Materials and Design 73 (2015) 60–66

Contents lists available at ScienceDirect

Materials and Design

journal homepage: www.elsevier .com/locate /matdes

High alkali-resistant basalt fiber for reinforcing concrete

http://dx.doi.org/10.1016/j.matdes.2015.02.0220261-3069/� 2015 Elsevier Ltd. All rights reserved.

⇑ Corresponding author at: Division of Chemical Technology and New Materials,Chemistry Department, Lomonosov Moscow State University, Moscow 119991,Russia.

E-mail address: yakov.lipatov@gmail.com (Y.V. Lipatov).

Ya.V. Lipatov ⇑, S.I. Gutnikov, M.S. Manylov, E.S. Zhukovskaya, B.I. LazoryakChemistry Department, Lomonosov Moscow State University, Moscow 119991, Russia

a r t i c l e i n f o a b s t r a c t

Article history:Received 13 November 2014Accepted 21 February 2015Available online 26 February 2015

Keywords:Basalt fiberAlkali resistanceAR-glass fiberZirconia dopingFiber reinforced concrete

Basalt glasses and fibers with zirconia content in the range from 0 to 7 wt% were obtained using ZrSiO4 asa zirconium source. Weight loss and tensile strength loss of fibers after refluxing in alkali solution weredetermined. Basalt fiber with 5.7 wt% ZrO2 had the best alkali resistance properties. Alkali treatmentresults in formation of protective surface layer on fibers. Morphology and chemical composition of sur-face layer were investigated. It was shown that alkali resistance of zirconia doped basalt fibers is causedby insoluble compounds of Zr4+, Fe3+ and Mg2+ in corrosion layer. Mechanical properties of initial andleached fibers were evaluated by a Weibull distribution. The properties of basalt fibers with ZrSiO4 werecompared with AR-glass fibers. The performance of concrete with obtained fibers was investigated.

� 2015 Elsevier Ltd. All rights reserved.

1. Introduction

Alkali-resistant glass (AR-glass) fiber was preferably designedto reinforce cementitious matrices which are widely used in con-struction industry. AR-glass fiber is applied in glass fiber reinforcedconcrete (GFRC) and textile reinforced concrete (TRC) production.Usually short AR-glass fibers are added to GFRC in order to preventdrying shrinkage of concretes at early ages and to increase the frac-ture toughness of the brittle matrix [1]. TRC consists of a fine-grained cement matrix and high-performance textile: fabrics,meshes, yarns. This composite material is notable for high strengthproperties and pseudo-ductile behavior, which is characterized bylarge deformations due to its tolerance to multiple cracking [2].TRC and GFRC can be used both in the creation of slender architec-tural constructions and in the strengthening of reinforced concretestructures. Most of these applications require that high tensilestrength and toughness of TRC do not degrade significantly withtime [3].

A significant alkalinity of Portland cement leads to corrosion ofglass fiber reinforced materials. Alkaline medium of cementremains not only at the hardening stage, but also hereafter dueto the presence of pore solution in the concrete. To evaluate thealkali resistance of glass fibers, the following solutions are mainlyused: NaOH, saturated Ca(OH)2, cement solution and mixtures ofNaOH and Na2CO3 [4]. The major factors, affecting the corrosionrate of glass in alkaline solutions, are as follows: the surface of area

exposed, the volume of the leaching solution, the nature of theleaching solution, its replenishing frequency, and the temperatureof leaching [5].

There are several stages in the reaction of glass with alkalinesolution. Initially OH� is adsorbed on the glass surface. Then thefollowing reactions occur:

−− +⋅→++−−−−−−− OHOmHnSiOOmHOHSiOSiOSi 222

|

|

|

|

|

|ð1Þ

−+ ++−−⇔+−− OHROHSiOHORSi|

|2

|

|ð2Þ

nSiO2 �mH2Oþ 2qNaOH! qNa2O � nSiO2 � pH2Oþ nH2O ð3Þ

At the final stage, the reaction products are removed from theglass surface. Since the number of active adsorption centers onthe glass surface is limited, the corrosion process is stabilized ata certain phase. The following glass destruction depends only ondissolution rate of the reaction products [6].

There are several approaches to increase alkali resistance offibers:

to improve glass composition [5,7];to apply new sizings and coatings [8–11];to use cement and concrete additives [12,13].

To realize the first approach, zirconia addition is most widelyused. Improvement of alkali resistance properties is caused bythe formation of a thin stable hydrated, zirconium-rich layer onthe glass surface. The layer, formed after the beginning of alkalineattack on the glass network, slows down the diffusion of OH� ions

Fig. 1. XRD patterns of deposits obtained on the bottom of the platinum crucible inthe glass preparation process (j – quartz SiO2, d – zirconia ZrO2).

