Construction and Building Materials 51 (2014) 225–234

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier .com/locate /conbui ldmat

Chloride and chemical resistance of self compacting concrete containingrice husk ash and metakaolin

0950-0618/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.conbuildmat.2013.10.050

⇑ Corresponding author. Tel.: +91 4630 261481, mobile: +91 9787211014; fax:+91 4630 261108.

E-mail addresses: kannan_civil81@yahoo.in (V. Kannan), ganesanhari@yahoo.com (K. Ganesan).

V. Kannan a,⇑, K. Ganesan b

a Department of Civil Engineering, National College of Engineering, Tirunelveli, Tamil Nadu, Indiab Department of Civil Engineering, Sudharsan Engineering College, Sathiyamangalam, Pudukkottai, Tamil Nadu, India

h i g h l i g h t s

� Rice husk ash (RHA), metakaolin (MK) and their combination are used as cement replacement materials.� The ternary systems of OPC, RHA and MK, enhance the mechanical properties of self compacting concrete.� The RHA and RHA in combination with MK showed a considerable resistance against acid attack.� Up to 30% the combination of RHA and MK as may be used as supplementary cement replacement materials.� Strong interrelationship occurred between weight loss due to acid attack and Silica ratio of blended SCC.

a r t i c l e i n f o

Article history:Received 18 July 2013Received in revised form 9 October 2013Accepted 31 October 2013Available online 27 November 2013

Keywords:Self compacting concrete (SCC)Rice husk ashMetakaolinChloride permeabilityAcid attackSilica ratio

a b s t r a c t

In this paper, the durability properties of self-compacting concrete (SCC) containing rice husk ash (RHA),metakaolin (MK) and a combination of MK and RHA (1:1 ratio) were evaluated and their relationshipsdiscussed. The durability properties of the various mixtures were studied. The results showed that SCCblended with RHA and a combination of RHA and MK showed a considerable improvement in durabilitythan unblended SCC (100% OPC). However, the performance of SCC blended with MK was unsatisfactoryin an acid environment. In addition, it was found that resistance to acid attack was directly related to thesilica ratio (SR).

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Self-compacting concrete (SCC) was first developed in 1988 inJapan as a partial response to the gradual reduction of skilled laborin the construction industry. The concrete’s qualities can beachieved without segregation and high deformability in threeways, which consist of limiting aggregate content, ensuring a lowwater-powder ratio and the use of superplasticizer [1]. Nowadays,SCC is gaining much popularity throughout the world because ofsome of its interesting structural properties [2]. However, it isnot completely accepted due to higher cost as well as the lack ofstandard specifications and testing procedures. In most cases, itis treated only as a special concrete [3]. The reason for the increas-ing cost of SCC production is the use of higher powder content (ce-

ment), which can be reduced by the use of various mineraladmixtures such as rice husk ash, fly ash, and metakaolin, etc. aspartial replacement of the cement. The mineral admixtures addi-tions also improve the structural properties of the SCC as well asreducing the CO2 emission [4,5].

RHA is an agro-waste and used as it enhances the excellentproperties of concrete; presently, it is well known as a cement-replacing pozzolanic material and there are also a number researchprojects being conducted on it. Usually, it can be obtained by burn-ing rice husks at about 600–800 �C in a controlled manner, whichcauses the formation of amorphous silica with a high surface area[6]. Finely ground RHA is responsible for the high reactivity in ce-ment and is used to reduce the porosity as well as the width of theinter-facial transition zone (ITZ) [7,8].

However, when the replacement level of RHA is increased(more than 10%), the high surface area decreases the workabilityof the SCC [9,10]; this property is unfavorable when it comes togenerating the SCC. Therefore, to increase the workability and

Table 1Sieve analysis and physical properties of fine and coarse aggregate.

Sieve size (mm) Fine aggregate(% of passing)

Coarse aggregate(% of passing)

20 100 10012.5 100 90.110 100 10.44.75 99.9 0.002.36 99.1 0.001.18 83.1 0.000.60 58.3 0.000.30 10.0 0.000.15 0.70 0.00Pan 0.00 0.00Bulk density (kg/m3) 1752 1640Specific gravity (g/cm3) 2.53 2.78Water absorption (%) 2.01 0.36

226 V. Kannan, K. Ganesan / Construction and Building Materials 51 (2014) 225–234

quantity of the total cement replacement level of the SCC, thewater-to-binder (W/B) ratio was increased with 2% superplasticiz-ers (SP) and RHA was used with MK to make ternary blended SCC.

MK is an amorphous material that is obtained by dehydratingkaolin at a temperature of about 800 �C [11]. The high reactivityof MK with cement and its usability to accelerate the cementhydration differentiates it from other pozzolanic materials [12]. Italso accelerates the initial setting time and improves the mechan-ical and transport properties, especially since it can also attain highcompressive strength at an early age [13,14].

From the previous studies, concrete blended with RHA and MKshows better performance in strength and in some properties ofdurability. However, few studies have investigated the durabilityof SCC containing RHA and MK, especially in an acid environment[15]. Therefore, the objectives of this study were to investigate theperformance of binary and ternary blended SCC containing RHAand MK in sulfuric and hydrochloric acid solutions. The effects ofRHA, in combination with and without MK, on durability weredetermined experimentally. In addition, the relationships betweenthe various durability properties were explored. Furthermore, therelationship between the silica ratio (SR) of the binder, SiO2/(Al2-

O3 + Fe2O3) and the resistance to acid attack on SCC was evaluated.

