Strength, durability, and micro-structural properties of concrete made with used-foundry sand (UFS)

Construction and Building Materials 25 (2011) 1916–1925

Contents lists available at ScienceDirect

Construction and Building Materials

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

Strength, durability, and micro-structural properties of concrete made withused-foundry sand (UFS)

Rafat Siddique a,⇑, Yogesh Aggarwal b, Paratibha Aggarwal b, El-Hadj Kadri c, Rachid Bennacer d

a Civil Engineering Department, Thapar University, Patiala 147004, Indiab Civil Engineering Department, National Institute of Technology, Kurukshetra, Indiac Department of Civil Engineering, University of Cergy Pontoise, Neuville-sur-Oise, 95031 Cergy-Pontoise, Franced Civil Engineering, LMT-Ecole Normale Supérieure – Cachan, 61 av. du président Wilson, F-94235 Cachan Cedex, France

a r t i c l e i n f o

Article history:Received 21 August 2010Received in revised form 21 October 2010Accepted 13 November 2010Available online 16 December 2010

Keywords:ConcreteFoundry sandStrength propertiesDurability propertiesMicrostructure

0950-0618/$ - see front matter � 2010 Elsevier Ltd. Adoi:10.1016/j.conbuildmat.2010.11.065

⇑ Corresponding author. Tel.: +91 175 239 3207; faE-mail address: [email protected] (R. Siddiq

a b s t r a c t

This paper presents the design of concrete mixes made with used-foundry (UFS) sand as partial replace-ment of fine aggregates. Various mechanical properties are evaluated (compressive strength, and split-tensile strength). Durability of the concrete regarding resistance to chloride penetration, and carbonationis also evaluated. Test results indicate that industrial by-products can produce concrete with sufficientstrength and durability to replace normal concrete. Compressive strength, and split-tensile strength,was determined at 28, 90 and 365 days along with carbonation and rapid chloride penetration resistanceat 90 and 365 days. Comparative strength development of foundry sand mixes in relation to the controlmix i.e. mix without foundry sand was observed. The maximum carbonation depth in natural environ-ment, for mixes containing foundry sand never exceeded 2.5 mm at 90 days and 5 mm at 365 days.The RCPT values, as per ASTM C 1202-97, were less than 750 coulombs at 90 days and 500 coulombsat 365 days which comes under very low category. Thereby, indicating effective use of foundry sand asan alternate material, as partial replacement of fine aggregates in concrete. Micro-structural investiga-tions of control mix and mixes with various percentages of foundry sand were also performed usingXRD and SEM techniques. The micro-structural investigations shed some light on the nature of variationin strength at the different replacements of fine aggregates with foundry sand, in concrete.

� 2010 Elsevier Ltd. All rights reserved.

1. Introduction

Concrete is the most widely used man-made product in theworld, and is second only to water as the world’s most utilized sub-stance. Slightly more than a ton of concrete is produced each yearfor every human being on the planet, some six billion tons a year.Concrete is an affordable and reliable material that is appliedthroughout the infrastructure of a nation’s construction, industrial,transportation, defense, utility, and residential sectors.

Fundamentally, concrete is economical, strong, and durable.Although concrete technology across the industry continues to riseto the demands of a changing marketplace, the industry recognizesthat considerable improvements are essential in productivity, productperformance, energy efficiency, and environmental performance. Theindustry will need to face and overcome a number of institutional,competitive, and technical challenges. One of the major challenges,with the environmental awareness and scarcity of space for landfill-ing, is the wastes/byproduct utilization as an alternative to disposal.

ll rights reserved.

x: +91 175 236 4498.ue).

Throughout the industrial sector, including the concrete industry,the cost of environmental compliance is high. Introduction of use ofindustrial by-products such as foundry sand, fly ash, bottom ash,and slag can result in significant improvements in overall industry en-ergy efficiency and environmental performance.

Foundry sand is high quality silica sand with uniform physicalcharacteristics. It is a by-product of ferrous and non-ferrous metalcasting industries, where sand has been used for centuries as amolding material because of its thermal conductivity. Foundriessuccessfully recycle and reuse the sand many times in a foundry.When the sand can no longer be reused in the foundry, it isremoved from the foundry and is termed as foundry sand. Used-foundry sand can be reused in various applications as an alterna-tive to sending it to landfill, and reuse options are well establishedin England, Europe and North America. Reuse options includecement manufacture, asphalt, concrete, bricks and free-flow fillfor certain construction applications. Some of these alternativesare starting to be adopted in India, but is still in early stage. Over-seas examples show that it is not only better for the environmentbut is profitable for the foundry to use the sand alternatively. Thesefoundries have significantly reduced the volume of waste sand

TabCh

TabMi

R. Siddique et al. / Construction and Building Materials 25 (2011) 1916–1925 1917

going to landfill and actually offset the total cost of transportingthe sand ‘in’ and ‘out’.

The objectives of this study are to investigate the effect of use offoundry sand as partial replacement of fine aggregates in variouspercentages (0–60%), on concrete properties such as mechanicaland durability characteristics of the concrete along with micro-structural analysis with XRD and SEM. Application of used-foundrysand in concrete will lead to diversion of large amounts of used-foundry sand from land filling to manufacturing of concrete.

3. Experimental methods

The effect of using foundry sand as partial replacement of fine aggregates in var-ious percentages, on concrete was investigated. Also, the effect of incorporatingfoundry sand in concrete on the mechanical, durability properties and microstruc-ture were evaluated.

