Composites: Part B 56 (2014) 83–91

Contents lists available at ScienceDirect

Composites: Part B

journal homepage: www.elsevier .com/locate /composi tesb

Combined effect of steel fiber and metakaolin incorporationon mechanical properties of concrete

1359-8368/$ - see front matter � 2013 Elsevier Ltd. All rights reserved.http://dx.doi.org/10.1016/j.compositesb.2013.08.002

⇑ Corresponding author. Tel.: +90 342 3172426; fax: +90 342 3601107.E-mail address: guneyisi@gantep.edu.tr (E. Güneyisi).

Erhan Güneyisi a,⇑, Mehmet Gesoglu a, Arass Omer Mawlod Akoi a, Kasım Mermerdas� b

a Department of Civil Engineering, Gaziantep University, 27310 Gaziantep, Turkeyb Department of Civil Engineering, Hasan Kalyoncu University, 27410, Gaziantep, Turkey

a r t i c l e i n f o

Article history:Received 24 May 2012Accepted 12 August 2013Available online 20 August 2013

Keywords:A. Ceramic–matrix compositesB. Mechanical propertiesB. StrengthSteel fiber reinforced concrete

a b s t r a c t

This study reports the results of an experimental study on mechanical properties of plain and metakaolin(MK) concretes with and without steel fiber. To develop the metakaolin included steel fiber reinforcedconcrete mixtures, Portland cement was partially replaced with MK as 10% by weight of the total bindercontent. Two types of hook ended steel fibers with length/aspect ratios of 60/80 and 30/40 were utilizedto produce fiber reinforced concretes. Two series of concrete groups were designed with water to binderratios (w/b) of 0.35 and 0.50. The effectiveness of MK and different types of steel reinforcement on thecompressive, flexural, splitting, and bonding strength of the concretes were investigated. All tests wereconducted at the end of 28 days of curing period. Analyses of variance on the experimental results werecarried out and the levels of the significance of the variables on the mechanical characteristics of the con-cretes were determined. Moreover, correlation between the measured parameters was carried out to bet-ter understand the interaction between mechanical properties of the concretes. The results revealed thatincorporation of MK and utilization of different types of steel fibers significantly affected the mechanicalproperties of the concretes, irrespective of w/b ratio.

� 2013 Elsevier Ltd. All rights reserved.

1. Introduction

Concrete is the most commonly used building material all overthe world because of its versatility and availability. Especially rein-forced concrete structural elements have been indispensable partsof construction works due to the ease in erection and relativelylower cost than the other structural materials. The proper adher-ence between reinforcing bars and concrete is the most desiredproperty due to the fact that structural performance of reinforcedconcrete members depends on the monolithic behavior. The prom-inent component controlling the competence of the bond is mostlythe quality of concrete. Because, the reinforcing steel bars are ob-tained from a fixed manufacturing process and the properties donot significantly fluctuate compared to concrete. However, struc-tural concretes have many different characteristics dependingmainly on the amount and type of the ingredients [1]. It is reportedthat concrete with improved mechanical property has superioradherence with reinforcing steel bars [2].

Apart from its excellent properties, concrete shows a rather lowperformance when subjected to tensile stress. For this reason, uti-lization of fibers to provide enhancement in tensile strengthbehavior of concrete has attracted the interest of the researchers[3–9]. Mechanical properties of concrete can be improved by

exploitation of reinforcement with randomly oriented short sepa-rated fibers, which obstruct and/or control initiation and propaga-tion of cracks. Fiber reinforced concrete (FRC) can keep on resistingmuch amount of loads even at deflections. The characteristics andperformance of FRC varies depending on matrix properties as wellas the fiber material, fiber concentration, fiber geometry, fiber ori-entation, and fiber distribution [8].

In order to improve the mechanical properties, particularlycompressive strength, use of some pozzolanic materials has beenreported by researchers for many years [10–16]. Pozzolans, suchas silica fume and fly ash, are the most commonly known mineraladmixtures used in production of high-strength concrete. Thesematerials impart additional performance to the concrete throughreacting with Portland cement hydration products to form second-ary C–S–H gel, the part of the paste mainly responsible for concretestrength [17].