2. Materials and methods

2.1. Materials

Ordinary Portland cement (OPC) conforming to ASTM C 150 (Type1) was used.The sieve analysis of fine aggregate (FA) and coarse aggregate (CA) was carried outin accordance with the ASTM C136 standard provision. The results of the sieve anal-ysis of FA and CA were tabulated in Table 1. The physical properties of FA and CAwere also presented in Table 1. Commercially available MK was used for this study.

Boiled fired RHA residue was collected from a modern rice mill. The mill-firedhusk residue ash was further burnt in a laboratory muffle furnace at a temperatureof 650 �C over a period of one hour [11]. The burnt material was ground in a labo-ratory pulverizer with a disc diameter of 175 mm for 1 h to a mean particle size of6.27 micron meters (lm) before it was used as a cement replacement material.

The physical and chemical analysis of OPC, RHA, and MK was carried outaccording to relevant Indian standard (IS) code provisions. Superplasticizers (SP)were used to increase the workability of SCC [16]. For this work, Sulphonated Nap-thalene Polymers based SP with the specific gravity of 1.220–1.225 was used as ahigh range water reducer (conforming to IS: 9103:1999 and ASTM-C-494 Type ‘F’depending on the dosages used) to improve the performance of SCC.

2.2. Mix proportion and preparation of the specimens

The mix design, which was based on previous studies, was modified using EFN-ARC guidelines [17]. In general, self-compactability can be greatly affected by thematerial properties and the mix proportion. In the trail mix, the fine and coarseaggregate contents are fixed so that self-compactability can easily be achieved byadjusting the water powder ratio and the superplasticizer dosage only. From thetrail mix, a suitable mix proportion was taken for further study. In this study, the

ratio of fine and coarse aggregate was fixed at 1.1, with a W/B (W/(C + RHA orMK or RHA + MK) ratio of 0.55 and 2% of the superplasticizer; the only variablewas RHA and MK to OPC.

The mix design was carried out to produce SCC without segregation and bleed-ing, with the target mean compressive strength of 38.5 N/mm2 (as M30 graded nor-mal vibrated concrete) at 28 days. For this study, a total of seventeen concretemixtures (RHA, MK with a range of 0%, 5%, 10%� � �30%, and a combination of RHAand MK with a range of 10%, 20%, 30% and 40% with one normal SCC) were prepared,and all the mixtures satisfied the target mean strength. These mixes were desig-nated as OPC (100%) and RHA5/MK5/RHA5 + MK5� � �RHA30/ MK30/RHA20 + MK20.The mix proportions are presented in Table 2.

During the production of SCC, the mixing order is very important to obtainhomogeneity and uniformity in all mixtures [17–19]. Initially, the batching processis carried out and all the materials were separately placed on a nonporous plate.Mixing sequences consisting of FA and CA are mixed for 30 s in the laboratory mix-ing machine to achieve homogeneity, then about 50% of the mixing water is addedto the mixer machine and mixing is continued for one more minute. Thereafter, themixing process is stopped to allow the aggregates to absorb the water for one min-ute. Before adding the cement and admixtures (RHA or MK or RHA + MK), they aremixed in the dry state, then added to the mixing drum. Finally, the SP is poured inthe remaining water and introduced to the mixture, and mixing is restarted for5 min. The mixed concrete is assessed to check its fresh state properties and thenplaced in the required molds for curing.

For all mixes, three specimens of 100 mm3 were cast for a compressive strengthtest; six specimens of 100 mm3 were cast for acid attack tests (H2SO4 and HCl); ninecylindrical specimens with a diameter of 100 mm and height of 50 mm were castfor permeability-related property tests (WA, Sorptivity and chloride penetration).After casting, all of the specimens were left in their casts for 24 h and then theywere unmolded and immersed in a water curing tank until they were requiredfor testing.

2.3. Testing methods

To check the fresh state properties such as filling ability, viscosity, and passingability of concrete, slump flow, V-funnel, and L-box tests were conducted accordingto European Federation of National Associations Representing the producers andapplicators of specialist building products for Concrete (or EFNARC) specifications[17]. Mineralogical and mean particle size analyses of RHA and MK were carriedout by X-ray diffraction (XRD). In addition, Scanning electron microscopic (SEM)with energy dispersive X-ray analysis (EDAX) were carried out to study the mor-phological behavior of concretes.

The compressive strength, permeability related tests (WA, Sorptivity and chlo-ride penetration) and the acid attack tests were conducted after 28 days of watercuring. The compressive strength tests were carried out according to IS 9013-1997.

Saturated water absorption values of RHA, MK and a combination of RHA andMK blended SCC specimens were measured according to ASTM C 642 and previousstudies [20].

Three specimens were used for sorptivity measurement. Measurements of cap-illary sorption were carried out using specimens preconditioned in the hot air ovenat about 50 �C until a constant weight was obtained. Then the concrete specimenswere cooled down to room temperature. As shown in Fig. 1, the test specimenswere exposed to the water on one face by placing them on a pan. The side facesof the specimens were coated with epoxy resin. The water level in the pan wasmaintained at about 5 mm above the base of the specimens during the experiment.

At suitable time intervals, each specimen was removed from the water, with ex-cess water removed by damp paper towel, and then the specimen was weighed. Itwas then replaced in the water and the stopwatch started again. The gain of massper unit area over the density of water is plotted versus the square root of theelapsed time. The slope of the line of best fit of these points was taken as the sorp-tivity value. The sorptivity values of RHA, MK and the combination of RHA and MKblended SCC specimens after 28 days of water curing were evaluated by the follow-ing expression (Eq. (1)):

i ¼ S t1=2 ð1Þ

where i is the cumulative water absorption per unit area of inflow surface (m3/m2),S is the sorptivity (m/s 1/2), and t is the time elapsed (s).

