e ect of marble dust on the corrosion resistance of glass...

Scientia Iranica A (2016) 23(3), 896{906

Sharif University of TechnologyScientia Iranica

Transactions A: Civil Engineeringwww.scientiairanica.com

E�ect of marble dust on the corrosion resistance ofglass �ber reinforced mortars

O. Kele�stemur� and S. Y�ld�z

Department of Civil Engineering, Faculty of Technology, Firat University, 23119, Elazig, Turkey.

Received 14 September 2015; received in revised form 11 June 2015; accepted 6 September 2015

KEYWORDSCorrosion;Mortar;Glass �ber;Marble dust;Microstructure.

Abstract. In this study, mechanical and physical properties of cement mortars containingmarble dust and glass �ber in various combinations were investigated, in addition toanalyzing the corrosion behavior of reinforcing steels embedded in these mortars. Tothis end, cement mortars containing glass �ber (0 kg/m3, 0.25 kg/m3, 0.50 kg/m3, and0.75 kg/m3) were prepared. Besides, with the purpose of determining e�ect of marble dustaddition on the corrosion resistance of the glass �ber reinforced mortars, marble dust wasadded to all series by replacing with �ller sand at proportions of 0%, 10%, 30%, and 50%by volume. Corrosion studies were carried out in two stages. Firstly, corrosion potentialof reinforcing steels in the mortars was measured every day for a period of 150 days basedon the ASTM C-876 standard. Secondly, cathodic polarization values of the steels wereobtained by using the Tafel extrapolation technique and, subsequently, corrosion currentdensities were determined. Thanks to the study, it was observed that a decrease in thecorrosion resistance of the mortars had taken place as a result of glass �ber addition. But, itwas determined that with the addition of marble dust into the mortars, corrosion resistanceof the specimens was signi�cantly increased.© 2016 Sharif University of Technology. All rights reserved.

1. Introduction

Leaving the waste materials in the environment candirectly cause environmental problems. In general, theindustry of dimensional stone marble has contributedto the development of major environmental problemsdue to waste generation at various stages of miningand processing operations [1]. Marble blocks are cutinto smaller blocks in order to give the desired smoothshape. During the cutting process, about 25% marblemass is lost in the form of dust. In Turkey, marbledust is settled by sedimentation and then dumpedaway, which results in environmental pollution andalso causes forming dust in the summer threatening

*. Corresponding author. Tel.: +90 4242370000;Fax: +90 4242367064E-mail addresses: okelestemur@�rat.edu.tr (O.Kele�stemur); syildiz@�rat.edu.tr (S. Y�ld�z)

both agriculture and public health. Therefore, useof the marble dust in various sectors, especially theconstruction, agriculture, glass, and paper industries,will help to protect the environment [2].

Many researchers were recently interested instudying the possibility of re-using the waste marbledust in useful industries, especially with regard to thebuilding and construction materials such as cement,concrete, and brick blocks [1]. The e�ect of marbledust on the mechanical and physical properties ofcement based materials was investigated; most of theresearch showed positive results and bene�ts [3-9]. But,not a single study has been encountered on corrosionperformance of cement mortars prepared by replacingvery �ne sand with waste marble dust in the openliterature.

Glass �bers are a type of high-strength �ber ma-terials and have many application areas in constructionsector. Glass �bers commonly use composite materials

O. Kele�stemur and S. Y�ld�z/Scientia Iranica, Transactions A: Civil Engineering 23 (2016) 896{906 897

in concrete owing to their great tensile strength andmodule; glass �bers have great potential to be used inconcrete due to their superior characteristics in termsof high sti�ness, low density and water absorption, hightensile strength, corrosion resistance, etc. [10-13].

Reinforced concrete is the most widely usedconstruction material all over the world [14]. Theuse of reinforced concrete structures is based on theprinciple that concrete is an ideal environment forreinforcing steel. The high alkalinity of concretecauses the formation of a saturated hydrated iron oxidepassivating �lm on the surface of the steel, whichprovides good protection against corrosion. However,depassivation may be occurred by carbonation of theconcrete cover or by the presence of chloride salts,thus initiating expansive corrosion of the steel, causingeventual damage to the surrounding concrete [15].A good-quality concrete with low permeability is aprimary corrosion control measure for reinforcing steelembedded in concrete. Previous studies have shownthat use of marble dust as sand replacement in concretemay reduce its permeability [5,7].

The main objective of this study was to inves-tigate the e�ect of marble dust addition on corrosionbehavior of reinforcing steels embedded in glass �berreinforced cement mortar specimens. Furthermore, thee�ect of marble dust on the mechanical and physicalproperties of the glass �ber reinforced mortar speci-mens was also investigated.

2. Materials and methods

Total of sixteen series of mortar specimens includingthe control specimen were prepared in order to examinethe corrosion behavior of reinforcing steels embeddedin mortar specimens with marble dust and glass �ber invarious combinations. For the corrosion tests, a total ofeighty pieces of mortar specimens were obtained, with�ve specimens being taken from each series.

