake-Durability Test. Int. J. Rock Mech. Min. Sci. 1972. 9 (3). Pp. 325-341. DOI: 10.1016/0148-9062(72)90001 -0
47. Franklin Trow Associates. Field Evaluation of Shales for Construction Projects. Research and Development Project No. 1404, Ministry of Transportation and Communication, Research and Development Branch, Ontario, 1979.
Information about author:
Branavan Arulmoly
ORCID: https://orcid.ora/0000-0001-9058-6567
E-mail: branavanarulmoly@sjp.ac.lk
Chaminda Konthesingha, PhD
E-mail: konthesinaha@sjp.ac.lk
Anura Nanayakkara, PhD
ORCID: https://orcid.ora/0000-0003-4887-0526
E-mail: sman@civil. mrt. ac. lk
Ranjith Premasiri, PhD
E-mail: ranjith@uom.lk
Received 02.01.2021. Approved after reviewing 08.10.2021. Accepted 09.10.2021.