The chloride ion penetration in all SCC specimens was determined by a rapidchloride ion penetration test (RCPT). The resistance to chloride ion penetration interms of total charge passed in coulombs of RHA, MK and the combination ofRHA and MK blended SCC specimens was measured according to the ASTM C1202 standard.

The acid resistance of the SCC specimens was assessed by immersing them intosulfuric acid (H2SO4) and hydrochloric acid (HCl) solutions. The specimens weretested after 28 days of curing and all the specimens were cleaned using a brush,in order to remove any loose material before testing. The initial weight was mea-sured and then the specimens were immersed into either a 5% sulfuric acid(H2SO4) or 5% hydrochloric acid (HCl) solution. A separate plastic beaker was usedfor each specimen for identification purposes. The solutions were replaced at regu-lar intervals (every week) to maintain a constant concentration throughout the test

Table 2Mix proportions of RHA, MK and combination of RHA and MK blended SCC.

Mix ID Mix description Mix systems W/B ratio Quantities (kg/m3)

Water Binder OPC RHA MK SP FA CA

M1 OPC (100%) Control 0.55 220 400 400 0 – 8 880 800M2 RHA5 Binary 0.55 220 400 380 20 – 8 880 800M3 RHA10 Binary 0.55 220 400 360 40 – 8 880 800M4 RHA15 Binary 0.55 220 400 340 60 – 8 880 800M5 RHA20 Binary 0.55 220 400 320 80 – 8 880 800M6 RHA25 Binary 0.55 220 400 300 100 – 8 880 800M7 RHA30 Binary 0.55 220 400 280 120 – 8 880 800M8 MK5 Binary 0.55 220 400 380 – 20 8 880 800M9 MK10 Binary 0.55 220 400 360 – 40 8 880 800M10 MK15 Binary 0.55 220 400 340 – 60 8 880 800M11 MK20 Binary 0.55 220 400 320 – 80 8 880 800M12 MK25 Binary 0.55 220 400 300 – 100 8 880 800M13 MK30 Binary 0.55 220 400 280 – 120 8 880 800M14 RHA5 + MK5 Ternary 0.55 220 400 360 20 20 8 880 800M15 RHA10 + MK10 Ternary 0.55 220 400 320 40 40 8 880 800M16 RHA15 + MK15 Ternary 0.55 220 400 280 60 60 8 880 800M17 RHA20 + MK20 Ternary 0.55 220 400 240 80 80 8 880 800

Fig. 1. Sorptivity test setup.

V. Kannan, K. Ganesan / Construction and Building Materials 51 (2014) 225–234 227

period. The specimens were weighed at regular intervals (every week) for 12 weeks.At the end of the immersion period the total weight loss for each specimen wascalculated.

Water absorption was chosen as an indicator of durability and its correlationwith other durability properties was determined using linear and exponentialregression. In order to ascertain the type and nature of interdependence betweenacid resistance (H2SO4 and HCl) and the silica ratio, SiO2/(Al2O3 + Fe2O3) the totalweight loss due to sulfuric and hydrochloric acid attack were correlated with thesilica ratio, SiO2/(Al2O3 + Fe2O3) by linear regression.

3. Results and discussion

3.1. Physical and chemical analysis of OPC, MK and RHA

The physical and chemical properties of the OPC, RHA and MKthat were tested are presented in Tables 3 and 4. Additionally,the particle size distribution curve for the binding materials is pre-sented in Fig. 2. From the particle size distribution curve, it isclearly noted that RHA, MK and their combinations are finer thanOPC. It is also confirmed from the fineness and particle size ofthe binder materials in physical properties results (Table 3).

The XRD patterns of RHA and MK that were used for this studyare shown in Fig. 3. By comparing with other sources [8,12,13,20],XRD analysis showed that the RHA was mainly in amorphous silicaform; the MK was also in an amorphous form, but deviated slightlyto a crystalline form at an angle of 26.8092�. Silica in the RHA ini-tially exists in an amorphous form, but may become crystallinewhen rice husk is burnt about 650 �C for 1 h. In addition, the silicain the RHA does not remain amorphous when combusted for a pro-longed period (1 h) under oxidizing conditions and then the quartzphase may be formed [21].

3.2. Fresh state properties

The fresh state properties of SCC containing RHA, MK or a com-bination of RHA and MK were studied and are presented in Figs. 4–6,respectively.

The slump flow values of SCC with RHA, MK, and RHA + MK areshown in Fig. 4. From the results, the slump flow values for differ-ent concrete mixes were calculated in the range of 495–740 mm.According to values recommended by EFNARC for fresh state prop-erties of SCC as presented (Table 5), all the mixtures examined fallunder the categories of slump flow classes 1 and 2 (SF1 and SF2)except RHA30 and RHA20 + MK20 mixes. The SF1 and SF2 classesin concrete mixes are used to indicate that these mixes are suitablefor applications such as deep foundation construction (SF1) and fornormal applications such as the building of columns (SF2). TheRHA30 and the RHA20 + MK20 mixes did not meet the EFNARCstandard specifications. From the results, it can be clearly notedthat the slump flow (or filling ability) value gradually reduced withthe increments of the replacement level of RHA, MK and RHA + MK.This condition may be caused by the high reactivity and higher sur-face area of RHA and MK when compared to OPC. It also may bedue to the lowest fineness modulus of fine aggregate (FA) (seeTable 1). Similar trends in the slump flow values were reportedin previous studies [7,12,22].