2.1. Preparing electrodesAs an electrode, the SAE1010 steel bar produced byEre�gli Iron and Steel Factories in Turkey was selectedfor this study. The as-received material was in the formof a hot-rolled bar 12 mm in diameter. The chemicalcomposition of the electrode is given in Table 1.

160 pieces of steel bars in 120 mm of length werecut out from the as-received material and their surfaceswere mechanically cleaned. Then, specimen surfaceswere polished with 1200 mesh sandpaper. Polishedsurfaces were cleaned with ethyl alcohol. Surface areas

Table 1. Chemical composition of the steel (% wt.).

C Mn Si P S Fe

0.17 0.250 0.050 0.005 0.050 Balance

Figure 1. Electrodes used in corrosion experiments.

(10 cm2) were kept open in the tips of electrodes,which would be embedded in the mortar. Screw threadwas machined at the other ends of the electrodes andcables were connected to these ends in order to makemeasurements during the experiment in an easier way.The remaining sections of the steel electrodes wereprotected against external e�ects by covering themwith epoxy resin at �rst and then with polyethylene.Figure 1 shows the electrodes used in the corrosionexperiments.

2.2. Preparing mortar specimens for thecorrosion tests

100 � 100 � 200 mm mortar specimens were preparedfor the corrosion experiments, in which the electrodesprepared in advance were embedded.

Commercial grade ASTM Type I Portland ce-ment, which is produced in Turkey as CEM I Portlandcement, was used in the preparation of all cement mor-tar specimens that were employed in the experimentswithin the scope of this study.

The marble sludge was obtained in wet form asan industrial by-product directly from the depositsof marble factories, which forms during the sawing,shaping, and polishing processes of the marbles in theElazig province of Turkey. The marble sludge was driedin an oven at approximately 50�C for 48 h before thepreparation of the cement mortar specimens. The driedmaterial was passed through 0.25 mm sieve and �nallythe marble dust was obtained to be used in the cementmortar specimens as very �ne sand. The chemicalproperties of the marble dust and cement used in theexperiments are presented in Table 2.

High-quality river sand was used as aggregate,which is widely employed in mortar production. Max-imum grain size of aggregate was 4 mm. The densityof the river sand was 2690 kg/m3. Various proportions(0%, 10%, 30% and 50%, by volume) of the very �nesand (passing through 0.25 mm sieve) were replacedwith waste marble dust. The grain size distributionsof very �ne sand and waste marble dust are shown inFigure 2.

The glass �bers were circular straight �bers ob-

898 O. Kele�stemur and S. Y�ld�z/Scientia Iranica, Transactions A: Civil Engineering 23 (2016) 896{906

Table 2. Chemical compositions of the cement andmarble dusts.

Oxide compounds(mass %)

CEM I 42.5 N Marbledust

SiO2 21.12 12.67

Al2O3 5.62 1.07

Fe2O3 3.24 3.79

CaO 62.94 40.44

MgO 2.73 10.84

Density (g/cm3) 3.10 2.78

tained from Camelsan. The properties of the glass �berused in the study are given in Table 3.

The mixture designs of all the cement mortargroups are presented in Table 4. A superplasticizer(SP) was used to improve workability of the mortarmixes. Modi�ed polycarboxilate based superplasti-cizer, obtained from SIKA, was used as 1% of cementweight. Regular tap water was used as the mixingwater during the preparation of the mortar specimens.

The cement mortar specimens were kept in moldsfor duration of 24 h at room temperature of about 20�2�C. After demolding, these specimens were cured forseven days at 25�C in 100% relative humidity, before

Figure 2. Grain size distributions of the very �ne sandand marble dust.

getting partially submerged in a 3% NaCl solution toinduce a corrosive environment.

2.3. Corrosion testsCorrosion tests were conducted in two stages. In the�rst stage, the corrosion potential of reinforcing steelsembedded in the mortar specimens was measured everyday for a period of 150 days in accordance with ASTMC876 method [16]. Saturated copper/copper sulfate(Cu/CuSO4) was used as the reference electrode, anda high impedance voltmeter was used as the mea-

Table 3. Properties of the glass �ber used in the experiments.

Color Length(mm)

Diameter(�m)

Density(g/cm3)

Young'sModulus (MPa)

TensileStrength (MPa)

White 6 13 2.68 72000 1700

Table 4. Details of the cement mortar mixes (kg/m3).

