Effect of shrinkage-reducing admixture on biaxial flexural behavior of ultra-high-performance fiber-reinforced concrete

Construction and Building Materials 89 (2015) 67–75

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Effect of shrinkage-reducing admixture on biaxial flexural behavior of ultra-high-performance fiber-reinforced concrete Doo-Yeol Yoo a, Jihwan Kim b,c, Goangseup Zi c, Young-Soo Yoon c,⇑ a

Department of Civil Engineering, The University of British Columbia, 6250 Applied Science Lane, Vancouver, BC V6T 1Z4, Canada Department of Civil and Environmental Engineering, University of Illinois at Urbana-Champaign, Urbana, IL 61801, USA c School of Civil, Environmental and Architectural Engineering, Korea University, 5-ga, Anam-dong, Seongbuk-gu, Seoul 136-713, Republic of Korea b

h i g h l i g h t s  Lower autogenous shrinkage of UHPFRC is obtained by including SRA.  The use of SRA deteriorates compressive strength and fiber pullout resistance.  The use of SRA decreases load carrying capacity and toughness of UHPFRC panel.  First cracking point of UHPFRC under biaxial flexure is suggested.  Deflection-hardening ratio and crack pattern are seldom influenced by SRA content.

a r t i c l e

i n f o

Article history: Received 14 October 2014 Received in revised form 10 March 2015 Accepted 22 April 2015

Keywords: Ultra-high-performance fiber-reinforced concrete Shrinkage-reducing admixture Compression Fiber pullout Biaxial flexure Toughness

a b s t r a c t This study aims to investigate the effect of shrinkage-reducing admixture (SRA) on the mechanical properties of ultra-high-performance fiber-reinforced concrete (UHPFRC). Three different SRA to cement weight ratios (0%, 1%, and 2%) were considered using the UHPFRC including 2% of smooth steel fibers by volume. The specimen without SRA exhibited the best performance in almost all aspects of the mechanical behaviors in compression, fiber pullout, and biaxial flexure including load carrying capacity, strain capacity, and energy absorption capacity (pullout energy and toughness). The mechanical performances deteriorated with the increase in the amount of SRA up to 2%. Finally, a suggestion was made to define the first cracking point of deflection-hardening UHPFRC under biaxial flexure stress. Ó 2015 Elsevier Ltd. All rights reserved.

1. Introduction Due to its superb performances including compressive strength, tensile strength, ductility, and durability, ultra-high-performance fiber-reinforced concrete (UHPFRC) is considered to be useful in structural applications where bending prevails. These advantages mean that the self-weight of the structural member made of UHPFRC is substantially reduced by decreasing the cross sectional area [1], and thus UHPFRC can be applied to the thin plate structures such as long span bridge decks, roofs, and thin walls [2–4]. UHPFRC develops a compressive strength exceeding 150 MPa with improved toughness and exhibits strain-hardening behavior, which exhibits a higher load carrying capacity after first cracking, ⇑ Corresponding author. Tel.: +82 2 3290 3320; fax: +82 2 928 7656. E-mail address: [email protected] (Y.-S. Yoon). http://dx.doi.org/10.1016/j.conbuildmat.2015.04.040 0950-0618/Ó 2015 Elsevier Ltd. All rights reserved.

by decreasing water-to-cementitious ratio (w/cm) and by adding high volume contents of steel fibers [5–7]. However, due to the low w/cm, UHPFRC is highly prone to early age shrinkage cracking and shows high ultimate autogenous shrinkage by approximately 800 le [8]. Therefore, research to prevent early age cracking in the manufacturing stage of UHPFRC structural members and to reduce the shrinkage strain has been conducted [6,8–13]. Due to its efficiency, many researchers have considered using a shrinkage-reducing admixture (SRA) to reduce drying and autogenous shrinkage of UHPFRC [8,9–12]. In addition, to mitigate the shrinkage cracking potential of concrete and mortar, using a combination of SRA and fibers has also been considered by several researchers [14,15]. These studies have shown that SRA degrades the surface tension of water in capillary pores, thus decreasing the magnitude of capillary stress, shrinkage, evaporation rate, and compressive strength. However, research related to the effect of SRA on the

68

D.-Y. Yoo et al. / Construction and Building Materials 89 (2015) 67–75

Table 1 Mix proportion. Relative weight ratios to cement

UH-S0 UH-S1 UH-S2

Steel fiber (Vf, %)

Cement

Water

Silica fume

Fine aggregate

Silica flour

Superplasticizer

Shrinkage-reducing admixture

1.00

0.25

0.25

1.10

0.30

0.018

0.00 0.01 0.02

2%

Where, UH = ultra-high-performance fiber-reinforced concrete, Sn = n% of shrinkage reducing admixture, Vf = volume fraction of fiber.