For the last two decades, there has been a growing attraction inthe beneficiation of metakaolin (MK) as a supplementary cement-ing material in concrete to enhance its properties. MK is an ultra-fine pozzolana, manufactured by calcination of purified kaolinclay at a temperature ranging from 650 to 900 �C to drive off thechemically bound water and destroy the crystalline structure[18–19]. Unlike other industrial by-product materials, MK needsa thorough process of manufacturing. It has to be carefully refinedto remove inert impurity and ground to particles of micron size.Research has demonstrated that concrete mixtures incorporating

Table 1Properties of Portland cement and metakaolin.

Item Portland Cement Metakaolin

Chemical properties CaO (%) 61.60 0.5SiO2 (%) 19.43 53Al2O3 (%) 5.64 43Fe2O3 (%) 4.00 1.2MgO (%) 2.41 0.4SO3 (%) 2.94 –K2O 0.78 –Na2O 0.11 –LOI (%) 1.85 0.4

Physical properties Specific gravity 3.19 2.60Fineness (m2/kg) 328a 18,000b

a Blaine specific surface area.b BET specific surface area.

Table 3Properties of steel fibers.

Designation of thesteel fiber

Diameter D(mm)

Length L(mm)

Aspect ratio(L/D)

SF1 0.75 60 80SF2 0.75 30 40

84 E. Güneyisi et al. / Composites: Part B 56 (2014) 83–91

high-reactivity MK present comparable performance to the oneswith other mineral admixtures in terms of mechanical propertiesas well as permeability and durability properties [20–28]. More-over, the utilization of this material is also environmentallyfriendly due to the reduction of CO2 emission to the atmosphereby decreasing Portland cement consumption.

In this study, effectiveness of MK and steel fiber on mechanicalproperties of concretes with and without steel fiber was examinedthrough an experimental program. The concretes dealt with thisstudy were produced by two different water/binder (w/b) ratios.For steel fiber reinforced concretes, two different types of steelfiber with length/aspect ratios of 60/80 and 30/40 were used. Thesteel fibers were added to concrete with 0.25% and 0.75% of thevolume of the concrete. The mechanical properties of the concreteswere measured through compressive, flexural and splitting tensilestrength testing at the end of 28 days of curing. Moreover, adher-ence between reinforcing steel bar and concrete were evaluatedby means of bonding strength test at the same age. The statisticalanalysis and calculation of the contributions of the independentfactors on mechanical behavior of concretes were realized by gen-eral linear model analysis of variance (GLM-ANOVA). Additionally,the relation between mechanical properties and the bondingstrength of the concretes were evaluated through correlating theexperimental data.

2. Experimental study

2.1. Materials

CEM I 42.5 R type Portland cement having specific gravity of3.14 and Blaine fineness of 328 m2/kg was utilized for preparing

Table 2Sieve analysis and physical properties of aggregates.

Sieve size, (mm) Passing (%)

Fine aggregate

River sand

Sieve analysis 31.5 10016.0 1008.0 99.74.0 94.52.0 58.71.0 38.20.50 24.90.25 5.4Fineness modulus 2.87

Physical properties Specific gravity 2.79Absorption,% 0.55

the concrete test specimens used in determination of mechanicalproperties. The chemical composition of the cement is shown inTable 1. The metakaolin used in this study is a white powder witha Dr. Lange whiteness value of 87. It has a specific gravity of about2.60, and specific surface area (Nitrogen BET Surface Area) of18,000 m2/kg. Physical and chemical properties of MK used in thisstudy are also given in Table 1. The origin of the MK is from CzechRepublic. Fine aggregate was a mix of river sand and crushed sandwhereas the coarse aggregate was river gravel with a maximumparticle size of 22 mm. Aggregates were obtained from localsources. Properties of the aggregates are presented in Table 2.Grading of the aggregate mixture was kept constant for all con-cretes. Sulphonated naphthalene formaldehyde based high rangewater-reducing admixture with specific gravity of 1.19 wasemployed to achieve slump value of 14 ± 2 cm for the ease ofhandling, placing, and consolidation in all concrete mixtures. Thesuperplasticizer was adjusted at the time of mixing to achievethe specified slump.

Two types of commercially available hooked end steel fibers(Dramix 60/80 and Kemerix 30/40) were used for production ofsteel fiber reinforced concretes. The geometrical properties and as-pect ratios of the steel fibers are given in Table 3.

Reinforcing ribbed steel bars having 16 mm diameter and min-imum yield strength of 420 MPa were utilized for preparing thereinforced concrete specimens to be used for testing the bondingstrength.