Exp.no

Designationof mixtures

Fiberglass

Marbledust

Aggregate(0� 0:25)

Aggregate(0:25� 4)

Water Cement SP

1 C (GF0-MD0) - - 407 1253.8 221.4 450 4.452 GF0-MD10 - 38.63 367 1253.8 221.4 450 4.453 GF0-MD30 - 115.9 285.2 1253.8 222.8 450 4.454 GF0-MD50 - 193.12 203.6 1253.8 222.8 450 4.455 GF0.25-MD0 0.25 - 407 1253.8 221.4 450 4.456 GF0.25-MD10 0.25 38.63 367 1253.8 221.4 450 4.457 GF0.25-MD30 0.25 115.9 285.2 1253.8 222.8 450 4.458 GF0.25-MD50 0.25 193.12 203.6 1253.8 222.8 450 4.459 GF0.50-MD0 0.50 - 407 1253.8 221.4 450 4.4510 GF0.50-MD10 0.50 38.63 367 1253.8 221.4 450 4.4511 GF0.50-MD30 0.50 115.9 285.2 1253.8 222.8 450 4.4512 GF0.50-MD50 0.50 193.12 203.6 1253.8 222.8 450 4.4513 GF0.75-MD0 0.75 - 407 1253.8 221.4 450 4.4514 GF0.75-MD10 0.75 38.63 367 1253.8 221.4 450 4.4515 GF0.75-MD30 0.75 115.9 285.2 1253.8 222.8 450 4.4516 GF0.75-MD50 0.75 193.12 203.6 1253.8 222.8 450 4.45


Table 5. Estimation of corrosion probability as determined by half-cell potential test.

Potential (mV) (CSE) Probability of the presence of active corrosion> �200 The probability for corrosion is very low

�200 � �350 Uncertain< �350 The probability for corrosion is very high

surement device in corrosion potential measurements.Changes in corrosion potentials versus time were in-dicated in graphs in order to determine whether thereinforcing steel bar was active or passive. Recommen-dations on evaluation of potential measurement resultsin ASTM C876 experiment method are speci�ed inTable 5 [17,18].

In the second stage, the anodic and cathodicpolarization values of the electrodes in the mortarswere obtained by using the galvanostatic technique andthen the corrosion current densities were determinedwith the aid of cathodic polarization curves. The Tafelextrapolation technique (TP) was used to determinethe polarization curves. The TP is electrochemicalmethod for calculating corrosion rate based on intensityof the corrosion current (Icorr) and the Tafel slopes.This technique is based upon application of eithersteady �xed levels of current followed by monitoringof the potential (galvanostatic) or application of aspeci�c potential followed by monitoring of the current(potentiostatic). The galvanostatic method was used inthis study. Su�cient Tafel slopes were obtained by ap-plying constant current densities from 0:01 �A/cm2 to20 �A/cm2. In this technique, generally, cathodic datais preferred since it is easier to be used for experimentalmeasurement. Furthermore, often, anodic curves maynot exhibit linear behavior near Ecorr. Therefore,corrosion current densities of the reinforcing steelsembedded in the mortar specimens were determinedby using cathodic polarization curves. The corrosioncurrent density, Icorr, was obtained from the Tafelplot by extrapolating the linear portion of the cathodiccurve to Ecorr.

The experimental setup used for the applicationof galvanostatic method is shown in Figure 3. In thiscircuit, the electrode that is connected to the positiveterminal is the anode and the other that is connectedto the negative terminal of the DC power source, whichsupplies a �xed voltage of 20 V DC to the system, isthe cathode. The same material (SAE1010 steel bar)has been used for anode and cathode electrodes.

2.4. Hardened concrete experimentsThe corrosion behaviors of the mortar specimensconsisting of marble dust and �ber glass at variousproportions as well as their mechanical and physicalproperties like compressive strength, exural strength,and porosity were investigated. Furthermore, sorptiv-ity measurement was also conducted on the mortar

Figure 3. The experimental setup used for theapplication of galvanostatic technique.

specimens. The data was interpreted together withthe corrosion rate of steels embedded in these mortarspecimens.

The mixtures were cast from the same batchinto prismatic (40 � 40 � 160 mm) and cubic (50 �50 � 50 mm) steel molds for exural and compressivestrength experiments, respectively. The specimenswere kept in the molds for 24 h at room temperatureof about 20 � 2�C. After demolding, these specimenswere cured in lime saturated water for 28 days.

Five specimens from each series were used todetermine the compressive and exural strengths at28 days according to TS EN 196-1. The compressionloads were applied at rates of 2400 N/s and 50 N/s todetermine the compressive and exural strengths of themortar specimens, respectively. For exural strength ofthe mortars, the specimens were loaded from their midspan and the clear distance between simple supportswas 120 mm.

The porosity of the cement mortar specimens wasmeasured applying the vacuum-saturation technique.The porosity measurements were carried out on slicesof 68 mm diameter cores cut out from the centerof 100 mm cubes (parallel to the casting direction).The specimens were dried in an oven at 100 � 5�Cuntil constant mass was achieved and they were thenplaced in a desiccator under vacuum for at least 3hours after which the desiccator was �lled with de-aired distilled water. The porosity value for eachmortar specimen was calculated through Eq. (1). Thismethod for measuring the porosity has previously been


reported [19-21].

P =(Wsat �Wdry)(Wsat �Wwat)

100; (1)

where, P stands for vacuum saturation porosity (%);Wsat weight in air of saturated specimen; Wwat weightin water of saturated specimen; and Wdry weight ofoven-dried specimen.