3.1. Materials and mix proportions

Portland Pozzolana Cement (53 MPa) conforming to Indian standard specifica-tions IS: 1489-1991 was used. Consistency was 27%, specific gravity was 3.56 andfineness as per specific surface of cement was 354 m2/kg. Locally available naturalsand with 4.75 mm maximum size was used as fine aggregate. It fulfilled therequirements of ASTM C 33-02a and crushed stone with 20 mm maximum sizewas used as coarse aggregate. The properties of fine aggregates and coarse aggre-gates were found to conform to IS: 383-1970 with specific gravity of sand as 2.63and coarse aggregate as 2.77. Unit weight of sand and coarse aggregate was 1890and 1650 kg/m3, respectively. Fineness modulus was observed to be 3.03 for sandand 6.74 for coarse aggregates. Locally available foundry sand was used as partialreplacement of fine aggregates (regular sand). Foundry sand with specific gravity

le 1emical properties of foundry sand.

Constituents % by weight (used inpresent study)

Requirements as per Americanfoundry men’s society, 1991

Loss onignition

2.15 5.15% (max)

Silica (SiO2) 78.81 87.9%Iron oxide

(Fe2O3)4.83 0.94%

Alumina(Al2O3)

6.32 4.70%

Calcium oxide(CaO)

1.88 0.14% (min)

Magnesiumoxide(MgO)

1.95 0.3%

Sulphate 0.05 0.09%Chloride 0.04 –

le 2x proportions of concrete mixes containing UFS.

Mix no. CM F10

Cement (kg/m3) 350 350Foundry sand (%) 0 10Foundry sand (kg/m3) 0 60.5Water (kg/m3) 175 175W/C 0.5 0.5Sand SSD (kg/m3) 605 544.5Coarse aggregate(kg/m3) 1260 1260Superplasticizer(kg/m3) 1.75 1.75Slump (mm) 30 30Compaction factor 0.83 0.88Vee-bee consistometer (sec) 5.98 4.40Air temperature (�C) 23 26Concrete temperature (�C) 25 29Fresh concrete density (kg/m3) 2437.7 2414.0pH value

90-days 11.70 11.73365-days 11.80 11.75

F20

35020121.01750.5484.012601.75300.844.9827272419.2

11.7511.91

2.61, unit weight 1638 kg/m3 and fineness modulus 1.78 was used. The foundrysand showed lower fineness modulus and bulk density than the regular sand. Asper the particle size distribution of the foundry sand, the size corresponding to50% of passing (d50) was around 33 lm and average diameter of foundry sand par-ticle was observed to be 28.8 lm. Table 1 gives the chemical composition of foun-dry sand. A polycarboxylic ether based superplasticizer of CICO brand complyingwith ASTM C-494 type F, IS: 9103 – 1999 and IS: 2645-2003 was used.

Seven mix proportions were prepared. First was control mix (without foundrysand), and the other six mixes contained foundry sand. Fine aggregate (sand) wasreplaced with foundry sand by weight. The proportions of fine aggregate replacedranged from 10% to 60% at the increment of 10%. Mix proportions are as given inTable 2. The control mix without foundry sand was proportioned as per Indian stan-dard specifications IS: 10262-1982, to obtain a 28-day cube compressive strength of36 MPa. Hand mixing was done for the all the concrete mixes.

3.2. Testing procedure

Fresh concrete properties such as slump flow, compaction factor, vee-bee con-sistometer were determined according to an Indian Standard specification IS: 1199-1959. The results are presented in Table 2. The 150 mm concrete cubes and150 � 300 mm cylinders were cast for compressive strength, and 150 � 300 mmcylinders for split-tensile strength. After required period of curing, the specimenswere taken out of the curing tank and their surfaces were wiped off. The varioustests performed were compressive strength test of cubes (150 mm side), cylinders(150 mm � 300 mm), and split-tensile strength of cylinders (150 mm � 300 mm)at 28, 90, and 365 days, as per IS: 516-1959.

The cylinders (100 mm � 200 mm) were cast for rapid chloride penetrationresistance test and were sliced 2-in. (51-mm) thick of 4-in. (102-mm) nominaldiameter. Rapid chloride penetration resistance test (according to ASTM C 1202-97) covered the determination of the electrical conductance of concrete to providea rapid indication of its resistance to the penetration of chloride ions. The test meth-od consisted of monitoring the amount of electrical current passed through 2-in.(51-mm) thick slices of 4-in. (102-mm) nominal diameter cores or cylinders for a6-h period. A potential difference of 60 V dc was maintained across the ends ofthe specimen, one of which was immersed in a sodium chloride solution, the otherin a sodium hydroxide solution. The total charge passed, in coulombs, was relatedto the resistance of the specimen to chloride ion penetration.

The cylinders (150 mm � 300 mm) were cast for carbonation test. Carbonationtest of depth of color less region using phenolphthalein indicator was determinedusing the cylinders (150 mm � 300 mm) as per RILEM CPC-18 [11]. After casting,test specimens were covered with plastic sheets and left in casting room for 24 hat room temperature and demolded and cured in water for 28 days. After that, spec-imens were air cured (open environment) for the required age say till 90-days or365-days, and then were split. The freshly split surface was cleaned and sprayedwith a phenolphthalein pH indicator. The indicator was a phenolphthalein 1% eth-anol solution (1 g phenolphthalein and 90 ml ethanol (95.0 V/V%) diluted in waterto 100 ml. The average depth ‘Xp’ of the colorless phenolphthalein region was mea-sured from three points, perpendicular to the two edges of the split face, immedi-ately after spraying the indicator.