Y.V. Lipatov et al. / Materials and Design 73 (2015) 60–66 61

into the bulk glass. Thus, further network breakdown can be sig-nificantly reduced [3]. The alkali resistance of glass fibers rises withincreasing zirconia content [14]. Most of AR-glass fibers containlarge amount of zirconia (16–20 wt%) providing therefore higheralkali resistance as compared to other glass fibers [5,7]. The maindisadvantage of AR-glass fiber is a high cost.

In recent years basalt continuous fiber (BCF) is increasinglyused in GFRC and TRC along with the AR-glass fiber. BCF is notablefor its low cost, high mechanical properties [15] and good resis-tance to alkaline attack both at room temperature and at elevatedtemperatures [16,17]. Application of basalt fiber results in anincrease of GFRC mechanical properties [18]. Studies indicate thefeasibility of using BCF as reinforcement material in TRC compos-ites [19]. Moreover, basalt fiber is cheaper than AR-glass fiber.However, alkali resistance properties of basalt fiber are inferiorto AR-glass fibers [6].

The present work is a continuation of previously publishedstudy [20] devoted to effect of ZrO2 addition on the alkali resis-tance and mechanical properties of basalt continuous fiber. In thispaper, to increase the alkali resistance of basalt fiber, zircon(ZrSiO4) was used as a zirconium source.

2. Experimental

2.1. Sample preparation

Basalt fibers with various zirconia contents were produced intwo stages. The first stage included obtaining of basalt bulk glasses.Basalt rock from the Sil’tsevskoe deposit (Carpathians, Ukraine)was used as a raw material. Zirconia doped basalt glasses were pre-pared by adding of ZrSiO4 to milled basalt batch. The batch mixturewas heated in platinum crucible in a high-temperature furnace at arate of 250 �C/h up to 1000 �C and at 30 �C/h in the range of 1000–1600 �C, then homogenized at 1600 �C for 24 h. The molten glasswas quenched in water from 1550–1590 �C. The chemical compo-sition of glasses is presented in Table 1. Total iron content isexpressed as Fe2O3 in the paper. In the case of ZrSi6–ZrSi16 glasses,a white deposit was observed on the bottom of platinum crucibleafter the molten glass had been poured out. XRD patterns of depos-its are presented in Fig. 1.

In the second stage, basalt fibers with various zirconia contentswere produced from obtained glasses on a laboratory scale system[21]. The deposit formed during the preparation of glassesZrSi6–ZrSi16 was not used. The fibers had filament diameters of10–12 lm. Sizing was not applied in the fiber production.

Basalt fiber properties were compared with commercial AR-glass fiber (manufacturer – Owens Corning) characteristics. It hasthe following chemical composition: 9.8% Na2O, 65.1% SiO2, 6.3%CaO, 18.8% ZrO2. AR-glass fiber was preliminarily annealed at500 �C to completely remove sizing, then it was placed in platinumcrucible and was heated up to 1600 �C. Glass quenching and the

Table 1XRF analysis of basalt glasses.

Sample Added to the basalt batch Chemical composition (w

ZrSiO4 (wt%) In terms of ZrO2 (wt%) ZrO2 Na2O M

ZrSi0 0 0 – 2.1(2) 3ZrSi2 2 1.3 1.3(1) 2.1(2) 3ZrSi4 4 2.7 2.7(1) 2.0(2) 3ZrSi6 6 4.0 3.6(1) 2.0(2) 3ZrSi8 8 5.4 4.9(1) 2.0(2) 3ZrSi10 10 6.7 5.4(1) 1.9(2) 3ZrSi12 12 8.1 6.3(1) 1.9(2) 3ZrSi14 14 9.4 6.9(1) 1.9(2) 3ZrSi16 16 10.8 7.0(1) 1.9(2) 3

fiber production were carried out according to the methoddescribed above.

The materials used in concrete preparation include cement CEMI 32.5N, mortar sand with particle size 0–1.6 mm and choppedfibers with 12 mm in length. Chopped fibers were prepared by cut-ting of continuous fibers obtained on a laboratory scale system. Thedensity of basalt, ZrSi10 and AR-glass fibers were equal to 2.65 g/cm3, 2.69 g/cm3 and 2.57 g/cm3 respectively. The mixture propor-tions of concrete are presented in Table 2. Dry mixture of compo-nents was mixed for 2 min and then water was added slowly. Theamount of water was chosen so that the slump was 105–115 mm.The fresh concrete was mixed for 3 min to ensure even dispersionof fibers in the concrete and was cast in molds. Concrete in moldswas vibrated for 3 min with frequency 50 Hz. Specimens werecured in a chamber at 20 ± 3 �C and 95 ± 5% relative humidityand were de-molded after 24 h. Hardened concrete was tested at28 days.

2.2. Testing methods

X-ray fluorescence (XRF) analysis of the glasses was performedon a PANalytical Axios Advanced spectrometer. CharacteristicX-rays were excited by an Rh-anode X-ray tube (current capacityup to 4 kW, maximal current 160 mA). For the XRF analysis, pow-dered glasses were pressed in pellets with polystyrol as a binder.