Five test specimens for sorptivity measurementwere prepared for each mixture. The sorptivity mea-surements were carried out on 50 mm cube specimens.Measurements of capillary sorption were carried outusing specimens pre-conditioned in the oven at about50�C until constant mass was achieved. Then, themortar specimens were cooled down to room temper-ature. The side surface of each specimen was sealedwith para�n to achieve unidirectional ow. The end ofthe specimen that would not be exposed to water wassealed using a loosely attached plastic bag to controlevaporation from this surface. The plastic bag wassecured using an elastic band. Rods supports wereplaced at the bottom of the pan and the pan was�lled with tap water so that the water level was 5 mmabove the top of the rods supports. The water levelin the pan was maintained 5 mm above the top of thesupport for the duration of the tests. As shown inFigure 4, the test surface of the specimen was placedon the rods supports. At certain times, the mass ofthe mortar specimens was measured using a balance;then, the amount of absorbed water was calculatedand normalized with respect to the cross-section area ofthe specimens subjected to the water at various timessuch as 0, 5, 10, 20, 30, 60, 180, 360, and 1440 min.The capillary absorption coe�cient (k) was obtainedby using the following equation:QA

= kpt; (2)

where Q stands for the amount of water absorbed

Figure 4. Schematic demonstration for the measurementof water capillary sorption.

(cm3); A the cross-section of specimen that was incontact with water (cm2); t time (s); and k thesorptivity coe�cient of the specimen (cm/s1=2). Todetermine the sorptivity coe�cient, Q=A was plottedagainst the square root of time (

pt), and then, k was

calculated from the slope of the linear relation betweenQ=A and

pt. This method of measuring the capillary

absorption has been used in many previous studies [21-25].

The microscopic analyses of the cement mortarspecimens were performed at the Electron MicroscopyLaboratory of Firat University using a Jeol JSM7001Fscanning electron microscope.

3. Results and discussion

3.1. Results of mechanical and physicalexperiments conducted on hardenedmortars

The data obtained through mechanical and physicalexperiments conducted on the hardened mortar speci-mens is presented in Table 6.

The results presented in Table 6 indicated thatthe compressive and exural strengths of the cementmortars increased with increasing the percentage ofmarble dust. Moreover, in contrast to marble duste�ect, the use of glass �ber in the mortar specimenswith and without marble dust increased the exuralstrength but decreased the compressive strength of themortars.

Table 6 demonstrates that compressive strengthof the mortar specimens decreased depending on theincrease in the amount of glass �ber. This outcomeis expected because glass �bers have more ductilestructure as per the cement matrix and aggregatewhen the glass �bers are added to the specimens;they cause discontinuity in the cement matrix. Thus,the compressive strength of the mortars decreases asthe amount of glass �ber content increases. Similarrelationships have also been reported by the studiesappearing in [26,27]. Figure 5 shows the SEM (Scan-ning Electron Microscope) images of the microstructureof glass �ber surface and cement matrix. It can beseen from Figure 5(a) that glass �ber surface has beencovered with the cement matrix. This phenomenonindicates a bond between the glass �ber and mortar.However, there are voids between �bers and matrixand the interface layer of the mortar is quite loose.The SEM micrograph shown in Figure 5(b) speci�esthat the glass �ber pulled out from the cement matrixpresents a smooth surface. This indicates that adebonding between cement matrix and glass �ber mayoccur in some areas of the cement mortar. This eventmay also result in a decrease in compressive strengthof the glass �ber reinforced mortar specimens.

In general, as can be clearly concluded from


Table 6. Results of mechanical and physical tests of the specimens.

SpecimensCompressive

strength(MPa)

Flexuralstrength(MPa)

Porosity(%)

Sorptivitycoe�cient

(10�3 cm.s�1=2)

C (GF0-MD0) 84.70 11.51 17.10 0.86

GF0-MD10 87.32 12.20 16.91 0.80

GF0-MD30 93.41 12.74 16.65 0.74

GF0-MD50 96.83 13.24 16.42 0.54

GF0.25-MD0 82.73 11.84 17.35 1.13

GF0.25-MD10 84.07 12.53 17.29 1.06

GF0.25-MD30 87.09 13.00 17.11 0.92

GF0.25-MD50 92.64 13.50 16.86 0.85

GF0.50-MD0 81.77 12.32 17.69 1.30

GF0.50-MD10 83.16 12.71 17.56 1.26

GF0.50-MD30 86.45 13.27 17.29 1.08

GF0.50-MD50 90.15 13.76 17.15 1.02

GF0.75-MD0 79.97 12.57 17.82 1.42

GF0.75-MD10 81.11 13.09 17.68 1.29

GF0.75-MD30 84.91 13.59 17.46 1.21

GF0.75-MD50 88.65 13.98 17.24 1.15

Figure 5. SEM images of the glass �bers in the cement mortar.