X-ray diffraction analysis (XRD) was done on Philips PW 1140/09. Diffractome-ter operated at 35 kV, using Cu ka radiation and Ni filler. The samples for X-ray dif-fraction analysis were prepared in powdered form. The concrete sample was takenfrom the inner core of the matrix. X-ray diffraction is a non-destructive techniqueused to determine the elements present in any particular substance. X-ray diffrac-tion is based on the fact that, in a mixture, the measured intensity of a diffractionpeak is directly proportional to the content of the substance producing it (Soroka

F30 F40 F50 F60

350 350 350 35030 40 50 60181.5 242.0 302.5 363.0175 179.24 185.6 196.20.5 0.512 0.53 0.56423.5 363.0 302.5 242.01260 1260 1260 12601.75 1.75 1.75 1.7530 30 30 200.85 0.81 0.81 0.784.97 5.65 5.34 6.2520 22 22 3423 24 24 282426.8 2420.1 2416.1 2402.2

11.75 11.72 11.72 11.9011.57 11.60 11.40 11.42

1918 R. Siddique et al. / Construction and Building Materials 25 (2011) 1916–1925

[12]). Since 2d is a known constant, the 2h setting of each peak corresponds to a cer-tain wave length.

The scanning electron microscopic studies of various concrete samples and con-stituent materials were carried out using Philips XL20 Scanning Electron Micro-scope. The concrete specimens were first cured in water for 365 days and thenoven dried at 105 �C for 24 h.

4. Results and discussions

4.1. Fresh concrete properties

The effect of using foundry sand as partial replacement of fineaggregates in various percentages, on concrete was investigated.Also, the effect of incorporating foundry sand in concrete on themechanical, durability properties and microstructure wereevaluated.

3.1. Materials and mix proportionsPortland Pozzolana Cement (53 MPa) conforming to Indian

standard specifications IS: 1489-1991 was used. Consistency was27%, specific gravity was 3.56 and fineness as per specific surfaceof cement was 354 m2/kg. Locally available natural sand with4.75 mm maximum size was used as fine aggregate. It fulfilledthe requirements of ASTM C 33-02a and crushed stone with20 mm maximum size was used as coarse aggregate. The proper-ties of fine aggregates and coarse aggregates were found to con-form to IS: 383-1970 with specific gravity of sand as 2.63 andcoarse aggregate as 2.77. Unit weight of sand and coarse aggregatewas 1890 and 1650 kg/m3, respectively. Fineness modulus was ob-served to be 3.03 for sand and 6.74 for coarse aggregates. Locallyavailable foundry sand was used as partial replacement of fineaggregates (regular sand). Foundry sand with specific gravity2.61, unit weight 1638 kg/m3 and fineness modulus 1.78 was used.The foundry sand showed lower fineness modulus and bulk den-sity than the regular sand. As per the particle size distribution ofthe foundry sand, the size corresponding to 50% of passing (d50)was around 33 lm and average diameter of foundry sand particlewas observed to be 28.8 lm. Table 1 gives the chemical composi-tion of foundry sand. A polycarboxylic ether based superplasticizerof CICO brand complying with ASTM C-494 type F, IS: 9103 – 1999and IS: 2645-2003 was used.

Seven mix proportions were prepared. First was control mix(without foundry sand), and the other six mixes contained foundrysand. Fine aggregate (sand) was replaced with foundry sand byweight. The proportions of fine aggregate replaced ranged from10% to 60% at the increment of 10%. Mix proportions are as givenin Table 2. The control mix without foundry sand was proportionedas per Indian standard specifications IS: 10262-1982, to obtain a28-day cube compressive strength of 36 MPa. Hand mixing wasdone for the all the concrete mixes.

3.2. Testing procedureFresh concrete properties such as slump flow, compaction fac-

tor, vee-bee consistometer were determined according to an IndianStandard specification IS: 1199-1959. The results are presented inTable 2. The 150 mm concrete cubes and 150 � 300 mm cylinderswere cast for compressive strength, and 150 � 300 mm cylindersfor split-tensile strength. After required period of curing, the spec-imens were taken out of the curing tank and their surfaces werewiped off. The various tests performed were compressive strengthtest of cubes (150 mm side), cylinders (150 mm � 300 mm), andsplit-tensile strength of cylinders (150 mm � 300 mm) at 28, 90,and 365 days, as per IS: 516-1959.

The cylinders (100 mm � 200 mm) were cast for rapid chloridepenetration resistance test and were sliced 2-in. (51-mm) thick of4-in. (102-mm) nominal diameter. Rapid chloride penetration

resistance test (according to ASTM C 1202-97) covered the deter-mination of the electrical conductance of concrete to provide a ra-pid indication of its resistance to the penetration of chloride ions.The test method consisted of monitoring the amount of electricalcurrent passed through 2-in. (51-mm) thick slices of 4-in. (102-mm) nominal diameter cores or cylinders for a 6-h period. A poten-tial difference of 60 V dc was maintained across the ends of thespecimen, one of which was immersed in a sodium chloride solu-tion, the other in a sodium hydroxide solution. The total chargepassed, in coulombs, was related to the resistance of the specimento chloride ion penetration.