X-ray diffraction (XRD) measurements were performed on aTHERMO ARL X’TRA powder diffractometer with a semiconductingPeltier-cooled detector (Cu Ka1 radiation, k = 1.54060 Å, Cu Ka2radiation, k = 1.54443 Å). XRD patterns were collected in the range

t%)

gO Al2O3 SiO2 K2O CaO TiO2 Fe2O3

.4(2) 15.2(5) 55.3(4) 1.5(2) 9.0(5) 1.10(8) 11.9(3)

.3(2) 14.9(5) 54.9(4) 1.5(2) 8.8(4) 1.08(8) 11.7(3)

.3(2) 14.6(5) 54.4(4) 1.4(1) 8.6(4) 1.06(7) 11.4(2)

.2(2) 14.4(5) 54.1(4) 1.4(1) 8.5(4) 1.04(7) 11.3(2)

.2(2) 14.1(5) 53.6(4) 1.4(1) 8.3(4) 1.02(7) 11.0(2)

.1(2) 14.0(5) 53.5(4) 1.4(1) 8.3(4) 1.01(7) 10.9(2)

.1(2) 13.8(5) 53.2(4) 1.4(1) 8.2(4) 1.00(7) 10.8(2)

.1(2) 13.6(5) 53.0(4) 1.4(1) 8.1(4) 0.99(7) 10.7(2)

.1(2) 13.6(5) 53.0(4) 1.3(1) 8.1(4) 0.99(7) 10.7(2)

Table 2Mixture proportions of concrete without fibers (C-0), with basalt (C-bas), ZrSi10(C-ZrSi10) and AR-glass (C-AR) fibers.

Sample Fiber content Cement(kg/m3)

Sand(kg/m3)

Water/cement ratio

By volume(%)

By weight(kg/m3)

C-0 0 0 500 1500 0.5C-bas 0.5 13.25 500 1500 0.6C-ZrSi10 0.5 13.45 500 1500 0.6C-AR 0.5 12.85 500 1500 0.6

Fig. 2. XRD patterns (a) and SEM image (b) of initial basalt fibers.

62 Y.V. Lipatov et al. / Materials and Design 73 (2015) 60–66

of 2h = 10–70� with a scan step 2h = 0.02� and scan rate of 1�/min.Phase analysis was determined using Crystallographica Search-Match software and ICDD PDF-2 database.

A JEOL JSM-6390LA microscope was used for the scanning elec-tron microscope (SEM) analysis. The accelerating voltage was set to20 kV. The morphology of the fibers was determined in secondaryelectron imaging (SEI) mode. Energy-dispersive X-ray (EDX) analy-sis was carried out on a JEOL EX-54175 JMH system. Before theexaminations all the fibers were coated with a conducting carbonlayer.

Alkali resistance of the fibers was determined by the weight lossafter alkali treatment. Samples of 10–12 lm diameter fiber withtotal surface area of 5000 cm2 were refluxed in a plastic flask with250 ml mixture of 1 M NaOH and 0.5 M Na2CO3 in the ratio of 1:1for 3 h in a water bath. The experimental temperature of alkalisolution was 98 �C. Then the fibers were washed thoroughly withdistilled water and were dried in air to constant weight.

The tensile strength of the fibers was measured on a TiniusOlsen H5KS universal testing machine. Monofilaments weresticked on in support paper frames with gauge length of 10 mm

using epoxy. Samples were tensed with a constant speed of5 mm/min till destruction. A set of 25 samples were tested for eachtype of fibers.

The performance of concrete with fibers was investigated usingHounsfield H100K-S universal testing machine. Flexural strengthwas tested through the third-point bending experiments conduct-ed on 40 � 40 � 160 mm prismatic specimens in accordance withRussian standard GOST 30744. The span of flexural experimentwas 100 mm. The load was applied by displacement control witha rate of 0.2 mm/min until the specimens failed. Compressivestrength was determined on 50 mm cubic specimens. Uniaxialtensile strength was measured on 40 � 40 � 280 mm prismaticspecimens with gauge length of 120 mm. Samples were tensedwith a constant speed of 0.2 mm/min till destruction.Compressive and uniaxial tensile strength tests were carried outaccording to GOST 10180. In this study, every test result consistsof three replicate tests.

3. Results and discussion

3.1. Basalt fibers with ZrSiO4

Based on data in Table 1, it can be concluded that there isincomplete reaction between ZrSiO4 and basalt batch starting withZrSi6 sample. XRD analysis of the deposit (Fig. 1) observed on thebottom of crucible indicates formation of crystalline zirconia andquartz. According to the literature [22], ZrSiO4 begins to decom-pose into ZrO2 and SiO2 in the range of 1450–1600 �C dependingon the purity and particle size of the reagent. At temperature above1650 �C the reaction rate increases many times. Despite the slowbatch heating up to 1600 �C and further exposure of melt at thistemperature for 24 h, ZrSiO4 did not quantitatively react withbasalt under these conditions. It can be assumed that a certain partof zirconium silicate interacts with glass matrix while the otherpart is decomposed into ZrO2 and SiO2 that are deposited on thecrucible bottom. The maximum zirconia content in basalt glassequals 7.0(1) wt%.