2.2. Mix proportions

Two series of concrete mixtures with water-to-binder ratios of0.35 and 0.50 were designed to produce plain and MK incorporatedconcretes. MK modified concretes were produced by 10% replace-ment of the cement with MK by the weight. For production of steelfiber (SF) reinforced concretes, each type of SF (SF1 and SF2) wereadded to the concrete by 0.25% and 0.75% of the total concretevolume. Therefore, 20 different types of concrete mixtures wereproduced for examining the mechanical properties of the con-cretes. The details of the concrete mixtures are given in Table 4.

The designations of each mix were made according to MK incor-poration, type of steel fiber, and volume fraction of steel fiber. For

Coarse aggregate

Crushed sand No. I (4–16 mm) No. II (16–22 mm)

100 100 100100 100 27.7100 31.5 0.699.2 1.0 0.163.3 0.5 0.043.7 0.5 0.028.4 0.5 0.016.4 0.4 0.02.57 5.66 6.72

2.42 2.72 2.730.92 0.45 0.42

Table 4Mix proportions of plain and steel fiber reinforced concretes with metakaolin (kg/m3).

Mix ID w/b ratio Water Cement Metakaolin Fine aggregate Coarse aggregate Steel fiber S.Pa

Natural sand Crushed sand No. I (4–16 mm) No. II (16–22 mm) SF 1 SF 2

Control I 0.35 157.5 450 0 663.1 284.2 568.4 378.9 0 0 11.25M0-25SF1 157.5 450 0 663.1 284.2 568.4 378.9 19.62 0 12.5M0-25SF2 157.5 450 0 663.1 284.2 568.4 378.9 0 19.62 16.0M0-75SF1 157.5 450 0 663.1 284.2 568.4 378.9 58.85 0 13.75M0-75SF2 157.5 450 0 663.1 284.2 568.4 378.9 0 58.85 19.5Control II 157.5 405 45 660.0 282.9 565.7 377.2 0 0 10M10-25SF1 157.5 405 45 660.0 282.9 565.7 377.2 19.62 0 11.25M10-25SF2 157.5 405 45 660.0 282.9 565.7 377.2 0 19.62 19.5M10-75SF1 157.5 405 45 660.0 282.9 565.7 377.2 58.85 0 13M10-75SF2 157.5 405 45 660.0 282.9 565.7 377.2 0 58.85 20Control I 0.50 175.0 350 0 656.3 281.25 562.5 375.0 0 0 4.75M0-25SF1 175.0 350 0 656.3 281.25 562.5 375.0 19.62 0 4.75M0-25SF2 175.0 350 0 656.3 281.25 562.5 375.0 0 19.62 6.0M0-75SF1 175.0 350 0 656.3 281.25 562.5 375.0 58.85 0 7.5M0-75SF2 175.0 350 0 656.3 281.25 562.5 375.0 0 58.85 10.25Control II 175.0 315 35 653.9 280.22 560.4 373.6 0 0 6.25M10-25SF1 175.0 315 35 653.9 280.22 560.4 373.6 19.62 0 6.75M10-25SF2 175.0 315 35 653.9 280.22 560.4 373.6 0 19.62 12.5M10-75SF1 175.0 315 35 653.9 280.22 560.4 373.6 58.85 0 8.5M10-75SF2 175.0 315 35 653.9 280.22 560.4 373.6 0 58.85 11.0

a SP: Superplasticiser.

E. Güneyisi et al. / Composites: Part B 56 (2014) 83–91 85

example, M10-75SF1 code stands for the concrete incorporatedwith 10% MK and 0.75% steel fiber type I (SF1).

Freshly poured concrete specimens were covered with plasticsheet and kept in laboratory at 21 ± 2 �C for 24 h. Then, the speci-mens were demoulded and transferred to a water tank for curingup to 28 day.

2.3. Test specimens

The concrete specimens having various dimensions were usedfor testing. Cubic specimens having 150 � 150 � 150 mm were uti-lized for compressive strength. For three point flexural tensilestrength testing, prismatic specimens with 100 � 100 � 500 mmdimensions were used to ensure 450 mm span length for testing.Splitting tensile strength of the concrete was measured from cylin-drical specimens having £150 � 300 mm dimensions. Bondingstrength between concrete and reinforcement were tested on cubicreinforced concrete specimen. In order to have a smooth surface to

Fig. 1. Details of the bonding strength test specimen.

provide uniform load distribution, the top surface of the pulloutspecimens were capped with gypsum coating. The details anddimensions of the pullout test specimen are illustrated in Fig. 1.