Table 6, in contrast to compressive strength comparedwith the control specimen, the glass �ber reinforcementincreased the exural strength of the cement mortarspecimens. Previous studies have reported that theaddition of �ber into cement based materials can signif-icantly improve exural strength of the specimens [28-30]. This may be attributed to the fact that after the exural cracking of the glass �ber reinforced mortarspecimens, the stress is transferred to the bridging�bers, and thus the development of macro-cracks isdelayed and exural strength of the mortar specimensis increased. These observations are in good agreementwith earlier �ndings [26-28,31,32].

It was also observed that the control mortar

specimen showed brittle behavior. When subjected tothe exural loading, the control specimen was brokeninto two parts once a bending crack occurred. But,the addition of the glass �ber e�ectively slowed downthe propagation of the cracks, and therefore exuralstrength of the mortar specimen increased.

The addition of marble dust as very �ne sandinto cement mortar specimens with and without glass�ber resulted in an increase in the compressive and exural strengths. This result may have one underlyingreason; the reason is that the marble dust has a muchsmall grain size compared to very �ne sand, enablingit to �ll in the voids. The �neness of an admixture ishighly critical for the modi�cation of aggregate/cement


Figure 6. SEM micrographs of a) control and b) GF0-MD50 mortar specimens.

paste interface zone, which is the weakest link of aconcrete's structure [33]. It can be seen clearly fromFigure 2 that average particle size of the very �ne sandis very large compared to marble dust, therefore its�ller e�ect may not be su�cient as marble dust. It canbe concluded that compressive and exural strengthsof the mortars without marble dust are relatively low,because the �ller e�ect of the very �ne sand is notas good as that of marble dust. The marble dustparticles, however, may act as ideal micro �ne �ller inthe interfaces between aggregate and cement paste orpores in the bulk paste. This �ller e�ect improves thecement matrix and transition zone properties. Thus,high levels of compressive and exural strengths areachieved in the mortars with marble dust compared tothat of the only glass �ber reinforced mortars. Theseobservations agreed fairly well with those obtained byother researchers [1,7,34,35].

Table 6 shows the change of porosity values of thecement mortar specimens. From this table, it can beseen that the glass �ber reinforced mortar specimensshowed the highest percentage of porosity. This can beexplained by the redistribution of the void structuredue to the inclusion of glass �bers.

It can be seen from Table 6 that the addition ofglass �ber into mortar mixture signi�cantly increasedthe sorptivity coe�cient. This may be due to distri-bution of the glass �bers and the increased porosityin the cement paste in the contact zone close to the�bers. Similar relationship has also been noticed byother researchers [36,37].

From Table 6, it can be seen that the usage ofmarble dust as a partial replacement of very �ne sandshowed a decrease in porosity value and sorptivitycoe�cient of the mortar specimens. This decreasecould be explained as a result of the pore-�lling e�ectof �ne marble dust that improves the properties of thetransition zone surrounding �ber and aggregate. The�ller e�ect of marble dust has also been reported bythe other researchers [1,7,34].

A number of SEM micrographs that illustrate themicro-structure characteristics of some cement mortarspecimens are shown in Figure 6. During scanningprocess, it was observed that marble dust blendedcement mortar specimens were denser and less porousthan the control specimen. The control and glass�ber reinforced mortars have larger capillary pores andpossess lower compressive strength, which results inthe higher capillary sorption in these mortars. Sincemarble dust is very �ne, pores in the bulk paste or inthe interfaces between glass �ber and mortar are �lledby this admixture. Therefore, the capillary pores arereduced in the mortars containing marble dust. Thebene�cial role of the marble dust causes an increasein the mechanical strength and a reduction in thecapillary sorption of the mortars with and withoutglass �ber. This �nding is in good agreement with theprevious study reported by Demirel [7].

3.2. Results of corrosion experimentsThe results obtained from corrosion potential mea-surements of reinforcing steel embedded in the mortarspecimens containing marble dust and glass �ber invarious combinations are shown in Figures 7-11.

As seen in Figure 7, the electronegative corrosionpotential is much more in all the glass �ber reinforced

Figure 7. Corrosion potentials of the C, GF0.25, GF0.50,and GF0.75 specimens.


Figure 8. Corrosion potentials of the C, MD10, MD30,and MD50 specimens.

Figure 9. Corrosion potentials of the C, GF0.25-MD10,GF0.25-MD30, and GF0.25-MD50 specimens.



mortar specimens of any type than that in the controlspecimen. Large increases were observed in electroneg-ative corrosion potential when the glass �ber amountwas increased. Indeed, the electrodes in the controlspecimen became more passive than those in the glass�ber reinforced specimens and they remained in thepassive zone in terms of corrosion during the 70 daysperiod. Then, the corrosion potentials of the electrodesin the control specimen entered the uncertain zoneand stabilized afterwards. The corrosion potentials ofthe reinforcing steels embedded in GF0.25 specimenreached the uncertain zone after the 2nd week andremained there for all the remaining duration of thetest. GF0.50 and GF0.75 mortar specimens, whichpossess the highest electronegative corrosion potential,reached the active zone after the 3rd month and re-mained there until the end of the 150th day. By takingASTM C876 as a reference, these observations wereconcluded to indicate that the corrosion still continuedin the GF0.50 and GF0.75 specimens even at the endof the 150th day. The GF0.75 specimen has the lowestcorrosion resistance among the glass �ber reinforcedmortar specimens. By taking account of the resultsof corrosion potentials by ASTM C876 standard, itcan be remarked that the control mortar specimen hashigher corrosion resistance than the mortar specimenscontaining glass �ber in various amounts.