The V-funnel times for different concrete mixes appear in Fig. 5.From the results, it can be noted that the V-funnel times varied inthe range of 3.9–8.4 s; all concrete mixes could be categorized intothe VF1 class except for the 20% RHA + 20% MK mix. According tothe EFNARC guidelines, a V-funnel time when exceeding 25 s only,it is not recommended (Table 5). From the results, the V-funneltimes for all concrete mixes were satisfied this requirement.

The L-box test results are shown in Fig. 6. From the results, theblocking ratio for different mixes varied from 0.59 to 0.94. A satis-factory blocking ratio was observed in up to 15% RHA, 15% MK, and30% RHA + MK mixes; the value of the ratio for the rest of the mixeswas found to be outside the EFNARC-recommended values. In aprevious study, Felekoglu et al. [23] concluded that a blocking ratiofrom 0.6 to 1 is acceptable for SCC to obtain satisfactory filling abil-ity. In this regard, all concrete mixes were satisfactory (that is,within the prescribed range) except for the 30% MK mix. In sum-mary, all of the mixes exhibited satisfactory fresh state properties(according to the criteria established by EFNARC and previousstudies) except for the 30% RHA, 30% MK, and 20% RHA + 20% MKmixes.

Table 3Physical properties of OPC, RHA and MK.

Materials Bulk density in (g/cm3) Specific gravity(g/cm3)

Fineness passing45l sieve (%)

Specific surface Area Mean particle size inmicron meter (lm)

Loose Dense Blain’s in (m2/kg) BET’s in (m2/g)

OPC 1.18 1.27 3.13 86 318 – 23.4RHA 0.46 0.51 2.08 91 943 36.47 6.27MK 0.50 0.52 2.58 99 2350 – 3.79

Table 4Chemical composition of OPC, RHA and MK (%).

Material SiO2 Al2O3 Fe2O3 CaO MgO Na2O K2O LOI

OPC (%) 20.25 5.04 3.16 63.31 4.20 0.08 0.51 3.08RHA (%) 87.89 0.19 0.28 0.73 0.47 0.66 3.43 4.36MK (%) 51.80 43.75 0.82 0.09 0.03 0.07 0.02 0.34

LOI- Loss on ignition.

Fig. 2. Particle size distribution curve for OPC, RHA, MK and RHA + MK.

Fig. 3. X-ray diffraction pattern of RHA and MK.

228 V. Kannan, K. Ganesan / Construction and Building Materials 51 (2014) 225–234

3.3. Compressive strength

From the results, it can be seen that SCC blended with MK andMK + RHA have a higher compressive strength than normal SCC(SCC with 100% OPC) (Fig. 7). The strength improvement of SCCblended with MK deteriorates with age; a finding that is confirmedby previous studies [6]. The compressive strength of SCC blendedwith 15% RHA was higher than that of normal SCC. However, thecompressive strength of SCC decreased when the quantity of RHAwas greater than 20%. This may occur due to the decreasing work-ability of the SCC blended with RHA, because of the higher surfacearea of RHA and its relative water demand [13,14,24]. The adverseself-compactibility of the SCC blended with RHA (exceeding 15%RHA) may be responsible for permeable voids in the concreteand a subsequent reduction in strength. This demonstrates thebenefit of using RHA in combination with MK to produce SCC witha higher cement replacement of around 40% (20% MK + 20% RHA).A comparison of the results shows that the compressive strengthsimproved up to 15% with RHA, up to 20% with MK and up to 30%with RHA + MK.

3.4. Water absorption

Results for permeability properties (water absorption, sorptivi-ty, and chloride penetration) are presented in Table 6. When MKand RHA + MK were used to partially replace OPC in SCC, the waterabsorption improved considerably (up to 30% and 40%,

respectively) and was lower than that for unblended SCC (100%OPC) [6]. However, SCC blended with RHA tended to show adverseperformance with respect to water absorption when the replace-ment level exceeded 15% by weight of binder. This adverse perfor-mance is attributable to RHA’s higher surface area, lower finenessmodulus value of FA, and subsequent water demand in the con-crete mixing process in comparison with unblended SCC. Withfresh state SCC blended with RHA, the workability gradually re-duced when RHA exceeded 15%. This reduction in the workabilityof RHA-blended SCC creates voids on the concrete, and may beresponsible for subsequent poor resistance to water absorption.

3.5. Sorptivity

The sorptivity is a key parameter in controlling the liquid trans-port in concrete and improving its durability [15]. From the test re-sults, (Table 6), it can be seen that the sorptivity progressivelydecreases with increasing RHA content up to 15%, MK content upto 20% and RHA + MK content up to 30%. SCC blended with MKand RHA + MK showed significant sorptivity due to the fact thatMK is finer than OPC, RHA and produce an additional calcium sili-cate hydrate (C–S–H) gel. In addition, MK blocks existing pores andalters pore structure [25], confirming that the addition of MK leadsto a reduction in the pore space. While considering RHA blendedSCC, the sorptivity values were increased for RHA20–RHA30 mixes.It implies when the replacement levels of RHA exceed 15%, the mi-cro pores may be developed on the concrete due to the lack ofworkability than other mixes. It was also observed from the sorp-tivity data that SCC specimens blended with 30% MK and 20%RHA + 20% MK showed a 18.82% and 19.10% reduction in sorptivityat 28 days respectively, compared to unblended SCC.

3.6. Chloride penetration

Chloride penetration for all concrete mixes in terms of the totalcharge passed is also presented in Table 6. Results for chloride

Fig. 4. Slump flow values for different mixtures.

Fig. 5. V-Funnel times for different mixtures.

Fig. 6. Blocking ratio (L-box test) for different mixtures.