For each test, three specimens were used. Each experimentalparameter was determined by averaging the results obtained fromthose specimens. All of the tests were performed at the end of28 day curing period.

2.4. Test methods

The compression test conforming to ASTM C39 [29] was carriedout on the specimens by a 3000 KN capacity testing machine.Three-point flexural tensile strength conforming to ASTM C293[30] was applied to the prismatic specimens through 100 kNcapacity bending frame. Splitting tensile strength was carried outaccording to the specification per ASTM C496 [31]. Bondingstrength of the concretes was determined in accordance with RI-LEM RC6 [32]. The bonding strength, s, is calculated by dividingthe tensile force by the surface area of the steel bar embedded inconcrete (Eq. (1)). For this test, specially modified test apparatuswas installed to 600 kN capacity universal testing machine (Fig. 2).

s ¼ Fp� d� L

ð1Þ

where F is the tensile load at failure (N), d and L are the diameter(mm) and embedment length (mm) of the reinforcing steel bar,respectively. In this study, d and L are 16 mm and 150 mm,respectively.

Scanning electron microscopy (SEM) image analysis was alsocarried out to observe the changes in cement paste matrix due toinclusion of MK. Moreover, the interface between steel fiber andmortar phase of the concrete was also observed for visual assess-ment of MK incorporation.

3. Test results and discussion

3.1. Compressive strength

Fig. 3 shows the variation in compressive strength of the plainand MK incorporated concretes with the increase in the amountof fiber reinforcement. The plain concretes’ compressive strengthvalues ranged between 62–72 MPa and 45–54 MPa for w/b ratios

Fig. 2. Photographic view of the pullout test device (a) installing the test apparatus and (b) testing the specimen.

86 E. Güneyisi et al. / Composites: Part B 56 (2014) 83–91

of 0.35 and 0.50, while MK incorporated ones had compressivestrength values between 66–76 MPa for the former and 49–59 MPa for the latter, respectively. The compressive strengthresults revealed that incorporation of MK had significant contribu-tion on the compressive strength of the concretes. Similar results

Fig. 3. Effect of steel fiber on compressive strength of (a) plain and (b) MKincorporated concretes.

have been reported by previous authors [10,23–28]. For example,in the study of Güneyisi et al. [25] concretes incorporated with5% and 15% replacement level of MK yielded relatively higherstrength than that of plain concretes at two different w/b ratios.As it can be seen from Fig. 3, increasing the amount of SF resulted

Fig. 4. Effect of steel fiber on the bond strength of (a) plain and (b) MK incorporatedconcretes.

E. Güneyisi et al. / Composites: Part B 56 (2014) 83–91 87

in rise of the compressive strength of the concretes withoutdepending on the incorporation of MK and w/b ratio. Nili and Afr-oughsabet [9] reported that 28 day compressive strengths of plainconcrete produced with w/b ratio of 0.46 were 41.3 MPa, 46.4 MPa,and 47.3 MPa for steel fiber volume fractions of 0%, 0.5%, and 1.0%,respectively. Moreover, the influence of aspect ratio was alsoclearly seen from Fig. 3. The higher the aspect ratio, the higherthe increase in compressive strength was observed, especially forMK incorporated ones. For instance, the plain concretes producedwith w/b ratio of 0.35 and steel fiber volume fraction of 0.75%had 71.4 MPa and 72.1 MPa for SF 2 and SF 1, respectively. How-ever, MK included concretes with the same parameters had72.8 MPa and 75.7 MPa for SF 2 and SF 1, respectively.

Fig. 6. Effect of steel fiber on the three-point flexural strength of (a) plain and (b)MK incorporated concretes.