The cylinders (150 mm � 300 mm) were cast for carbonationtest. Carbonation test of depth of color less region using phenol-phthalein indicator was determined using the cylinders (150 mm �300 mm) as per RILEM CPC-18 [11]. After casting, test specimenswere covered with plastic sheets and left in casting room for 24 hat room temperature and demolded and cured in water for 28 days.After that, specimens were air cured (open environment) for therequired age say till 90-days or 365-days, and then were split. Thefreshly split surface was cleaned and sprayed with a phenolphthaleinpH indicator. The indicator was a phenolphthalein 1% ethanolsolution (1 g phenolphthalein and 90 ml ethanol (95.0 V/V%) dilutedin water to 100 ml. The average depth ‘Xp’ of the colorless phenol-phthalein region was measured from three points, perpendicular tothe two edges of the split face, immediately after spraying theindicator.

X-ray diffraction analysis (XRD) was done on Philips PW 1140/09. Diffractometer operated at 35 kV, using Cu ka radiation and Nifiller. The samples for X-ray diffraction analysis were prepared inpowdered form. The concrete sample was taken from the innercore of the matrix. X-ray diffraction is a non-destructive techniqueused to determine the elements present in any particular sub-stance. X-ray diffraction is based on the fact that, in a mixture,the measured intensity of a diffraction peak is directly proportionalto the content of the substance producing it (Soroka [12]). Since 2dis a known constant, the 2h setting of each peak corresponds to acertain wave length.

The scanning electron microscopic studies of various concretesamples and constituent materials were carried out using PhilipsXL20 Scanning Electron Microscope. The concrete specimens werefirst cured in water for 365 days and then oven dried at 105 �C for 24 h.

4. Results and discussions

4.1. Fresh concrete propertiesThe workability of fresh concrete is a composite property which

includes the diverse requirements of stability, mobility, compacti-bility, placeability, and finishability. Slump is a measure indicatingthe consistency or workability of concrete. Slump for control mixCM was 30 mm and for the F mixes it was observed to be 30 to40 mm. The compaction factor values for control mix, and F mixescorresponded to the slump flow values as per Table 2. The presenceof finer foundry sand particles in concrete lead to the increase inthe water demand, as compared to the regular sand particles. Thus,to maintain the workability within specified range, the water con-tent was constant till F30 and thereafter increased. The values ofvee-bee time for control mix CM, and F mixes corresponds to theslump flow values and compaction factor values as per ACI Com-mittee 211 [13].

4.2. Mechanical properties4.2.1. Compressive strength. Cube compressive strength results ofmixes made with various percentages of foundry sand i.e., CM(0% FS), F10 (10% FS), F20 (20% FS), F30 (30% FS) F40 (40% FS),F50 (50% FS), F60 (60% FS), at ages of 28, 90 and 365 days areshown in Table 3. There is marginal decrease in the compressive

R. Siddique et al. / Construction and Building Materials 25 (2011) 1916–1925 1919

strength of concrete mixes with the inclusion of foundry sand asreplacement of regular sand. The maximum strength was obtainedat 30% foundry sand in all the replaced mixes and was more thancontrol mix at 28 days, 90 days and 365 days. The strength varia-tion for the various percentages of replacements was observed tobe linearly increasing from 10% to 50% with different behaviourof mix F30, attaining a comparatively high strength almost closeror more than the strength of the CM mix. The strength of F60mix reduced drastically.

The strengths of F30, F40, and F50 mixes were observed to bemore than the strength of CM mix at the ages of 28 days, 90 daysand 365 days. Also, the strength of F20 mix was also observed tobe more than the strength of CM mix at 365 days. Lam et al. [14]reported gain of strength between 7 days and 90 days for normalconcrete (slump about 75 mm) as 72% which is characteristic ofnormal concrete. At 90 days, the rate of gain was closer to that ofcontrol mix and at 365 days, the rate of gain for all the F mixeswas higher than the CM mix.

As in cube compressive strength, the cylinder compressivestrength as indicated in Table 2 of mix F30 was highest amongall F mixes, at all ages of 28, 90 and 365-days and even higher thancontrol mix CM at 365-days. Also, linear increase of 28-day, 90-day, and 365-day strength was observed from F10 to F50 exceptfor mix F30. F mixes showed the 28-day cylinder compressivestrengths between 0.64 to 0.70 times; the 90-day cylinder com-pressive strengths between 0.65 to 0.76 times and the 365-daycompressive strengths between 0.67 to 0.74 times the correspond-ing cube compressive strength of concrete.

The increase in compressive strength (cube and cylinder) withthe inclusion of foundry sand could probably be due to the fact thatfoundry sand was finer than regular sand which resulted in thedenser concrete matrix. Inclusion of foundry sand has not ad-versely affected the 28-day compressive strengths of concretemixes made with foundry sand, but has shown a linear increase be-tween mixes F10 and F50 depending upon the foundry sandcontent.

4.2.2. Split-tensile strength. The results for split-tensile strength areshown in Table 4. The variation in the split-tensile strength withfoundry sand content was similar to that observed in the case ofthe compressive strength. When compared to control mix CM, itwas observed that F30, F40, and F50 mixes showed higher strengthat 28 days, 90 days, and 365-days. The optimum replacement ofsand with foundry sand can no doubt be assumed as F30, as ithas attained highest strength among all mixes at all ages. A de-crease in strength for mix F60 is observed, after a linear increaseof split-tensile strength from F10 to F50, except F30 mix at all ages.28-day split-tensile strength of showed decrease of 11.5%, 4.8% and17.30% for F10, F20 and F60 mixes and increase of 24.03%, 19.23%,and 14.42% for F30, F40 and F50 mixes, in comparison with thestrength of the control mix CM. At 90 days, a decrease of 11.65%,4.51%, and 18.79% for mixes F10, F20, and F60 and increase of25.18%, 22.55%, and 19.92% for F30, F40, and F50 was observed

Table 3Cube and cylinder compressive strength of concrete mixes.