As clearly seen from Figures 8-11, marble dustaddition increased the passivation rate of reinforcingsteels embedded in the mortar specimens with andwithout glass �ber. In fact, the reinforcing steels inthe specimens with only marble dust (MD10, MD30,and MD50) became more passive compared to the Cspecimen and remained in the passive zone in termsof corrosion during the 150 days test period. Thepassivation rate of the steels in the mortars withmarble dust increased depending on the increase inthe percentage of marble dust. Furthermore, marbledust increased the passivation rate of the steel barsin the glass �ber reinforced mortars. So that, theglass �ber reinforced mortars of any type remainedin the uncertain zone even at the 150th day. Thesesituations are an indication of that; while the corrosionresistance of a mortar is reduced by adding glass�ber, it increases again as marble dust is added tothose mortars containing glass �ber. The additionof marble dust such that it would replace with very�ne sand in various proportions in mortars consistingof glass �ber up to 0.75 kg/m3 (GF0.25, GF0.50,and GF0.75) has enhanced the corrosion resistanceof the mortar specimens and carried some glass �berreinforced specimens (GF0.25-MD50, GF0.25-MD30,and GF0.50-MD50) above that of the C specimen.

The corrosion potential measurement providesqualitative observations and probably points out thecorrosion of reinforcing steel bar embedded in mortar to


Figure 12. Changes in the Icorr of the mortar specimens.

a large extent. On the other hand, reliable quantitativedata on the corrosion of steel bar embedded in mortarspecimen can be obtained by measuring resistance orcurrent density of the steel.

Results of corrosion polarization studies on thedeveloped mortar specimens with respect to coppersulfate electrode are shown in Figure 12. The data ofcorrosion current density (Icorr) shown in Figure 12has been derived from the experimentally obtainedcathodic polarization curves using Tafel's linear extrap-olation method.

The corrosion current density values seem tocon�rm the corrosion potential values. As is clearlyseen in Figure 12, the C specimen exhibits the highestcorrosion resistance compared to the mortar specimensproduced with only glass �ber. Furthermore, corrosionresistance has decreased in parallel to the increase involume fraction of glass �ber in the mortar specimen.This decrease in corrosion of the specimens results fromthe voids between the glass �bers and matrix. As it isknown, oxygen input into the mortar is facilitated athigher porosities. An increase in oxygen concentrationincreases the current density and corrosion rate of thespecimen. Thus, the corrosion resistance of the mortarspecimens with glass �ber drops o� since the oxygeninput increases due to the voids emerging after theglass �ber addition. As seen in Figure 12, however,the corrosion resistance of the mortar specimens hasbeen increased as a result of marble dust addition tothe mortars with and without glass �ber. As alreadyspeci�ed, it is attributed to the fact that in the Cmortar, the average particle size of the very �ne sandis higher than the specimens with marble dust andthe pores in bulk paste and interfaces are not �lledcompletely. The �ne particles result in a relativelydense structure and more discontinuous pores in thecement mortar. It also reduces the volume of largevoids and capillaries. The resultant cement matrixis more chemically resistant, and it has a denser mi-croscopic pore structure and a relatively impermeablemortar specimen. Table 6 shows that cement mortarscontaining marble dust demonstrate the lowest water

absorption ability and porosity among all the mortarspecimens studied in this work. Water normally entersthe mortar through capillary action; the fewer andsmaller the capillary pores are, the less water entersthe mortar and the lower the rates of oxygen di�usion.Due to the high alkalinity in cement mortar, thereare very few H+ ions to be consumed in the cathodicreaction, so that the oxygen has to be reduced in orderto sustain the cathodic reaction. Therefore, when theoxygen reduction rate falls, so does the corrosion rate.Chloride ions tend to eliminate the normal passivationstate of reinforcing steel bars in mortar. Low air voidcontents and low water absorptivity help keep chlorideions in going through the mortar and reaching thesurface of the steel bars [33,38]. Hereby, a mortarcontaining marble dust, which has a denser microscopicpore structure, is relatively impermeable to chlorideions. Thus, in this study, addition of the marble dust asvery �ne sand into the mortars with and without glass�ber e�ectively increased the corrosion resistance of thespecimens which were submerged in chloride solutions.