According to XRD data (Fig. 2), all obtained fibers are amor-phous. The surface of fibers is smooth and has not defects andmicroinclusions. In our previous paper [20] it was shown thatthe zirconia addition to basalt batch in baddeleyite form(monoclinic ZrO2) leads to crystallization of monoclinic and tetra-gonal ZrO2 on the fiber surface during fiber melt spinning. Usingzircon as a zirconia source in this study has not such disadvantage.It should be pointed out that the solubility limit of zirconia inbasalt glass remains constant and is amounted to 7.0 wt% in thecase of both baddeleyite and zircon addition. At the same time,use of ZrSiO4 does not cause phase crystallization during fiber meltspinning, and the fiber chemical composition is almost identical tothe glass chemical composition. Therefore, the maximum zirconiacontent in fibers with ZrSiO4 addition equals 6.9(2) wt% (Table 3)and in fibers with ZrO2 addition equals 3.1(2) wt% [20]. We believethat zirconia addition to basalt matrix in form of ZrSiO4 occursmuch readily due to the chemical bonds between zirconium andsilicon–oxygen tetrahedra in zircon, and zirconium incorporatesin the basalt matrix structure in a bound state.

3.2. Alkali resistance

The formation of corrosion layer was observed on the fiber sur-face after refluxing in mixture of 1 M NaOH and 0.5 M NaOH. Thesurface layer had a thickness in the range of 0.2–1.0 lm dependingon the fiber composition (Fig. 3). A similar surface layer formed bythe products of reaction between glass fiber and alkali solution isdescribed in many papers [4,17,23,24]. As follows from the EDX

Table 3EDX analysis of fiber surface composition (wt%): I – initial fiber, L – leached fiber, ±D – confidential interval (a = 0.05).

Oxide ZrSi0 ZrSi2 ZrSi4 ZrSi6 ZrSi8 ZrSi10 ZrSi12 ZrSi14 ZrSi16 AR ±D

I L I L I L I L I L I L I L I L I L I L

Na2O 1.9 0.6 1.7 1.3 1.9 1.5 2.1 1.4 1.9 1.6 1.9 1.4 2.0 1.4 1.9 1.5 2.0 1.5 9.8 6.8 0.3MgO 3.7 6.2 3.5 5.4 3.5 5.5 3.3 4.9 3.1 4.8 3.2 6.1 3.2 7.0 3.0 6.9 3.0 6.7 – – 0.4Al2O3 15.5 8.2 15.6 10.5 14.9 12.1 14.6 13.7 14.2 13.7 13.9 12.5 13.7 11.8 13.6 11.0 13.5 11.3 – – 0.6SiO2 55.6 34.1 55.5 41.5 54.1 43.9 53.8 47.1 53.8 48.5 53.7 46.0 53.7 44.7 53.4 41.9 53.5 41.5 65.1 60.5 0.6K2O 1.8 1.3 1.7 1.2 1.6 1.3 1.7 1.3 1.6 1.3 1.5 1.0 1.5 1.1 1.4 1.0 1.3 1.1 – – 0.2CaO 8.7 18.4 8.5 13.8 8.5 10.5 8.6 8.8 8.0 8.7 8.5 9.3 7.9 8.6 8.0 8.9 8.2 8.9 6.3 7.8 0.5TiO2 1.0 3.1 1.0 2.5 1.1 1.9 1.1 1.6 1.0 1.5 0.9 1.4 1.0 1.3 1.0 1.4 1.0 1.3 – – 0.2Fe2O3 11.8 28.1 11.4 21.9 11.6 18.7 11.3 15.6 11.3 13.1 10.7 14.0 10.8 14.6 10.9 16.2 10.6 16.1 – – 0.6ZrO2 – – 1.1 1.9 2.8 4.6 3.5 5.6 5.1 6.8 5.7 8.3 6.2 9.5 6.8 11.2 6.9 11.6 18.8 24.9 0.2

Fig. 3. Cross-section (a) and surface layer (b) of ZrSi4 basalt fiber after refluxing inmixture (1:1) of 1 M NaOH and 0.5 M Na2CO3 for 3 h at 98 �C.