Mix 28-day (MPa) 90-day (MPa

Cube Cylinder Cube

CM 36.27 26.35 43.91F10 31.05 21.87 37.25F20 32.52 22.68 40.08F30 38.03 24.94 46.59F40 36.42 23.58 44.23F50 37.14 24.17 45.18F60 29.86 20.73 33.13

when compared to control mix CM. Split-tensile strength wasfound to increase with age.

For F mixes, the ratio of split-tensile strength to cube compres-sive strength of concrete varied from 6% to 7% with 5% for F60 mix.It was reported that for normal strength mixes, the ratio of split-tensile strength to cube compressive strength of concrete is inthe range of 7–9% (Mehta & Monterio, [15]).

Compressive strength is assumed as an adequate index for alltypes of strength, and therefore a direct relationship ought to existbetween the compressive and tensile strength of a given concrete.It has been observed that relationship among various types ofstrength is influenced by factors like the methods by which thetensile strength is measured (i.e., direct tension test, splitting test,or flexure test), the quality of concrete (i.e., low-, moderate- orhigh-strength), the aggregate characteristics (e.g., surface textureand mineralogy), and admixtures (e.g., air-entraining and mineraladmixtures). In the present study, the relationship betweencompressive strength and split-tensile strength were found tocorrespond to that of normal concrete i.e. the ratio of tensile-to-compressive strength ratio was observed to be lying between 5%and 7% for F mixes, due to the inclusion of foundry sand in F mixes,affecting the type of fine aggregate in the concrete mix.

5. Durability

5.1. CarbonationThe test results on the carbonation depth of concrete specimens

measured until 12 months are represented by the values of corre-sponding carbonation coefficient (C). C is inversely proportional tothe carbonation resistance of concrete and was estimated using theempirical relationship,

X ¼ CðTÞ0:5 ð1Þ

where X is the tested carbonation depth (mm), T the period of expo-sure (month), C is the corresponding carbonation coefficient (mm/month0.5).

This formula is based on the square-root-t-law, which is gener-ally used to compare the carbonation resistance of concrete andhas been adopted by numerous researchers (Sulapha et al. [16];Wee et al. [17]; Castroa et al. [18]). The results of carbonationdepth in natural environment are expressed in Fig. 1, at 90 and365 days.

It can be seen that the carbonation depth increases with an in-crease in the age. Similar results have been reported for the controlmix that carbonation increases with age (Corinaldesi and Moriconi[19]). From Fig. 1, it is evident that foundry sand incorporation ofits own demonstrated increase in carbonation depth with increasein percentage of foundry sand in F mixes. For every increase in 10%foundry sand i.e. mix F10 to F60, there is at an average 0.17 mm in-crease in carbonation depth in F mixes at 90 days. Similarly, at365 days at an average 0.33 mm increase in carbonation depthwas observed in F mixes. The maximum carbonation depth ob-served for F mixes was for F60 mix (containing 60% foundry sand)

) 365-day (MPa)

Cylinder Cube Cylinder

30.50 44.42 31.9426.17 43.09 30.7628.00 47.08 33.1830.44 54.15 36.4029.57 50.12 35.5029.89 51.71 36.1025.28 36.19 27.09

Table 4Split tensile strength of concrete mixes.

Mix 28-day (MPa) 90-day (MPa) 365-day (MPa)

CM 2.08 2.66 2.97F10 1.84 2.35 2.66F20 1.98 2.54 2.86F30 2.58 3.33 3.50F40 2.48 3.26 3.42F50 2.38 3.19 3.33F60 1.72 2.16 2.52

1920 R. Siddique et al. / Construction and Building Materials 25 (2011) 1916–1925

was about 2.17 mm at 90 days and 5 mm at 365 days, which is farless than the cover of reinforcing steel bars to cause corrosion.

The typical values for coefficient C never exceeded 1.5, so as perTable 5, the concrete can be adjudged as good concrete (Castroaet al. [18]). Thus, it was observed that the carbonation depth ofmaximum 5 mm observed at 365-day age, in the present studywas very less as compared to carbonation depth observed from lit-eratures. Also, pH value for various mixes ranged between 11.72and 11.90 at 90 days and 11.40 and 11.80 at 365 days.

5.2. Rapid chloride penetration resistanceThe ability of concrete to resist the penetration of chloride ions

is a critical parameter in determining the service life of steel-rein-forced concrete structures exposed to deicing salts or marine envi-ronments. The effect of fly ash on the mass transfer properties ofconcrete has been well documented; however, no documentationof foundry sand as replacement of fine aggregates in concretemixes is available. The measurement concerns the chloride ionsthat come into concrete and also those flowing through the sam-ples. The RCPT value of F mixes containing foundry sand at theage of 90 and 365 days is shown in Fig. 2. It can be seen that theRCPT value decreases with increase in age. At 90 days, the RCPT va-lue for F mixes was found to more than that of CM mix except forF20 and F30 mixes. The maximum value was observed for F60 mix.It is observed that cement type, w/c ratio, curing condition, andtesting age have effect on chloride permeability of concrete. Thenormal concretes or the concrete with various additives could varyin above parameters thereby effecting RCPT values. The F mixes inthe present study showed very less RCPT value thereby indicatinggood permeability on addition of foundry sand in concrete.