V. Kannan, K. Ganesan / Construction and Building Materials 51 (2014) 225–234 229

penetration were compared with ASTM C1202 standards. As perASTM C1202, total charge-passed values of between 0 and 100,

100 and 1000, 1000 and 2000, 2000 and 4000, and 4000 and5000 are considered to correspond with ‘‘negligible’’, ‘‘very low,’’

Table 5EFNARC recommended values for fresh state properties of SCC.

Slump flow test V-funnel test L-box test

Slump flow classes Slump flow (mm) Viscosity classes V-funnel times (s) Passing ability classes Blocking ratio (H2/H1)

SF1 550–650 VF1 68 PA1 P0.8 with 2 barsSF2 660–750 VF2 9–25 PA2 P0.8 with 3 barsSF3 760–850

Fig. 7. Compressive strength of RHA and MK blended concrete.

Table 6Permeability related properties of RHA, MK and combination of RHA and MK blended SCC

Mix designation MK/RHA/MK + RHA (%) Water absorption (%)

OPC (100%) 0 4.54RHA05 5 4.53RHA10 10 4.1RHA15 15 3.93RHA20 20 3.92RHA25 25 4.47RHA30 30 4.92MK5 5 3.59MK10 10 3.57MK15 15 3.48MK20 20 2.88MK25 25 2.83MK30 30 2.78RHA5 + MK5 10 3.23RHA10 + MK10 20 3.02RHA15 + MK15 30 2.98RHA20 + MK20 40 3.17

Fig. 8. Weight losses due to immersion in 5% sulfuric acid solution

230 V. Kannan, K. Ganesan / Construction and Building Materials 51 (2014) 225–234

‘‘low,’’ ‘‘moderate,’’ and ‘‘high possibility’’ of chloride penetration,respectively, through the concrete specimen. The chloride ion per-meability of the SCC blended with RHA, MK and RHA + MK de-creased up to 15%, 30% and 40%, respectively. The SCC blendedwith MK and RHA + MK showed the best performance for chlorideion penetration. The total charges passed values of 28.23, and25.43 coulombs for SCC blended with MK and RHA + MK, respec-tively, with these values being rated as ‘very low’ for chloride ionpenetration. As reported in previous studies [9,26], the resistanceto chloride ion penetration is significantly improved through thepozzolanic reaction of MK with Ca(OH)2 and the discontinuity ofthe porosity network due to the fineness and chemical propertiesof the material. The mixtures with a higher content of MK andRHA + MK (with high amounts of alumina contents) exhibited amuch better resistance to chloride ion penetration. From the re-sults, it was also noted that the total charge passed for all blendedmixes were lower (in the category of low) than the unblended SCC.Previous studies have also confirmed that the chloride ion perme-ability of SCC with pozzolanic admixtures is lower than that of un-blended SCC [27].

.

Sorptivity � 10�6 (m/s1/2) Total charge passed by RCPT (C)

3.56 1486.283.64 438.623.43 389.183.31 306.224.06 876.966.41 904.79.2 10891.75 431.232.69 299.322.5 292.982.29 59.742.5 34.672.89 28.233.38 286.722.71 173.62.64 86.32.88 25.43

of (a) RHA concrete (b) MK concrete (c) RHA + MK concrete.

Fig. 9. Deterioration of (a) RHA blended SCC, (b) MK blended SCC and (c) RHA + MKblended SCC specimens after 12 weeks of immersion in 5% sulfuric acid solution.

Table 7EDAX spectrum results.

Elements Mass (%)

100% OPC 15% RHA 20% MK 15% RHA + 15% MK

C K 3.32 1.68 3.05 4.6O K 18.75 30.56 21.57 15.3Mg K 0.78 9.59 – –Al K 5.37 6.47 8.76 3.76Si K 14.79 23.23 19.75 38.03Ca K 52.04 22.56 51.78 33.64Fe K 3.07 4.77 4.94 –Cu K 1.88 – – 3.25K K – 1.14 – 1.43

V. Kannan, K. Ganesan / Construction and Building Materials 51 (2014) 225–234 231

3.7. Sulfuric acid attack

The results from the specimens exposed to 5% sulfuric acid(H2SO4) showed that the SCC blends with RHA and RHA + MK were

Fig. 10. Weight losses due to immersion in 5% hydrochloric acid solu

more resistant against sulfuric acid attack than was SCC blendedwith MK (Fig. 8). The lowest weight losses in SCC blended withRHA, MK and RHA + MK were obtained at replacement levels of25% for RHA, 5% for MK, and 40% for RHA + MK. From the results,it can be also noted that the weight of the specimen increases grad-ually between 7 and 14 days depending on the mix. Weight gainmay occur due to the filling of specimen pores by acid solutions.Upon visual inspection (Fig. 9), it was clear that the surface ofthe specimens in which SCC was blended with MK started to dete-riorate after 2 weeks in the acid solution; the level of deteriorationgradually increased for up to 12 weeks. Weight loss and deteriora-tion in specimen condition was greatest in the SCC blended withMK due to the high concentration of alumina in MK, which pro-duces calcium sulfoaluminate (ettringite). Previous studies haveconfirmed that the formation of ettringite depends purely on thereaction between alumina (Al2O3) and sulfate (SO4) and the ratioof the two compounds [28]. The formation of ettringite leads toexpansion in the concrete, which causes disruption of the set ce-ment paste [29]. In contrast, RHA and RHA + MK contain lower alu-mina content with higher silica content compared to MK. It is alsoconfirmed by the EDAX spectra analysis results (Table 7). The SCCblended with RHA and RHA + MK showed better resistance to sul-furic acid attack for this reason.