Y.V. Lipatov et al. / Materials and Design 73 (2015) 60–66 63

data, the chemical composition in the center of fiber remains thesame after leaching, while the chemical composition of surface lay-er undergoes significant changes (Table 3). Alkali treatment leadsto a decrease of sodium, potassium, aluminum and silicon on thesurface of all basalt fibers. The content of magnesium, calcium, tita-nium, iron and zirconium (in zirconia doped fibers) simultaneouslyincreases. It seems that at first the potassium and sodium ions,which weakly bound with glass network, are removed from thefiber surface. A decrease of silicon and aluminum is caused bydegradation of the glass network. In this case, the reaction prod-ucts are readily soluble in alkali solution and are removed fromthe fiber surface. Poorly soluble compounds of magnesium, calci-um, titanium, iron and zirconium form a surface shell that slows

down the further fiber damage. It is worth noting that the alkaliresistance of basalt fibers without zirconia is primarily due toinsoluble compounds of Fe3+, Ca2+, Mg2+ and Ti4+, the content ofwhich in the surface layer increases by 1.5–3 times after leaching.At the same time, Zr4+, Fe3+ and Mg2+ play key roles in alkali resis-tance of zirconia doped basalt fiber. The chemical composition ofAR-glass fiber is simpler than basalt, therefore Zr4+ is the maincomponent increasing alkali resistance. Solubility of the zirconiumcompounds is less than solubility of calcium, iron and magnesiumcompounds formed after alkali treatment. The solubility productconstants of Ca(OH)2, Fe(OH)3, Mg(OH)2, Zr(OH)4 equal 5.5 � 10�6,6.3 � 10�38, 6 � 10�10, 1 � 10�52 respectively [25]. Therefore, the sur-face layer of ZrSi0–ZrSi6 fibers is loose and porous. Water mole-cules lead to swelling of the surface layer and increasing itsvolume. Corrosion shell starts to peel from the fiber surface(Fig. 4). The increase of zirconium on the surface of ZrSi8–ZrSi10fibers after alkali treatment leads to formation of compact layer.Compact layer reduces the corrosion rate of basalt fiber. Despitethe high ZrO2 content in ZrSi14–ZrSi16 fibers, the iron and magne-sium content significantly increased after alkali treatment. Itseems that the high content of these elements results in formationof porous layer similar to the ZrSi0–ZrSi6 surface layers.

Thus, it can be concluded the leaching of zirconia doped basaltfibers under study begins at the surface. If the reaction products offiber with alkali solution are soluble, it will be removed from thefiber surface. Otherwise, the products remain on the surface andform the insoluble protective layer retarding the diffusion ofhydroxyl anions to the intact glass surface underneath the shelland inhibiting further fiber corrosion.

The addition of ZrSiO4 to the basalt glasses results in alkaliresistance enhancement of the fibers. With increasing zirconia upto 5.7(2) wt% the weight loss of fiber decreases from 0.065(6) to0.028(6) mg/cm2 after leaching. Further enhancement of ZrO2 con-tent up to 6.9(2) wt% leads to an increase of weight loss up to0.063(6) mg/cm2 (Fig. 5). Such behavior can be explained by theformation of firmly surface shell on the fiber with the highest alkaliresistance (ZrSi8–ZrSi10). According to the EDX analysis, reductionof SiO2 by 10% and Al2O3 by 3.5% on the ZrSi8 surface is the lowestamong the all basalt fibers (Table 3), i.e. this sample corrodes inalkali solution with a minimum rate. In the case of ZrSi0–ZrSi6and ZrSi14 fibers, the surface layer is partially peeled off in someareas (Fig. 4) resulting in increase of weight loss after leaching. Itseems that the corrosion shell of ZrSi16 fiber has been completelyremoved and then the formation of the protective surface layer hasstarted anew.

It was experimentally found that weight loss of AR-glass fiberafter leaching is amounted to 0.019(6) mg/cm2, while the weightloss of the most alkali resistant basalt fiber ZrSi10 equals0.028(6) mg/cm2. It is emphasized that the zirconia content inZrSi10 fiber equals 5.7(2) wt% that the less than in AR-glass fiber(18.8(2) wt% ZrO2). As in the case of ZrSi8–ZrSi10, a firmly

Fig. 4. SEM images of leached basalt fibers.

Fig. 5. Weight loss of basalt fibers after refluxing in mixture (1:1) of 1 M NaOH and0.5 M Na2CO3 for 3 h at 98 �C.

64 Y.V. Lipatov et al. / Materials and Design 73 (2015) 60–66

protective layer completely covered the entire surface of AR-glassfibers after leaching (Fig. 6).

Fig. 6. SEM images of AR-glass fibers: (a) initial fibers, (b) leached fibers.

3.3. Tensile strength

Mechanical test data were analyzed by the Weibull distribution,based on the concept of the failure of the weakest link [26].According to Weibill distribution probability of failure P of indi-vidual filament at the applied stress r can be written as

Table 4Mechanical properties of fibers: hri – average tensile strength, ±D – confidential interval (a = 0.05), r0 – scale parameter of Weibull distribution, m –shape parameter of Weibulldistribution, R2 – coefficient of determination, Dm – decrease of shape parameter after leaching.