5.3. X-ray diffraction (XRD) studies of mixesThe X-ray diffraction pattern and analysis of the concrete mixes

i.e. control mix, and F mixes at 365 days are shown in Fig. 3a–g. Inall the mixes, C2S, C3S, and C4AF peaks are not visible indicatingthat they are totally consumed. Also, the consumption of lime isindicated due to lowering of pH from 13 to between 10.9 and 12.As shown in Fig. 3a–g, SiO2 peak indicating free silica, in CM mixwas observed at 1800. For the F10 mix almost same peak was ob-tained as the addition of foundry sand was nominal. For F20 mix,

0

1

2

3

4

5

0 10 20 30 40 50 60Foundry Sand (%)

Car

bona

tion

Dep

th (m

m)

90 Days

365 Days

Fig. 1. Variation of carbonation depth for CM & F mixes at different ages.

increase in peak was observed to 3700. It was observed that forF30 mix the peak was minimum at 1300 in all mixes indicatingthe maximum utilization of silica in C–S–H gel. The F40 mixshowed the peak increase to 5000. Again a decrease in intensitypeak was observed for F50 mix at 1600 and finally for F60 mix itwas observed to increase to 3800.

5.4. SEM analysisThe type, amount, size, shape, and distribution of phases pres-

ent in a solid constitute its microstructure. It is the application oftransmission and scanning electron microscopy techniques whichhas made it possible to resolve the microstructure of the materialsto a fraction of 1 lm. Although, concrete is the most widely usedstructural material, its microstructure is heterogeneous and highlycomplex. Also, the microstructure–property relationships in con-crete are not fully developed. Original microstructure and mor-phology of the hydrate mixes were observed on fracturedsurfaces. Fractured small samples were mounted on the SEM stubswith gold coating.

It is well known that, the calcium–silica–hydrate (C–S–H) ismajor phase present. The factors that influence the mechanicalbehaviour of C–S–H phases are: size and shape of the particles, dis-tribution of particles, particle concentration, particle orientation,topology of the mix, composition of the dispersed/continuousphases and the pore structure. Considering various scanning elec-tron microscope images, assume that the bright and dark matterin the images stands for C–S–H gel/paste and inert aggregates,and the medium dark particles are for foundry sand particles.The assumptions regarding presence of particles is based on thefacts that these medium dark particles are seen in almost everysample except the control mix CM (every sample except the con-trol mix CM contains foundry sand). These assumptions can be jus-tified based on the fact that the basic structure of the concrete in allthe samples is the same i.e. the mix designed for the control mixhas been kept constant in all the samples changing only the foun-dry sand percentages in these mixes.

Fig. 4 is micrograph of control mix i.e. the SEM image at 1.5 KXmagnifications. It shows the formation of proper C–S–H gel in var-ious stages. The gel formation is clearly visible in the micrograph.The encircled portions represent the voids while rest of the pictureconsists of C–S–H gel and inert aggregates (both fine and coarse).In the micrograph C–S–H gel i.e. the bright masses with nodulesand big chalky gel parts are spread over the entire micrograph.Also, it is evident from various literatures, that the C–S–H gel getsspread over the aggregates thus acting as binder for the paste.

Fig. 5a, micrograph of F10 mix shows two major features.Firstly, the number of voids in the mix has significantly reduced(the big voids as seen in the control mix are not visible) and sec-ondly, the C–S–H gel paste is not as widely spread as it was inthe control mix showing some aversion to the binder paste butmore importantly the effect of foundry sand has been negativeon the strength because of lesser quantity of foundry sand thanthe optimum amount of the foundry sand required, this is clearlyevident from the mechanical properties as all the strengths ofthe mix has deteriorated significantly. The microstructure alsoshows the presence of foundry sand particles of various sizes atvarious places. The decrease in strength could be attributed tothe non formation of proper C–S–H gel as compared to CM mixmicrostructure. Although, at few places the formation of C–S–Hgel could be detected as the percentage of the foundry sand addedwas only 10%. Fig. 5b, micrograph of F10 mix at higher (2.50 KX)magnification clearly indicates that the increased fluidity in thepaste has tremendously reduced its capacity for attaining higherstrength because lesser amount of paste is available than requiredin addition to the foundry sand.

Table 5Typical value of C for Poor, average and good concreteCastroa et al., 2000.

C (mm/yr 0.5) Concrete

>9 Poor9 > C > 6 Average<6 Good

200

300

400

500

600

700

800

0 10 20 30 40 50 60Foundry Sand (%)

Cha

rge

Pass

ed (C

oulo

mbs

)

90 Days 365 Days

Fig. 2. Variation of charge passed for CM & F Mixes at different ages.

R. Siddique et al. / Construction and Building Materials 25 (2011) 1916–1925 1921

Fig. 6, micrograph of F20 mix shows the presence of more foun-dry sand particles at various places since the percentage of thefoundry sand increased to 20%. Also, formation of C–S–H gel at veryfew places can be observed in the microstructure. The C–S–H gelshows better spread than in the previous mix and formation ofnodules is also higher than in the previous mix. The combined ef-fect of both can be correlated to the strength values obtained forthis mix, though the strength has been higher than the previousmix but it is still less than the control mix, this evidently states thatthe amount of the foundry sand is still less than that required forthe mix to be optimum.

Fig. 7 shows three observations different from the previousmixes. Firstly and the foremost is the significant formation of ten-drils (the pointed thin strands), secondly the paste has spread veryfinely and firmly throughout the sample, and lastly the effect ofcombined paste and foundry sand has led to a formation of higherstrength concrete.