3.8. Hydrochloric acid attack

The results of the blended SCC specimens exposed to 5% hydro-chloric acid (HCl) solutions show that the lowest weight losseswere obtained at the replacement level of 20% RHA, 5% MK and40% RHA + MK (Fig. 10). Similar to the sulfuric acid attack, SCCblended with RHA and RHA + MK performed best in the hydrochlo-ric acid solution. SCC blended with MK was less resistant to thehydrochloric acid compared to SCC blended with RHA andRHA + MK. It was also observed that the weight loss for all mix-tures in the hydrochloric acid solution was lower than in the sulfu-ric acid solution. Other than SCC blended with 30% MK, allmixtures showed better resistance (lower weight loss) to thehydrochloric acid attack than unblended SCC (100% OPC). This isdue to the fact that the RHA and MK in fresh state SCC reducethe capillary pores by the formation of a new C–S–H gel from thepozzolanic reaction and portlandite hydrate resulting from the ce-ment hydration [30]. In the visual inspection, the differences in thedegraded state of the various specimens after 12 weeks of immer-sion in 5% hydrochloric acid solutions were apparent (Fig. 11).While most of the concrete mixtures had undergone a smallchange, the cubic form of the SCC-MK25 and the SCC-MK30 waslost.

tion of (a) RHA concrete (b) MK concrete (c) RHA + MK concrete.

Fig. 11. Deterioration of (a) RHA blended SCC, (b) MK blended SCC and (c)RHA + MK blended SCC specimens after 12 weeks of immersion in 5% hydrochloricacid solution.

232 V. Kannan, K. Ganesan / Construction and Building Materials 51 (2014) 225–234

3.9. Scanning electron microscopic (SEM) analysis

SEM images and the EDAX spectra of unblended SCC (100%OPC), 15% RHA, 20% MK, and 15% RHA + 15% MK mixes are shownin Fig. 12 and Table 7 respectively. From the SEM image inFig. 12(a), it can be observed that the unblended concrete sample

Fig. 12. Scanning electron microscopic (SEM) image of (a) unblended SCC (100% OPC), (b)

consisted of irregular particle with micro pores; this may be themain reason for the poor performance in relation to strength anddurability compared to blended concretes. However, in the caseof the SEM images of blended concrete with RHA and MK(Fig. 12b–d), it can be seen that the samples were not porous.Due to the hydration progress, the pore structure of the RHA andMK blended concretes were greatly reduced, giving a more uniformstructure than in the unblended SCC. The SEM image also showsthat the ettringite (a spine-like crystal present in unblended SCC)is much reduced due to the pozzolanic reaction of RHA and MK(see Fig. 12) [31]. From the EDAX spectrum analysis (Table 7), itcan be clearly noted that the silica content of blended concretewith RHA and MK is greater than that of unblended SCC. The highersilica content observed in the 15% RHA + 15% MK mix indicatesthat a good pozzolanic reaction and synergic effect has occurredin OPC + RHA + MK [32].

3.10. Relationships between durability properties and the effect ofsilica ratio (SR)

The relationships between the various durability properties areimportant in correlating mix formulations with long term durabil-ity parameters [15]. The regression of water absorption with otherdurability parameters shows a poor correlation for binary blendedSCC containing RHA + MK, whereas ternary blended SCC containingRHA + MK showed a very good correlation between water absorp-tion and the other durability properties (R2 = 56.03% with sorptiv-ity, R2 = 99.60% with chloride penetration, R2 = 98.27% with weightloss from sulfuric acid attack and R2 = 99.33% with weight lossfrom to hydrochloric acid attack) (Table 8). This implies a beneficialsynergistic effect of RHA + MK replacement on different durabilityrelated properties.

The synergic effect may have occurred because of the goodpacking behavior of granular materials (RHA and MK) in cement,as well as the cement hydration process. In previous studies[12,33], it has been reported that the particle size of the pozzolanicmaterials may play an important role in the packing density of theconcrete. Since the RHA and MK particle size (see Table 3 and

15% RHA blended SCC, (c) 20% MK blended SCC and (d) 15% RHA + 15% MK blended.

Table 8Regression between other durability properties (y) and Water absorption (x) of RHA, MK and RHA + MK blended SCC.

Mix Fit Sorptivity in(m/s1/2)

Chloride penetrationIn (C)

Weight loss due toH2SO4 in (%)

Weight loss due toHCl in (%)

RHA blended SCC Linear y = a + bx A �1.371 � 10�5 �1.927 � 103 0.063 1.153B 4.262 � 10�6 624.28 1.927 1.074R2 0.41474 0.149 �0.179 �0.110

Exponential y = a � bcx A 5.811 � 10�4 1.2421 � 104 9.960 6.291B 5.951 � 10�4 1.4692 � 104 6.017 � 1029 5.093 � 1015

C 0.992 0.947 3.933 � 10�8 1.100 � 10�4

R2 0.267 �0.063 �0.175 �0.155

MK blended SCC Linear y = a + bx A 1.115 � 10�6 �2.260 � 103 39.505 9.967B 4.842 � 10�7 779.79 �5.808 �0.654R2 0.475 0.880 0.706 �0.125

Exponential y = a � bcx A 2.470 � 10�4 3.093 � 105 14.969 6.867B 2.459 � 10�4 3.116 � 105 �2.327 � 103 �6.469 � 1018

C 0.998 0.997 0.140 2.730 � 10�7

R2 0.343 0.849 0.881 0.579

RHA + MK blended SCC Linear y = a + bx A 1.418 � 10�6 �2.556 � 103 �12.729 �1.181B 4.841 � 10�7 889.64 6.341 2.197R2 0.525 0.997 0.740 0.911

Exponential y = a � bcx A 3.606 � 10�6 2.600 � 105 15.483 8.934B 0.001 2.625 � 105 1.137 � 105 601.262C 0.079 0.996 0.043 0.180R2 0.560 0.996 0.982 0.993

Fig. 13. Relationship between weight loss due to sulfuric acid attack and silica ratioof RHA, MK and RHA + MK blended SCC.