3.2. Tensile strength

The tensile strength of plain and MK incorporated concreteswere monitored with respect to modulus of rupture and splittingtensile strength. The test results of flexural and splitting tensilestrength tests are presented in Figs. 6 and 7, respectively, to dem-onstrate the effectiveness of steel fiber reinforcement. The resultsrevealed that steel fiber addition provided increase in flexuralstrength capacity of plain concretes by 34.0% and 24.6% for w/b ra-tios of 0.35 and 0.50, respectively. However, MK incorporated onesexhibited 13.5% and 18.0% for those w/b ratios. It was reported thatthe main and major contribution of the steel fibers was due to theincrease of tensile strain capacity of the concrete [33]. The sametrend was also observed for splitting tensile strength of the con-cretes. The maximum splitting tensile strength values of 7.17 and5.86 MPa were observed for concretes coded M10-75SF2 for w/bratios of 0.35 and 0.50, respectively. Except for modulus of ruptureresults of MK incorporated concretes, tensile strength values of fi-ber reinforced concretes appeared to be very close to each other at0.25% volume fractions. Another noticeable finding from the tensilestrength testing is that unlike previous results, the contribution ofSF2 was observed to be better than that of SF1. This situation maybe attributed to the distribution of the steel reinforcement withinthe cement matrix. Namely, the shorter the steel fiber, the morehomogenous distribution may be achieved. Sanal and Özyurt [34]investigated the effect of orientation of steel fibers on the mechan-ical performance of the concretes. They reported that, short-cutsteel fibers have a tendency to align in the flow direction and great-er orientation density in the casting direction resulted in a greaterflexural toughness.

3.3. Bonding strength

Bonding strength of the concretes versus the amount of thesteel fiber reinforcement is plotted in Fig. 4. The figure depictedthat the increase in the volume fraction of SF resulted in great

Fig. 5. Typical failure patterns of concretes (a)

change in the bonding strength. However, the bonding strengthvalues at 0.25% volume fraction appeared to have close values,regardless the incorporation of MK. Nevertheless, introduction ofMK to the concretes imparted additional performance in terms ofbonding strength. For example, at w/b ratio of 0.35, the highestbonding strength for MK modified concretes was observed as16.9 MPa, while the minimum value for plain concrete was ob-served as 12.5 MPa. Therefore, 35% enhancement in bondingstrength capacity was accomplished by combined incorporationof MK and steel fibers. The similar trend was also observed forthe concrete group with w/b ratio of 0.5. Baran et al. [35] statedthat steel fibers improve the pull-out resistance of strands by con-trolling the crack growth inside concrete blocks. They stated that,

without steel fiber and (b) with steel fiber.

Fig. 7. Effect of steel fiber on the splitting tensile strength of (a) plain and (b) MKincorporated concretes.

88 E. Güneyisi et al. / Composites: Part B 56 (2014) 83–91

by this way, the level of confinement at the strand-concrete inter-face was increased, which resulted in improvements in both fric-tion and mechanical bond components of the resistance. Theirresults also indicated that more than 30% increase was achievedin pull-out strength due to fiber reinforcement.

Being one of the most popular mineral admixtures MK is knownto have comparable contribution to the mechanical and durabilityperformance of concretes as silica fume does [22,24,25]. However,the studies regarding the effect of inclusion of MK on the bondingstrength between concrete and steel bars has not yet attracted theadequate attention. The previous results presented for silica fume

Fig. 8. Correlation between bond strength and compressive strength.

incorporated steel fiber reinforced concretes may highlight theeffectiveness of utilization of MK for this purpose. In the study ofChan and Chu [36], the influence of silica fume on the bond prop-erties of steel fiber in matrix of reactive powder concrete (RPC)were studied. They performed pullout tests in their experimentalprogram, with the silica fume content as the primary variable. Theyindicated that the incorporation of silica fume in RCP matrixgreatly enhanced the fiber–matrix bond. Abu-Lebdeh et al. [37]also revealed that the quality of matrix has prominent importanceon the bonding and tensile strain capacity of steel fibers in highstrength concrete. Consequently, owing to its superior enhance-ment in cement matrix as a result of pore size refinement [28],MK provided improvement in the pullout capacity of the reinforcedconcretes.

Photographic views of the pullout specimens tested in the cur-rent study are given in Fig. 5. As can be seen, after failure, the rein-forcing steel bars were separated from the concretes without steelfiber, whereas steel fiber reinforced concretes did not release thesteel bars.

3.4. SEM image analysis

The change in cement paste matrix of the concrete made withw/b ratio of 0.35 due to the incorporation of MK is demonstratedin Fig. 11. As seen from Fig. 11a, there were large pores in plainconcrete while a great pore refinement was provided by 10% inclu-sion of MK into the concrete. Refinement of the pore structure ofthe concrete led to better mechanical properties mentioned above.When observing Fig. 12, it was pointed out that better interfacialtransition zone (ITZ) between steel fiber and concrete was ob-served in M10-75SF1 concrete than M0-75SF1. As a result of theimprovement of adherence between SF and concrete, toughness,ductility, and fracture energy of concrete may be enhanced.