4. Conclusions

On the basis of experimental study that has beencarried out and presented in this paper, the followingconclusions can be drawn:

1. When marble dust was added to the mortar speci-mens with and without glass �ber, it improved boththe mechanical strength and other performanceparameters (porosity and sorptivity values). This�nding is due to the fact that marble dust particlesmay act as ideal micro �ller in the interfacesbetween glass �ber and mortar or pores in the bulkpaste;

2. In contrast to marble dust e�ect, the use of glass�ber in the cement mortars increased the exuralstrength, porosity, and sorptivity values, but de-creased the compressive strength of the specimens.The reduction in the compressive strength and risein the porosity and sorptivity values of the mortarscan be attributed to the voids between the glass�ber and mortar. Furthermore, glass �bers giverise to discontinuity in the cement matrix due totheir more ductile structure than that of the cementmatrix;

3. As a result of the experiments conducted for thepurpose of determining the corrosion resistance ofglass �ber reinforced mortars, it has been observedthat the corrosion resistance of the specimens de-creases in line with increasing amount of glass �ber.However, addition of the marble dust as very �nesand into the glass �ber reinforces mortar specimensand signi�cantly increases the corrosion resistanceof these specimens;


4. The experimental �ndings indicate that the addi-tion of the marble dust in a way that it wouldreplace the very �ne sand at proportions of 10%,30%, and 50% by volume into the mortars withoutglass �ber has reduced the corrosion rate of thesteel electrodes embedded in these mortars to evenlower levels than that of the electrodes in the controlmortar.

References

1. Aliabdo, A.A., Elmoaty, M.A. and Auda, E.M. \Re-use of waste marble dust in the production of cementand concrete", Const. Build. Mat., 50, pp. 28-41(2014).

2. Kara�sahin, M. and Terzi, S. \Evaluation of marblewaste dust in the mixture of asphaltic concrete",Const. Build. Mat., 21, pp. 616-620 (2007).

3. Y�ilmaz, H., G�ur�u, M., Day�, M. and Tekin, I. \Uti-lization of waste marble dust as an additive in cementproduction", Mater Des., 31, pp. 4039-42 (2010).

4. Corinaldesi, V., Moriconi, G. and Naik, T.R. \Charac-terization of marble powder for its use in mortar andconcrete", Const. Build. Mat., 24, pp. 113-7 (2010).

5. Binici, H., Kaplan, H. and Yilmaz, S. \In uenceof marble and limestone dusts as additives on somemechanical properties of concrete", Sci. Res. Essay.,2(9), pp. 372-9 (2007).

6. Erg�un, A. \E�ects of the usage of diatomite and wastemarble powder as partial replacement of cement onthe mechanical properties of concrete", Const. Build.Mat., 25, pp. 806-12 (2011).

7. Demirel, B. \The e�ect of the using waste marbledust as �ne sand on the mechanical properties ofthe concrete", Int. J. Phys. Sci., 5(9), pp. 1372-1380(2010).

8. Kelecstemur, O., Ar�c�, E., Y�ld�z, S. and G�ok�cer, B.\Performance evaluation of cement mortars containingmarble dust and glass �ber exposed to high tempera-ture by using Taguchi method", Const. Build. Mat.,60, pp. 17-24 (2014).

9. Kele�stemur, O., Y�ld�z, S., G�ok�cer, B. and Ar�c�,E. \Statistical analysis for freeze-thaw resistance ofcement mortars containing marble dust and glass�ber", Mater. Design, 60, pp. 548-555 (2014).

10. Emiroglu, M., Beycioglu, A. and Yildiz, S. \ANFISand statistical based approach to prediction the peakpressure load of concrete pipes including glass �ber",Exp. Syst. with Applica., 39, pp. 2877-2883 (2012).

11. Asokan, P., Osmani, M. and Price, A.D.F. \Assessingthe recycling potential of glass �ber reinforced plasticwaste in concrete and cement composites", J. Clean.Product., 17, pp. 821-829 (2009).

12. Asokan, P., Osmani, M. and Price, A.D.F. \Im-provement of the mechanical properties of glass �brereinforced plastic waste powder �lled concrete", Const.Build. Mat., 24, pp. 448-460 (2010).

13. Bai, W., Zhang, J., Yan, P. and Wang, X. \Study onvibration alleviating properties of glass �ber reinforcedpolymer concrete through orthogonal tests", Mater.Design, 30, pp. 1417-1421 (2009).

14. Sideris, K.K. and Savva, A.E. \Durability of mixturescontaining calcium nitrite based corrosion inhibitor",Cem. Conc. Comp., 27(2), pp. 277-287 (2005).

15. Gonzalez, J.A., Molina, A., Otero, E. and Lopez, W.\On the mechanism of steel corrosion in concrete: Therole of oxygen di�usion", Mag. Conc. Res., 42(150),pp. 23-27 (1990).

16. ASTM C876-91, Standard Test Method for Half-CellPotentials of Uncoated Reinforcing Steel in Concrete,ASTM International, West Conshohocken, PA (1999).

17. Kele�stemur, O., Aksoy, M., Y�ld�z, S. \Corrosionbehavior of tempered dual-phase steel embedded inconcrete", Int. J. Miner. Metall. Mater., 16(1) pp. 43-50 (2009).