Fig. 14. Relationship between weight loss due to hydrochloric acid attack and silicaratio of RHA, MK and RHA + MK blended SCC.

V. Kannan, K. Ganesan / Construction and Building Materials 51 (2014) 225–234 233

Fig. 2) is less than that of cement, RHA and MK can fill the cementparticle gaps effectively. It should also be noted that the synergiceffect may have occurred due to the effect of SCM’s (RHA andMK) on the cement hydration process [7,12]. With the additionof RHA to cement, the hydration is increased at a later stage [34],whereas in MK blended cement, hydration increases earlier thanRHA [35]. Therefore, the combination of RHA and MK blended ce-ment may increase the hydration at both the early and late stages.This indicates that the combination of RHA and MK can have agood synergic effect on the strength and durability properties ofSCC.

The linear regression of the total weight loss due to sulfuric acidattack and the silica ratio (SR)(SiO2/Al2O3 + Fe2O3) shows that theyare well correlated with each other (R2 = 71.11% for SCC blendedwith RHA, 82.53% for SCC blended with MK, and 99.85% for SCCblended with RHA + MK) (Fig. 13). From the analysis, it is to benoted that the sulfuric acid resistance is dependent on the SR forRHA, MK, and RHA + MK blended SCC, implying that the SR signif-

icantly influences concrete’s resistance to sulfuric acid. While con-sidering weight loss due to sulfuric acid attack for various mixes(Fig. 8), considerable resistance against sulfuric acid attack was ob-served in the SR range between 2.1 and 5.5. When the SR is lessthan 2.1 (mainly obtained in MK-blended SCC), the sulfuric acidresistance of concrete considerably decreased in this study anddeterioration occurred. Conversely, the correlation between SRand weight loss due to hydrochloric acid attack is not perfectly cor-related (R2 = 34.93% for SCC blended with RHA, 54.93% for SCCblended with MK, and 80.23% for SCC blended with RHA + MK)(Fig. 14). From this analysis, it is noted that the correlation be-tween SR and weight loss due to HCl attack is strong only forSCC blended with RHA + MK. In the same way, when consideringweight loss due to HCl attack (Fig. 10), all mixes showed strongerresistance against HCl in comparison with unblended SCC (100%OPC). Hence, the useful SR range for blended SCC with RHA, MK,and RHA + MK to help maximize HCl acid resistance is from 1.7to 4.8.

234 V. Kannan, K. Ganesan / Construction and Building Materials 51 (2014) 225–234

4. Conclusions

Based on the experimental studies presented in this paper, thefollowing conclusions may be drawn:

(1) As high as 40% by weight of OPC can be replaced with a com-bination of RHA and MK without any adverse effect on thedurability of SCC.

(2) Replacement with 30% MK in blended SCC mixtures leads toa beneficial improvement in the strength and permeability,when compared to those of unblended SCC. However, in anacid environment the SCC blended with MK gave an unfavor-able performance.

(3) Even though the compressive strength of SCC blended withRHA was lower than SCC blended with MK at a 30% replace-ment level, it showed a beneficial improvement in durability(especially acid attack).

(4) From the regression analysis, strong correlations were foundfor SCC blended with RHA + MK as follows:

These strong correlations clearly indicate that the combinationof MK and RHA caused them to act as highly reactive pozzolanicmaterials in SCC.

(a) Water absorption and chloride penetration (R2 = 99.73%).(b) Water absorption and weight loss due to sulfuric acid attack

(R2 = 98.27%).(c) Water absorption and weight loss due to hydrochloric acid

attack (R2 = 99.33%).(d) The silica ratio and sulfuric acid attack (R2 = 99.85%).(1) There was a good synergistic effect between RHA and MK on

the mechanical properties and durability of SCC. The SCCmixture with 15% RHA and 15% MK was the most durableand it may be considered as an optimum level of replace-ment for OPC as supplementary cementitious materials.

References

[1] Okamura Hajime, Ouchi Masahiro. Self-compacting concrete. J Adv ConcrTechnol 2003;1(1):5–15.

[2] Hameed M Shagul, Sekar ASS, Saraswathy V. Strength and permeabilitycharacteristics study of self-compacting concrete using crusher rock dust andmarble sludge powder. Arab J Sci Eng 2012;37:561–74.

[3] Hemant Sood, Khitoliya RK, Pathak SS. Incorporating European standards fortesting self compacting concrete in Indian conditions. Inter J Rec Tren Eng2009;1(6):41–5.

[4] Nor Atan MD, Hanizam Awang. The compressive and flexural strengths of selfcompacting concrete using raw rice husk ash. J Eng Sci Technol2011;6(6):720–32.

[5] Damtoft JS, Lukasik J, Herfort D, Sorrentino D, Gartner EM. Sustainabledevelopment and climate change initiatives. Cem Concr Res2008;38(2):115–27.

[6] Fortes Revilla Carlos, Ramirez Martinez Sagrario, Blanco Varela Maria Teresa.Modelling of slaked lime–metakaolin mortar engineering characteristics interms of process variables. Cem Concr Compos 2006;28:458–67.

[7] Vejmelkova Eva, Keppert Martin, Grzeszczyk Stefania, Skalinski Bartłomiej,Cerny Robert. Properties of self-compacting concrete mixtures containingmetakaolin and blast furnace slag. Constr Build Mater 2011;25:1325–31.