Sample Initial fiber Leached fiber Dm

hri (MPa) ±D (MPa) r0 (MPa) m R2 hri (MPa) ±D (MPa) r0 (MPa) m R2

ZrSi0 1820 119 1962 5.1 0.96 623 173 663 1.4 0.99 3.7ZrSi2 2210 143 2336 6.6 0.97 938 99 1011 3.7 0.98 2.9ZrSi4 2117 105 2275 5.5 0.96 930 78 1005 3.5 0.96 2.0ZrSi6 2113 145 2295 4.1 0.98 1067 96 1145 3.1 0.96 1.0ZrSi8 2093 122 2260 4.0 0.98 1109 115 1181 3.5 0.95 0.5ZrSi10 2103 154 2295 4.1 0.98 1155 118 1266 3.5 0.95 0.6ZrSi12 2064 163 2263 3.8 0.97 1042 84 1116 3.2 0.94 0.6ZrSi14 1959 158 2142 3.8 0.96 1056 150 1182 2.4 0.95 1.4ZrSi16 1702 125 1872 4.0 0.95 973 112 1046 3.7 0.95 0.3AR 1559 94 1668 5.8 0.97 1422 98 1503 5.4 0.97 0.4

Table 5Mechanical properties of hardened concrete at 28 days without fibers (C-0), withbasalt (C-bas), ZrSi10 (C-ZrSi10) and AR-glass (C-AR) fibers.

Sample Compressivestrength (MPa)

Flexuralstrength (MPa)

Uniaxial tensilestrength (MPa)

Density(kg/m3)

C-0 33.1 ± 1.5 3.9 ± 0.1 0.7 ± 0.1 2197 ± 19C-bas 27.7 ± 1.6 4.2 ± 0.2 0.8 ± 0.1 2092 ± 27C-ZrSi10 26.9 ± 1.1 4.6 ± 0.2 1.0 ± 0.1 2080 ± 14C-AR 27.5 ± 1.6 4.8 ± 0.3 1.0 ± 0.1 2105 ± 21

Y.V. Lipatov et al. / Materials and Design 73 (2015) 60–66 65

PðrÞ ¼ 1� e�ðrr0Þm; ð4Þ

where r0 – scale parameter and m – shape parameter. The shapeparameter m or Weibull modulus is a measure of the strength vari-ability of a material; the scale parameter r0 is the stress level forwhich the failure probability is 63.2%. In order to determine r0

and m, graph ln(�ln(1 � P)) versus lnr was plotted.Mechanical properties of the fibers and the tensile strength

evaluation by Weibull distribution function are listed in Table 4.The addition of a small amount of ZrSiO4 to the basalt fibers leadsto rise of tensile strength from 1.8(1) GPa to 2.2(1) GPa (ZrSi0–ZrSi2). With increasing zirconia content from 2.8(2) to 6.2(2) wt%(ZrSi4–ZrSi12), the tensile strength remains constant and on theaverage equals 2.1(2) GPa. On reaching the maximum content ofZrO2 in basalt fiber (ZrSi14–ZrSi16), the tensile strength reduces.The tensile strength evaluation by Weibull distribution indicatesthat the shape parameter increases from 5.1 to 6.6 in the rangefrom ZrSi0 to ZrSi2 and further reduces to 3.8–4.0. The high valuesof the shape parameters indicate a high degree of surface homo-geneity, resulting in a breakage of fibers in a narrow interval of fail-ure stresses. A decrease of the shape parameter points out widersize distribution of defects on the fiber surface. It can be assumedthat the phenomenon is related to the structural features of zirco-nia doped basalt glasses. According to the literature, there is nosingle point of view on the zirconium role in glass: a number ofstudies [27,28] states that zirconium acts as a network former, inother papers [29] it is indicated that zirconium is a modifier. Thefirst small amounts of Zr4+ appear to behave as a network formerinto the glass structure and to increase the polymerization degreeof glass network, resulting in enhancement of the tensile strength.The further additives of Zr4+ act as a modifier filling the vacantpositions in the glass network and the tensile strength does notchange. When a certain threshold is reached, zirconium surplusleads to ‘‘loose’’ of the glass structure and to fall of the tensilestrength.

Alkali resistance of fibers is characterized not only by weightloss, but also by tensile strength loss after leaching. The retainedtensile strength of basalt fibers without zirconia equals0.6(2) GPa; the retained tensile strength equals 1.2(1) GPa for themost alkali resistant basalt fiber ZrSi10. The tensile strength ofAR-glass fibers after leaching is amounted to 1.4(1) GPa (Table 4).Besides, it can be noted that leaching leads to reduction of theshape parameters for all types of fibers. The attack of OH� groupsstarts with existing surface defects and the corrosion rate will bedifferent depending on defect size. Moreover, alkaline mediummay leads to formation of new defects on fiber surface. Thus,amount and size of defects increase. As a result of these processes,significant changes take place in the morphology of fiber surface(Fig. 4). The shape parameter greatly decreases for fibers with par-tially peeled off surface layer. ZrSi16 fiber has negligible traces of

old corrosion shell. In this case smooth fiber surface leads todecrease in difference of shape parameter between initial and lea-ched fibers. The obtained results are in good agreement with paper[4], which deals with corrosion mechanism of glass fibers in alkalisolutions. It can be summarized that after leaching in identical con-ditions basalt fibers with different ZrO2 content are at various cor-rosion stages. The fibers with high alkali resistance are completelycovered by firmly surface layer. Corrosion shell starts to peel offfrom the surface of fibers with medium resistance. In the case ofbasalt fibers with low alkali resistance, corrosion shell is removed,and formation of new surface layer is repeated.