4. Statistical evaluation of the test results

A general linear model analysis of variance (GLM-ANOVA) wascarried out at a 0.05 level of significance to examine the variationin the tested features of the fiber reinforced concretes in a quanti-tative manner. For this, compressive strength, splitting tensilestrength, flexural strength, and bonding strength of the concreteswere assigned as the dependent variables while the type of thesteel fibers used, incorporation of MK and w/b ratio were the fac-tors. A statistical analysis was performed to specify the statisticallysignificant (p-level < 0.05) factors. The contributions of the factorson the measured test results are also presented in Table 5. Thecolumn under the percent contribution provides an idea aboutthe degree of effectiveness of the independent factors on the

Fig. 9. Correlation between bond strength and modulus of rupture.

Fig. 10. Correlation between bond strength and splitting tensile strength.

E. Güneyisi et al. / Composites: Part B 56 (2014) 83–91 89

measured response such that the higher the contribution, theeffectiveness of the factors to that particular response was higher.Likewise, if the percent contribution is low, the contribution of thefactors to that particular response is less. It was observed in Table 5that all of the independent variables had significant effect on themechanical properties of fiber reinforced concretes. When observ-ing the contribution levels of the factors, it was noticed that the

Fig. 11. Refinement in pore structure of the paste matrix fo

Fig. 12. Interface between steel fiber and mortar phase of concrete (a) M0

most important parameter in variation of the compressive strengthand splitting tensile strength of the fiber reinforced concretes is w/b ratio. However, the influence of using different type of steel rein-forcement was observed to be more dominant at bonding strengthand flexural strength of the fiber reinforced concretes. Besides, theutilization of MK was also proved to be effective on all of theparameters at moderate levels.

5. Correlating between mechanical properties of the concretes

Correlating the experimental data is one of the most commonpractices among the researchers for assessment of the findings re-ported. Theoretically, the main elements controlling the mechani-cal properties of concrete are the relative volume fractions of pastematrix and aggregate, as well as their quality. As mentioned earlierhigher compressive strength reflects improved mechanicalbehavior. To evaluate the bonding strength between reinforcementand concrete, correlating other mechanical properties with thisparameter was carried out. For this, correlation between bondingstrength and other mechanical properties for both w/b contentsis respectively presented in Figs. 8–10. Based on the facts pre-sented above to specify the possible correlation between themechanical characteristics of plain and MK concretes with andwithout reinforcement, the correlation coefficients (R2) were calcu-lated and presented on those figures as well. The data used for

r w/b ratio of 0.35 (a) plain and (b) MK incorporated.

-75SF1 for w/b ratio of 0.35 and (b) M10-75SF1 for w/b ratio of 0.35.

Table 5Statistical analysis of the test result.

Dependent variable Independent variable Sequential sum of squares Computed F P Value Significance Contribution (%)

Compressive strength w/b Ratio 1304.11 459.23 0.000 YES 88.2MK replacement 54.65 19.24 0.001 YES 3.7Type of steel fiber 86.16000 30.34 0.000 YES 5.8Error 34.08000 – – – 2.3Total 1479 – – – –

Bonding strength w/b Ratio 6.4389 13.03 0.004 YES 19.8MK replacement 5.5578 11.24 0.006 YES 17.1Type of steel fiber 14.5733 29.48 0.000 YES 44.8Error 5.9315 – – – 18.2Total 32.5015 – – – –

Flexural strength w/b Ratio 2.4492 10.68 0.000 YES 24.9MK replacement 1.21 61.60 0.005 YES 12.3Type of steel fiber 4.9062 17.85 0.000 YES 49.9Error 1.2605 – – – 12.8Total 9.8259 – – – –

Splitting tensile strength w/b Ratio 4.6118 9.47 0.000 YES 68.8MK replacement 1.1503 89.43 0.000 YES 17.2Type of steel fiber 0.357 30.72 0.002 YES 5.3Error 0.581 – – – 8.7Total 6.7001 – – – –

90 E. Güneyisi et al. / Composites: Part B 56 (2014) 83–91

these figures cover the entire test results obtained. Figs. 8 and 9 re-vealed that the best correlation was achieved by polynomial curvefitting for compressive and flexural tensile strength while for split-ting tensile strength exponential curve fitting yielded the highestcorrelation, without depending on the w/b ratio. Since the scatterof the data for bonding vs. flexural tensile strength was observedto be more irregular than the others, R2 value determined was rel-atively lower. However, as a result of the noticeable differencesand uniformity for the measured values between bonding andcompressive strengths, the strongest correlation was observed totake place between these parameters.