The formation of tendrils has been attributed to various factors,but in literature (Lea’s, [20]) the most prominent of these being thequick-set, that is the mix may have subjected itself to some chem-ical as well as physical changes to modify the profoundness as wellas initial setting of the mix, i.e. the mix due to its reactivity has gi-ven a quickset. Secondly, if these tendrils are neglected, then theC–S–H paste behind shows some distinct characteristics that werenon-existent in earlier mixes, the C–S–H gel is more finely spreadthan before, this spread plus the extra amount of availability of thegel has resulted in higher strength, even higher than the controlmix. This mix has shown a tremendous response to the foundrysand and highest strength among all the mixes. Due to the C–S–H gel formation at its best the mix became denser which caused in-crease in the strength, thereby attributing to the maximumstrength of F30 mixes in all F mixes. There is indication of prefectreaction of all the components in this mix. This finding is in accor-dance with the mechanical tests results obtained for strength.

Fig. 8 is micrograph of F40 mix shows the presence of foundrysand particles at various places in mixed state and as separate par-ticles, but prominently visible. It is observed that mix is not pene-trated by foundry sand at many places. The gel formation could beseen at some places in the micrograph. Firstly, the foundry sand inthe sample has spread out rather than confirming to the mix, this

has led to an abrupt disrupting of the paste. The paste in this casehas shown nodules formation, because of the more amount offoundry sand present than is needed for the equilibrium to beset. The excess amount of foundry sand has come out of the gelleading to a disrupture in the paste, which finally comes down inthe form of lower strength of the sample, when compared to pre-vious mix.

Fig. 9, a micrograph of F50 mix shows the formation of good C–S–H gel but slightly less than that observed in F30 mix. Also, thestrength observed in F50 mix is slightly less than that observedin F30 mix. The mix also shows a dense matrix with no pores orcracks. The gel formation is clearly visible at all places in the micro-graph. Mix with 50% replacement of the foundry sand, has furtherattempted on itself to regain the equilibrium again. In this mix, theC–S–H gel further has reconciled itself leading to some increase instrength and betterment of the paste. This further increase instrength and high equilibrium can be attributed to equilibrium for-mation (both physical and chemical) have increased the limit to ahigher value, thus increasing the strength and increasing the qual-ity of the paste as is clear from the micrograph.

Fig. 10 is micrograph of F60 mix showing the presence of biggerand more foundry sand particles at various places. The decrease instrength could be attributed to the development of micro cracks atdifferent places due to coming out of foundry sand particles as isclearly visible from the micrograph. The proper C–S–H gel forma-tion is not visible in the micrograph. The microstructure of F60mix is also porous and weak and the equilibrium of the systemhas finally collapsed leading to a failure, with the lowest strengthand visible micro cracking. The marked areas with the polygonsin the micrograph represent microcracks. Thus, more than 50% ofsand replacement by foundry sand leads to significant reductionin strength. The microstructure of control mix, and F mixes hasshown formation of C–S–H gel at various places and some microcracks were also observed in 60% replacement mixes.

EDX analysis indicated that various mixes showed Ca/Siratios as 1.88 (CM); 2.47(F10); 2.15(F20); 1.87(F30); 1.97(F40);1.89(F50); 10.0(F60). Low Ca/Si ratio (1.88) C–S–H gel was formedin CM mix. The reactivity of the foundry sand was more since highcontact areas allow more reactivity between the activated CaO andSiO2 to produce C–S–H. Low Ca/Si ratio C–S–H gel phase in the CMmix and almost equal Ca/Si ratio in F30 mix show higher reactivityand thus attaining higher strength as compared to other F mixes. Itcould be observed that the variation of Ca/Si C–S–H was in correla-tion with mechanical properties of different F mixes.

In fact, in the present study the mixes with amount of replace-ment of sand more than 50% with foundry sand, also lead to crum-bling at the time of curing. These results simply imply that morethan 50% replacement of sand by foundry sand leads to flaws inconcrete, but the best mixture in any case is inarguably the 30%replacement mix. Further, F30 mix showed large formation ofC–S–H gel thus, development of dense microstructure. The fibrousC–S–H formation acts as a thick impermeable membrane for theingress of chloride ions into concrete. This makes the concretemore resistant to aggressive environment as observed from RCPTvalues also.

6. Conclusion

The following conclusions could be arrived at from the study:

1. The fresh properties for all the mixes were observed to be com-parative with the control mix. The replacement of fine aggre-gate with foundry sand was found to be optimum at 30% andshould not exceed 50%. The rate of gain was closer to that ofcontrol mix at 90 days and at 365 days the rate of gain for allthe mixes with foundry sand was higher than the CM mix.

(a)

(b)

(c)

(d)

Fig. 3. X-ray diffraction patterns of various F mixes.

1922 R. Siddique et al. / Construction and Building Materials 25 (2011) 1916–1925

(e)

(f)

(g)

Fig. 3 (continued)

Fig. 4. Micrograph of CM (1.50 KX).

R. Siddique et al. / Construction and Building Materials 25 (2011) 1916–1925 1923

2. The compressive strengths of cubes and cylinders, and split-tensile strengths were observed to increase with age. Thevarious strengths were found to be marginally lower than thecontrol mix at 28 days, which increased with age and at365 days, was either higher or equal to control mix, thusenabling the use of foundry sand as construction material forstructures and also indicating that the strength differencebetween foundry sand concrete specimens and control concretespecimens became less distinct after 28 days.

3. The concrete with foundry sand F mixes showed goodresistance to carbonation and rapid chloride penetrationresistance as per ASTM C 1202-97 was observed under thecategory of very low. Thus, these by-products can be easily uti-lized in concrete, enhancing the durability properties ofconcrete.