3.4. Fiber reinforced concrete

The performance of fibers in cementitious matrix was deter-mined by adding chopped fibers in concrete. Basalt, zirconia dopedbasalt (ZrSi10) and AR-glass fibers were used for producing fiberreinforced concrete. Mechanical properties of hardened concreteat 28 days are presented in Table 5.

The addition of 0.5% fibers to concrete resulted in 16–19%reduction of compressive strength for all samples. Similar behaviorwas also reported in papers [30,31]. During testing, the specimensof concrete without fibers were broken into several parts. For con-cretes with fibers, the specimens had a small amount of cracks andremained a good appearance after fracture. The differences in com-pressive strength between concretes with various types of fibersare negligible. Fresh concretes with fibers needed more water toreach necessary workability that negatively affects their hydration.Moreover, the addition of fibers rises the number of voids incementitious matrix and reduces concrete density.

Chopped fibers increase significantly the flexural and uniaxialtensile strength of concrete. Compared with concrete withoutfibers, the flexural strength of concretes with basalt, ZrSi10 andAR-glass fibers increases by 8%, 18% and 23% respectively. Thecontribution of fibers in the uniaxial tensile strength of concretesis displayed an even greater degree. The increase of the uniaxialtensile strength equals 14% for concrete with basalt fibers andequals 43% for concrete with ZrSi10 and AR-glass fibers. The fiberseffectively slow down the propagation of crack and thus improve

66 Y.V. Lipatov et al. / Materials and Design 73 (2015) 60–66

the mechanical behavior and toughness of the specimen [32]. It isworth noting that concretes with ZrSi10 and AR-glass fibers havehigher flexural and uniaxial tensile strength compared to concretewith basalt fibers. Despite the alkali resistance of fibers was deter-mined under more severe conditions than conditions surroundingfibers in cementitious matrix, zirconia doped basalt fibers hadcomparable performance with AR-glass fibers in concrete.

4. Conclusions

In the present paper it is shown that the maximum zirconiacontent in basalt fibers is amounted to 6.9(2) wt% using ZrSiO4.Zirconia doped fibers with smooth and homogeneous surface havegood mechanical properties. The surface morphology undergoessignificant changes after alkali treatment. Formation of corrosionlayer with high content of Zr4+, Fe3+ and Mg2+ takes place, whichpartially or even completely peeled off from the fiber surface.The tensile strength evaluation by Weibull distribution functionindicates increasing of both amount and size variability of the sur-face defects. The weight loss of ZrSi10 basalt fiber with 5.7(2) wt%ZrO2 is comparable with commercial AR-glass fiber containing18.8(2) wt% ZrO2. An important factor in alkali resistance evalua-tion of fiber is also retained tensile strength after leaching. Thetensile strength of zirconia doped basalt fiber with optimal compo-sition ZrSi10 and AR-glass fiber after alkali treatment equals1.2(1) GPa and 1.4(1) GPa respectively. To evaluate the effect ofchopped fibers on concrete properties, obtained fibers were usedfor producing fiber reinforced concrete. Zirconia doped basaltfibers demonstrated high performance in concrete. Thus, it canbe concluded that the basalt fibers with ZrSiO4 addition, studiedin this paper, are very promising material for reinforcing concrete.

References

[1] G. Barluenga, Fiber–matrix interaction at early ages of concrete with shortfibers, Cem. Concr. Res. 40 (2010) 802–809.

[2] R. Barhum, V. Mechtcherine, Effect of short dispersed glass and carbon fibreson the behavior of textile-reinforced concrete under tensile loading, Eng. Fract.Mech. 92 (2012) 56–71.

[3] M. Butler, V. Mechtcherine, S. Hempel, Experimental investigations on thedurability of fibre–matrix interfaces in textile-reinforced concrete, Cem. Concr.Compos. 31 (2009) 221–231.

[4] C. Scheffler, T. Förster, E. Mäder, G. Heinrich, S. Hempel, V. Mechtcherine, Agingof alkali-resistant glass and basalt fibers in alkaline solutions: evaluation of thefailure stress by Weibull distribution function, J. Non-Cryst. Solids 355 (2009)2588–2595.

[5] A. Paul, Chemical durability of glasses; a thermodynamic approach, J. Mater.Sci. 12 (1977) 2246–2268.

[6] A.A. Pashchenko, V.P. Serbin, V.S. Klimenko, et al., Armirovanieneorganicheskih vyazhushhih veshhestv mineral’nymi voloknami, Moscow,Strojizdat, 1988 (in Russian).