[8] Jutice JM, Kurtis KE. Influence of metakaolin surface area on properties ofcement based materials. ASCE J Mater Civil Eng 2007;19(9):762–71.

[9] Shekarchi M, Bonakdar A, Bakhshi M, Mirdamadi A, Mobasher B. Transportproperties in metakaolin blended concrete. Constr Build Mater2010;24:2217–23.

[10] Jiping BAI, Stan WILD, Albinas GAILIUS. Accelerating early strengthdevelopment of concrete using metakaolin as an admixture. Mater Sci2004;10(4):338–44.

[11] Ganesan K, Rajagopal K, Thangavelu K. Rice husk ash blended cement:assessment of optimal level of replacement for strength and permeabilityproperties of concrete. Constr Build Mater 2007;22:1675–83.

[12] Tuan Nguyen Van, Guaug ye, Breugel Klaas van, Fraaij Alex LA, Bui Danh Dai.The study of using rice husk ash to produce ultra high performance concrete.Constr Build Mater 2011;25:2030–5.

[13] Safiuddin MD, West JS, Soudki KA. Hardened properties of self-consolidatinghigh performance concrete including rice husk ash. Cem Concr Compos2010;32:708–17.

[14] Abhilash, Shukla, Singh CK, Sharma Arbind Kumar. Study of the properties ofconcrete by partial replacement of ordinary Portland cement by rice husk ash.Int J Earth Sci Eng 2011;4(6):965–8.

[15] Kanellopoulos Antonios, Petrou Michael F, Ioannou Ioannis. Durabilityperformance of self-compacting concrete. Constr Build Mater 2012;37:320–5.

[16] Nadeem Siddiqui A. Influence of fluidity on reliability of SCC produced usinglocal Saudi materials. Arab J Sci Eng 2011;36(2):203–14.

[17] EFNARC. The European guidelines for self-compacting concrete, Europeanproject group; 2005.

[18] Sonebi M. Medium strength self-compacting concrete containing fly ash:modeling using factorial experimental plans. Cem Concr Res2004;34(7):1199–208.

[19] Khayat KH, Bickley J, Lessard M. Performance of self-consolidating concrete forcasting basement and foundation walls. ACI Mater J 2000;97(3):374–80.

[20] Safiuddin MD, Mahmud HB, Jumaat MZ. Efficacy of ASTM saturationtechniques for measuring the water absorption of concrete. Arab J Sci Eng2011;35(5):761–8.

[21] Ramezanianpour AA, Mahdi khani M, Ahmadibeni GH. The effect of rice huskash on mechanical properties and durability of sustainable concretes. Int J CivilEng 2007;7(2):83–91.

[22] Guneyisi E, Gesoglu M. Properties of self-compacting mortars with binary andternary cementitious blends of fly ash and metakaolin. Mater Struct2008;41:1519–31.

[23] Felekoglu B, Turkel S, Baradan B. Effect of water/cement ratio on the fresh andhardened properties of self-compacting concrete. Build Environ2007;42:1795–802.

[24] Feng Q, Yamamichi H, Shoya M, Sugita S. Study on the pozzolanic properties ofrice husk ash by hydrochloric acid pretreatment. Cem Concr Res2004;34(3):521–6.

[25] De Oliveira Luiz Antonio Pereira, de Castro Gomes Joao Paulo, Pereira CristianaNadir Gonilho. Study of sorptivity of self-compacting concrete with mineraladditives. J Civ Eng Manag 2006;7:215–20.

[26] Uysal Mucteba, Yilmaz Kemalettin, Ipek Metin. The effect of mineraladmixtures on mechanical properties, chloride ion permeability andimpermeability of self-compacting. Concr, Constr Build Mater2012;27:263–70.

[27] Zhu W, Peter JMB. Permeation properties of self compacting concrete. CemConcr Res 2003;33:921–6.

[28] Merlini Marco, Artioli Gilberto, Cerulli Tiziano, Cella Fiorenza, Bravo Anna.Tricalcium aluminate hydration in additivated systems. A crystallographicstudy by SR-XRPD. Cem Concr Res 2008;38(4):477–86.

[29] Rawal P. Development of durable concrete from pulverized fly ash andpozzolan derived from agricultural wastes. Master of Engineering. ThesisNumber ST-82-1, Asian Institute Technology, May, 2003.

[30] Siad H, Mesbah HA, Khelafi H, Kamali Bernard S, Mouli M. Influence of naturalpozzolan on the behavior of self-compacting concrete under sulphuric andhydrochloric acid attacks, comparative study. Arab J Sci Eng2010;35(1):183–95.

[31] Ramezanianpour AA, Bahrami Jovein H. Influence of metakaolin assupplementary cementing material on strength and durability of concretes.Constr Build Mater 2012;30:470–9.

[32] Khan Rawaid, Jabbar Abdul, Ahmad Irshad, Khan Wajid, Khan Akhtar Naeem,Mirza Jahangir. Reduction in environmental problems using rice-husk ash inconcrete. Constr Build Mater 2012;30:360–5.

[33] Detwiler RJ, Mehta PK. Chemical and physical effects of silica fume on themechanical behavior of concrete. ACI Mater J 1989;86(6):609–14.

[34] Nguyen VT, Ye G, Van Breugel K, Guo Z, Bui DD. Hydration process of rice huskash and silica fume in cement paste by means of isothermal calorimetry. In:the sixth international conference on concrete under severe conditionsenvironment and loading (CONSEC’10), Yucatan, Mexico; June 2010. p.1521–28.

[35] Antoni M, Rossen J, Martirena F, Scrivener K. Cement substitution by acombination of metakaolin and limestone. Cement Concr Res2012;42:1579–89.