Fig. 5a. Micrograph of F10 mix (1.50 KX).

Fig. 5b. Micrograph of F10 mix (2.50 KX).

Fig. 6. Micrograph of F20 mix (1.50 KX).

Fig. 7. Micrograph of F30 mix (1.50 KX).

Fig. 8. Micrograph of F40 mix (1.50 KX).

Fig. 9. Micrograph of F50 mix (1.50 KX).

Fig. 10. Micrograph of F60 mix (1.50 KX).

1924 R. Siddique et al. / Construction and Building Materials 25 (2011) 1916–1925

4. In all the mixes, C2S, C3S, and C4AF peaks were not visible indi-cating that they are totally consumed. Also, the consumption oflime was indicated due to lowering of pH from 13 to between11.40 and 11.80 at 365 days.

5. The presence of calcium hydroxide was not detected in any ofthe mixes, which confirmed the consumption of calciumhydroxide in the hydration reaction, making dense micro struc-ture and additional development of C–S–H gel, causingincreased strength and resistance to aggressive environment.

6. The microstructure as studied by SEM for concretes with foun-dry sand has shown the reduced voids and C–S–H gel paste isnot as widely spread when compared to control mix.

7. Also, presence of foundry sand particles could be seen at variousplaces in various F mixes. Better spread of C–S–H gel and forma-tion of nodules was observed to increase from F20 to F50 mixwith maximum for F30 mix.

R. Siddique et al. / Construction and Building Materials 25 (2011) 1916–1925 1925

8. The formation of maximum tendrils in F30 mix could be attrib-uted to quick set. Also, the findings of SEM were in accordancewith the mechanical test results obtained for strength. Microcracks were observed in the mix F60.

9. A good correlation between SEM micrographs and mechanicalproperties was observed. Results of this investigation suggestthat used-foundry sand could be very conveniently used inmaking good quality concrete and construction materials.

Even though the strength development is marginally less forfoundry sand at some percentages of replacements, it can be equa-ted to lower grade of normal concrete and making utilization ofwaste material justifies the concrete mix-development. Foundrysand used as fine aggregates replacement enables the large utiliza-tion of waste product.

References

[1] Naik TR, Parikh DM, Tharaniyil MP. Beneficial utilization of used foundry sandsas construction materials. Report No. CBU-1992-22, Center of by-productutilization, Dept Civil Eng Mech University of Wisconsin, Milwaukee; 1992.

[2] Javed S, Lovell CW. Use of waste foundry sand in civil engineering. Transportresearch record 1486. In: Washington, DC: Transportation Res Board; 1994b. p.109–3.

[3] Naik TR, Patel VM, Parikh DM, Tharaniyii MP. Utilization of used foundry sandin concrete. J Mater Civil Eng 1994;6(2):254–63.

[4] Naik TR, Singh SS, Ramme BW. Performance and leaching assessment of flowable slurry. J Environ Eng 2001:359–68.

[5] Khatib JM, Ellis DJ. Mechanical properties of concrete containing foundry sand.ACI special publication SP-200. American Concrete Institute; 2001. p. 733–48.

[6] Naik TR, Kraus RN, Chun YM, Ramme BW, Singh SS. Properties of fieldmanufactured cast-concrete products utilizing recycled materials. J Mater CivilEng ASCE 2003:400–7.

[7] Naik TR, Kraus RN, Chun YM, Ramme WB, Siddique R. Precast concreteproducts using industrial by-products. ACI Mater J 2004;101(3):199–206.

[8] Bakis R, Koyuncu H, Demirbas A. An investigation of waste foundry sand inasphalt concrete mixtures. Waste Mange Res 2006;24:269–74.

[9] Fiore S, Zanetti MC. Foundry wastes reuse and recycling in concreteproduction. Am J Env Sci 2007;3(3):135–42.

[10] Siddique R, Schutter GD, Noumowe A. Effect of used-foundry sand on themechanical properties of concrete. Cons Build Mater 2009;23:976–80.

[11] RILEM Committee CPC-18. Measurement of hardened concrete carbonationdepth. Mater Struct 1988;18: 453–5.

[12] Soroka I. Portland cement paste and concrete. Macmillan Press Ltd; 1979.[13] ACI Committee 211. Recommended practice for selecting proportions for

normal and heavy weight concrete. J Am Concr Inst 1977;71(11):59–60.[14] Lam L, Wong YL, Poon CS. Effect of fly ash and silica fume on compressive and

fracture behaviours of concrete. Cem Concr Res 1998;28(2):271–83.[15] Mehta PK, Monterio PJM. Concrete microstructure, properties, and

materials. Tata McGraw-Hill Edition; 2006.[16] Sulapha P, Wong SF, Wee TH, Swaddiwudhipong S. Carbonation of concrete

containing mineral admixture. J Mater Civil Eng 2003;15(2):134–43.[17] Wee TH, Suryavanshi AK, Logendran D. Pore structure controlling the

carbonation of a hardened cement matrix blended with mineral admixture.Adv Cem Res 1999;11(2):81–95.

[18] Castroa P, Sanjuan MA, Ganesca J. Carbonation of concrete in Mexico Gulf.Build Env 2000;35:145–9.

[19] Corinaldesi V, Moriconi G. Influence of mineral additions on the performanceof 100% recycled aggregate concrete. Constr Build Mater 2009;23:2869–76.

[20] Hewlett Peter C, editor. Lea’s chemistry of cement and concrete. Jordan Hill,Oxford OX2 8DP: Butterworth-Heinemann Lincare House; 2001.