[7] A.J. Majumdar, J.M. West, L.J. Larner, Properties of glass fibres in cementenvironment, J. Mater. Sci. 12 (1977) 927–936.

[8] S.G. Gao, E. Mäder, A. Abdkader, P. Offermann, Sizings on alkali-resistant glassfibers: environmental effects on mechanical properties, Langmuir 19 (2003)2496–2506.

[9] S.L. Gao, E. Mäder, R. Plonka, Coatings for glass fibers in a cementitious matrix,Acta Mater. 52 (2004) 4745–4755.

[10] V.A. Rybin, A.V. Utkin, N.I. Baklanova, Alkali resistance, microstructural andmechanical performance of zirconia-coated basalt fibers, Cem. Concr. Res. 53(2013) 1–8.

[11] T.H. Jung, R.V. Subramanian, Alkali resistance enhancement of basalt fibers byhydrated zirconia films formed by the sol–gel process, J. Mater. Res. Soc. 9(1994) 1006–1013.

[12] A. Nourredine, Influence of curing conditions on durability of alkali-resistantglass fibres in cement matrix, Bull. Mater. Sci. 34 (4) (2011) 775–783.

[13] M. Butler, V. Mechtcherine, S. Hempel, Durability of textile reinforced concretemade with AR glass fibre: effect of the matrix composition, Mater. Struct. 43(2010) 1351–1368.

[14] R.G. Simnah, Chemical durability of ZrO2 containing glasses, J. Non-Cryst.Solids 54 (1983) 335–343.

[15] S.I. Gutnikov, A.P. Malakho, B.I. Lazoryak, V.S. Loginov, Influence of alumina onthe properties of continuous basalt fibers, Russ. J. Inorg. Chem. 54 (2009) 191–196.

[16] V. Velpari, B.E. Ramachandran, T.A. Bhaskaran, B.C. Pai, N. Balasubramanian,Alkali resistance of fibres in cement, J. Mater. Sci. 15 (1980) 1579–1584.

[17] B.E. Ramachandran, V. Velpari, N. Balasubramanian, Chemical durabilitystudies on basalt fibres, J. Mater. Sci. 16 (1981) 3393–3397.

[18] T.M. Borhan, Properties of glass concrete reinforced with short basalt fibre,Mater. Des. 42 (2012) 265–271.

[19] F.J. De Caso y Basalo, F. Matta, A. Nanni, Fiber reinforced cement-basedcomposite system for concrete confinement, Constr. Build. Mater. 32 (2012)55–65.

[20] Ya.V. Lipatov, S.I. Gutnikov, M.S. Manylov, B.I. Lazoryak, Effect of ZrO2 on thealkali resistance and mechanical properties of basalt fibers, Inorg. Mater. 48(2012) 751–756.

[21] S.I. Gutnikov, M.S. Manylov, Ya.V. Lipatov, B.I. Lazoryak, K.V. Pokholok, Effect ofthe reduction treatment on the basalt continuous fiber crystallizationproperties, J. Non-Cryst. Solids 368 (2013) 45–50.

[22] A. Kaiser, M. Lobert, R. Telle, Thermal stability of zircon (ZrSiO4), J. Eur. Ceram.Soc. 28 (2008) 2199–2211.

[23] M. Friedrich, A. Schulze, G. Prösch, C. Walter, D. Weikert, N.M. Binh, et al.,Investigation of chemically treated basalt and glass fibres, Mikcochim. Acta133 (2000) 171–174.

[24] B. Wei, H.-L. Cao, S.-H. Song, Tensile behavior contrast of basalt and glass fibersafter chemical treatment, Mater. Des. 21 (2010) 4244–4250.

[25] Yu.Yu. Lure, Spravochnik po analiticheskoj himii, Moscow, Himija, 1979 (inRussian).

[26] W. Weibull, A statistical distribution function of wide applicability, J. Appl.Mech. 18 (1951) 293–297.

[27] A.E. Ringwood, The principles governing trace element distribution duringmagmatic crystallization Part I: The influence of electronegativity, Geochim.Cosmochim. Acta 7 (1955) 242–254.

[28] O.B. Lapina, D.F. Khabibulin, V.V. Terskikh, Multinuclear NMR study of silicafiberglass modified with zirconia, Solid State Nucl. Magn. Reson. 39 (2011) 47–57.

[29] K. Linthout, Alkali-zirconosilicates in peralkaline rocks, Contrib. Mineral.Petrol. 86 (1984) 155–158.

[30] D.P. Dias, C. Thaumaturgo, Fracture toughness of geopolymeric concretesreinforced with basalt fibers, Cem. Concr. Compos. 27 (2005) 49–54.

[31] N. Kabay, Abrasion resistance and fracture energy of concretes with basaltfiber, Constr. Build. Mater. 50 (2014) 95–101.

[32] C. Jiang, K. Fan, F. Wu, D. Chen, Experimental study on the mechanicalproperties and microstructure of chopped basalt fibre reinforced concrete,Mater. Des. 58 (2014) 187–193.