6. Conclusions

The following conclusions may be drawn based on the experi-mental results presented above.

� Use of MK as a replacement material resulted in enhancedmechanical properties of concretes compared to plain ones forboth w/b ratios. The highest compressive strength values weremeasured as 75.7 and 58.8 MPa for concrete groups with w/bratios of 0.35 and 0.50, respectively. The inclusion of steel fibersalso contributed to the compressive strength. The long fibers(SF1) provided higher compressive strength development thanSF2 incorporated concretes with increase in volume fraction.The level of improvement was more pronounced for MK con-cretes than plain ones.� By incorporation of steel fibers remarkable improvement in

bonding and tensile strength capacities of the concretes wereobserved. The steel fibers with higher length/aspect ratio(SF1) demonstrated higher development in bonding strengthwhile the concretes incorporated with SF2 fibers had better per-formance in modulus of rupture and splitting tensile strength.This difference in the behavior of steel fiber reinforced con-cretes may be attributed to the dispersion and orientation ofthe steel fibers within the concrete.� 0.25% volume fraction of the steel fibers revealed similar trends

in contribution to the mechanical properties of the concretes forboth of the SF type used. However, for the higher volume frac-tion of steel fiber, the behavior was more distinguishable.� SEM image analysis visually proved that ITZ between steel fiber

and concrete was improved as a result of pore refinement ofconcretes by MK incorporation.

� Statistical analysis revealed that w/b ratio, type of steel fiberand incorporation of MK were all influential factors at varyinglevels on mechanical properties of the concretes. Especially,for bonding and flexural strength, the type of the steel fiberhad the greatest effect. Besides, the contribution levels of MKincorporation on the mechanical properties was observed tovary up to 18%.

References

[1] Mehta PK, Monteiro PJM. Concrete: microstructure, properties, and materials.3rd ed. USA: McGraw-Hill; 2006.

[2] Ersoy U, Özcebe G, Tankut T. Reinforced concrete. Ankara, Turkey: METU Press;2003.

[3] Bentur A, Mindess S, Diamond S. Pull out processes in steel fiber reinforcedcement. Int J Cem Compos Lightweight Concr 1985;7(1):29–38.

[4] Barros JAO, Cruz JS. Fracture energy of steel fiber-reinforced concrete. MechCompos Mater Struct 2001;8(1):29–45.

[5] Banthia N, Trottier JF. Concrete reinforced with deformed steel fibres. Part II:Toughness characterization. ACI Mater J 1995;92(2):146–54.

[6] Khayat KH, Roussel Y. Testing and performance of fiber reinforced, selfconsolidating concrete. Mater Struct 2000;33:391–7.

[7] Quian CX, Stroeven P. Development of hybrid polypropylene–steel fibrereinforced concrete. Cem Concr Res 2000;30:63–9.

[8] Yurtseven AE. Determination of mechanical properties of hybrid fibereinforced concrete. MSc thesis, METU, Ankara, Turkey; 2004.

[9] Nili M, Afroughsabet V. Property assessment of steel–fibre reinforced concretemade with silica fume. Constr Build Mater 2012;28:664–9.

[10] Güneyisi E, Gesoglu M, Mermerdas� K. Improving strength, drying shrinkage,and pore structure of concrete using metakaolin. Mater Struct2008;41:937–49.

[11] Al-Khaja WA. Strength and time-depended deformations of silica fumeconcrete for use in Bahrain. Constr Build Mater 1994;8:169–72.

[12] Brooks JJ, Megat Johari MA, Mazloom M. Effect of admixtures on the settingtimes of high-strength concrete. Cem Concr Compos 2000;22:293–301.

[13] Hooton RD. Influence of silica fume replacement of cement on physicalproperties and resistance to sulfate attack, freezing and thawing, and alkalisilica reactivity. ACI Mater J 1993;90:143–51.

[14] Chindaprasirt P, Homwuttiwong S, Sirivivatnanon V. Influence of fly ashfineness on strength, drying shrinkage and sulfate resistance of blendedcement mortar. Cem Concr Res 2004;34:1087–92.