18. Kele�stemur, O. and Y�ld�z, S. \E�ect of various dual-phase heat treatments on the corrosion behavior ofreinforcing steel used in the reinforced concrete struc-tures", Const. Build. Mat., 23, pp. 78-84 (2009).

19. Mydin, M.A.O. \Modeling of transient heat transferin foamed concrete slab", J. Eng. Sci. Tech., 8(3), pp.326-343 (2013).

20. Cabrera, J.G. and Lynsdale, C.J. \A new gas per-meameter for measuring the permeability of mortarand concrete", Mag. Conc. Res., 40(144), pp. 177-182(1988).

21. Gonen, T. and Yazicioglu, S. \The in uence of mineraladmixtures on the short and long-term performance ofconcrete", Build. Env., 42, pp. 3080-3085 (2007).

22. Gonen, T. and Yazicioglu, S. \The in uence of com-paction pores on sorptivity and carbonation of con-crete", Const. Build. Mat., 21, pp. 1040-1045 (2007).

23. Tasdemir, C. \Combined e�ects of mineral admixturesand curing conditions on the sorptivity coe�cient ofconcrete", Cem. Conc. Res., 33, pp. 1637-1642 (2003).

24. Turkmen, I. \In uence of di�erent curing conditionson the physical and mechanical properties of concreteswith admixtures of silica fume and blast furnace slag",Mat. Lett., 57(29), pp. 4560-4569 (2003).

25. Demirel, B. and Kele�stemur, O. \E�ect of elevatedtemperature on the mechanical properties of concreteproduced with �nely ground pumice and silica fume",Fire Safe. J., 45, pp. 385-391 (2010).

26. C�avdar, A. \A study on the e�ects of high temperatureon mechanical properties of �ber reinforced cemen-titious composites", Compos. B., 43, pp. 2452-2463(2012).

27. C�avdar, A. \Investigation of freze-thaw e�ects on me-chanical properties of �ber reinforced cement mortars",Compos. B., 58, pp. 463-472 (2014).

28. Jiang, C., Fan, K., Wu, F. and Chen, D. \Experimentalstudy on the mechanical properties and microstructure


of chopped basalt �bre reinforced concrete", Mater.Des., 58, pp. 187-193 (2014).

29. Sivakumar, A. and Santhanam, M. \Mechanical prop-erties of high strength concrete reinforced with metallicand non-metallic �bres", Cem. Conc. Comp., 29(8),pp. 603-608 (2007).

30. Song, P.S. and Hwang, S. \Mechanical properties ofhigh-strength steel �ber-reinforced concrete", Const.Build. Mat., 18(9), pp. 669-673 (2004).

31. Qureshi, L.A. and Ahmed, A. \An investigation onstrength properties of glass �ber reinforced concrete",Int. J. Eng. Res. Tech., 2(4), pp. 2567-2572 (2013).

32. Kene, K.S., Vairagade, V.S. and Sathawane, S. \Ex-perimental study on behavior of steel and glass �berreinforced concrete composites", Bonfring Int. J. In-dust. Eng. Manag. Sci., 2(4), pp. 125-130 (2012).

33. Kele�stemur, O. and Demirel, B. \Corrosion behavior ofreinforcing steel embedded in concrete produced with�nely ground pumice and silica fume", Const. Build.Mat., 24, pp. 1898-1905 (2010).

34. Kristulovic, P., Kamenic, N. and Popovic, K. \A newapproach in evaluation of �ller e�ect in cement", Cem.Conc. Res., 24(4), pp. 721-727 (1994).

35. Topcu, I.B., Bilir, T. and Uygunoglu, T. \E�ect ofwaste marble dust content as �ller on properties ofself-compacting concrete", Const. Build. Mat., 23(5),pp. 1947-1953 (2009).

36. Duran Ati�s, C. and Karahan, O. \The durabilityproperties of polypropylene �ber reinforced y ashconcrete", Mater. Des., 32, pp. 1044-1049 (2011).

37. Duran Ati�s, C. and Karahan, O. \Properties of steel�ber reinforced y ash concrete", Const. Build. Mat.,23, pp. 392-399 (2009).

38. Hou, J. and Chung, D.D.L. \E�ect of admixtures inconcrete on the corrosion resistance of steel-reinforcedconcrete", Corros. Sci., 42, pp. 1489-1507 (2000).

Biographies

O�guzhan Kele�stemur received his MS and PhD de-grees in Construction Education from Firat University,Turkey, in 2003 and 2008. He is currently AssociateProfessor in the Civil Engineering Department of theTechnology Faculty at Firat University. His researchinterests include: concrete, corrosion, dual-phase steel,construction materials.

Servet Y�ld�z received his MS and PhD degrees inCivil Engineering from Firat University, Turkey, in1989 and 1998. He is currently Associate Professor inthe Civil Engineering Department of the TechnologyFaculty at Firat University. He currently serves asVocational School Manager at Firat University. Hisresearch interests include: concrete, construction ma-terials.

e ect of marble dust on the corrosion resistance of glass...

Documents