[15] Khatri RP, Sirivivatnanon V, Yu LK. Effect of curing on water permeability ofconcretes prepared with normal Portland cement and with slag and silicafume. Mag Concr Res 1997;49:162–72.

[16] Ramezanianpour AA, Malhotra VM. Effect of curing on the compressivestrength, resistance to chloride ion penetration and porosity of concretesincorporating slag, fly ash, or silica fume. Cem Concr Compos 1995;17:125–33.

[17] Neville AM. Properties of concrete. 4th and final ed. England: Addison WesleyLogman; 1996.

[18] Kakali G, Perraki T, Tsivilis S, Badogiannis E. Thermal treatment of kaolin: theeffect of mineralogy on the pozzolanic activity. Appl Clay Sci 2001;20:73–80.

E. Güneyisi et al. / Composites: Part B 56 (2014) 83–91 91

[19] Al-Akhras NM. Durability of metakaolin concrete to sulfate attack. Cem ConcrRes 2006;36:1727–34.

[20] Coleman NJ, Page CI. Aspect of the pore solution chemistry of hydrated cementpastes containing metakaolin. Cem Concr Res 1997;27:147–54.

[21] Brooks JJ, Johari MAM. Effect of metakaolin on creep and shrinkage of concrete.Cem Concr Compos 2001;23:495–502.

[22] Boddy A, Hooton RA, Gruber KA. Long-term testing of the chloride-penetrationresistance of concrete containing high-reactivity metakaolin. Cem Concr Res2001;31:759–65.

[23] Ding JT, Li Z. Effects of metakaolin and silica fume on properties of concretes.ACI Mater J 2002;99:393–8.

[24] Güneyisi E, Mermerdas� K. Comparative study on strength, sorptivity, andchloride ingress characteristics of air-cured and water-cured concretesmodified with metakaolin. Mater Struct 2007;40:1161–71.

[25] Güneyisi E, Gesoglu M, Karaoglu S, Mermerdas� K. Strength, permeability andshrinkage cracking of silica fume and metakaolin concretes. Constr BuildMater 2012;34:120–30.

[26] Poon CS, Kou SC, Lam L. Compressive strength, chloride diffusivity and porestructure of high performance metakaolin and silica fume concrete. ConstrBuild Mater 2006;20(10):858–65.

[27] Kim HS, Lee SH, Moon HY. Strength properties and durability aspects of highstrength concrete using Korean metakaolin. Constr Build Mater2007;21:1229–37.

[28] Wild S, Khatip JM, Jones A. Relative strength, pozzolanic activity and cementhydration in superplasticized metakaolin concrete. Cem Concr Res1996;26:1537–44.

[29] ASTM C39/C39M-12. Standard test method for compressive strength ofcylindrical concrete specimens. Annual Book of ASTM Standard, Philadelphia,vol. 04–02; March 2012. 7p.

[30] ASTM C293/C293M-10. Standard test method for flexural strength of concrete(using simple beam with center-point loading). Annual Book of ASTMStandard, Philadelphia, vol. 04–02; September 2010. 3p.

[31] ASTM C496. Standard test method for splitting tensile strength of cylindricalconcrete specimens. Annual Book of ASTM Standard, Philadelphia, vol. 04–02;August 2011. 5p.

[32] RILEM RC 6. Recommendations for the testing and use of constructionsmaterials bond test for reinforcement steel. 2. Pull-out test; 1996. 3p.

[33] Kayali O, Haque MN, Zhu B. Some characteristics of high strength fiberreinforced lightweight aggregate concrete. Cem Concr Compos2003;25:207–13.

[34] Sanal _I, Özyurt N. Effects of formwork dimensions on the mechanicalperformance of fiber-reinforced cement based materials. In: 9thInternational congress on advances in civil engineering, 27–30 KaradenizTechnical University, Trabzon, Turkey; September 2010.

[35] Baran E, Akis T, Yesilmen S. Pull-out behavior of prestressing strands in steelfiber reinforced concrete. Constr Build Mater 2012;28:362–71.

[36] Chan Y, Chu S. Effect of silica fume on steel fiber bond characteristics inreactive powder concrete. Cem Concr Res 2004;34:1167–72.

[37] Abu-Lebdeh T, Hamoush S, Heard W, Zornig B. Effect of matrix strength onpullout behavior of steel fiber reinforced very-high strength concretecomposites. Constr Build Mater 2011;25:39–46.