Table 2 Chemical compositions and physical properties of cementitious materials. Composition% (mass)

Cement

Silica fume

CaO Al2O3 SiO2 Fe2O3 MgO SO3 Specific surface (cm2/g) Density (g/cm3)

61.33 6.40 21.01 3.12 3.02 2.30 3413 3.15

0.38 0.25 96.00 0.12 0.10 – 200,000 2.10

tensile and flexural behaviors of fiber-reinforced concrete (FRC) is very limited [16]. According to Wang et al. [16], the flexural behavior of high-strength FRC containing SRA is inferior to that of high-strength FRC without SRA due to the undernourished transition zone between the fiber and the matrix. It is well known that the superb tensile properties including strength and ductility are the major reasons for using UHPFRC in the structures, and thereby the effect of SRA on the tensile and flexural properties of UHPFRC should be investigated before using it. Meanwhile, because of the excellent tensile performances of UHPFRC, research and practical applications of UHPFRC have focused on the thin plate structures [2–4]. Thin plate structures are subjected to a multi-axial stress state, rather than a uniaxial tensile stress, owing to the geometry and complex loading form [17]. However, for practical reasons, the uniaxial strength value was decided as a reference in many applications, and the most previous studies only evaluated the uniaxial tensile and flexural behaviors using dog-bone test and three- or four-point bending test [6,9,16,18–20]. The strength is not a constant parameter, but depends on the stress state [21]. Therefore, from these test methods, the structural behavior of thin plate structures cannot be directly evaluated, and it needs to assess the flexural behavior of UHPFRC under biaxial stress state. There are two different test methods available for the investigation of the biaxial flexural behavior of concrete; the flexural toughness test (ASTM C 1550) [22] and the biaxial flexure test (BFT) [17,21]. A circular plate is tested in both methods. The stress field is three-fold symmetric, and equi-biaxial only at the center, in the ASTM C 1550 test in which the specimen is supported by three pivots and loaded at its top point of the center [17]. Due to the stress distribution, specimens are fractured by three axisymmetric cracks initiating from the center of the specimen because the stress is maximum, also equi-biaxial, at the center. Unlike the ASTM C

1550 test, the stress in the BFT is equi-biaxial in the finite region enclosed by the loading ring. This feature allows taking into account the stochastic variation of the biaxial tensile strength of concrete. By the same reason that the four-point bending test is preferred over the three-point bending test in the uniaxial stress condition, the BFT is adopted in this study. Accordingly, this study investigates the effect of SRA on the mechanical behaviors of UHPFRC including compressive, fiber pullout, and biaxial flexural behaviors. The specific objectives were to investigate the effect of SRA on; (1) the compressive strength, elastic modulus, strain capacity and Poisson’s ratio, (2) the bond strength and pullout energy of the fiber embedded in the matrix, and (3) the biaxial flexural strength, toughness and crack patterns. 2. Research significance Very few investigations of the effect of SRA on the tensile and flexural behaviors of UHPFRC have been conducted, although SRA has been used for decreasing shrinkage and evaporation. In addition, even though UHPFRC is mostly applied to thin plate structures subjected to multi-axial stress, uniaxial tensile and flexural behaviors have been generally evaluated. This study therefore examines the influence of SRA on the interfacial bond properties of fiber and matrix (which profoundly influence in tensile behavior), and on the flexural properties of UHPFRC under biaxial stress state. 3. Experimental program 3.1. Materials and mix proportion The details of the mix proportion investigated in this study are given in Table 1. This optimized mix proportion was determined based on the packing density theory and the results from rheological and mechanical tests [23]. Type 1 Portland cement and silica fume (SF) were used as cementitious materials, the chemical compositions of which are listed in Table 2. Fine aggregate (silica sand) with grain size below 0.5 mm and 2 lm diameter silica flour including 98% SiO2 were also included in the mixture without coarse aggregate. In order to evaluate the effect of SRA on mechanical properties, glycol based SRA produced in Germany was applied for the tests. For all test specimens, 2 vol.% of smooth and high strength steel fibers was added, and the properties of the fibers used are given in Table 3. A high performance water-reducing agent, polycarboxylate superplasticizer (SP) with a density of 1.06 g/cm3, was also added to achieve workability. Test specimens were covered with plastic sheets immediately after concrete casting and were cured

LVDTs

Table 3 Properties of steel fiber. Type of fiber

Diameter (mm)

Length (mm)

Aspect ratio

Density (g/cm3)

Tensile strength (MPa)

Elastic modulus (GPa)

Smooth fiber

0.2

13

65

7.8

2500

200

Where, aspect ratio = length of fiber/diameter of fiber (13/0.2 = 65).

Specimen ( 100 200 mm)

Compressometer

Fig. 1. Uniaxial compression test.

69

D.-Y. Yoo et al. / Construction and Building Materials 89 (2015) 67–75

cross the crack surface [26]. To minimize the effect of surface roughness, the circular side of the specimens was capped using high strength gypsum of 2 mm thickness as shown in Fig. 3, and soft rubber pads were placed at the interface between the rings and the specimens. Each specimen was placed over the ring support, and was loaded by another smaller ring at the center of the specimen using UTM. Because the specimen is placed on the top of the annular steel ring and the soft rubber, the settlement of the support should be taken into account carefully. An annual steel frame was mounted onto the round specimen and the deflection was measured using an LVDT attached to the steel frame ring as shown in Fig. 4. In addition, to determine the first crack strength and deflection, two strain gages were attached to the bottom surface of each specimen.

Load cell LVDT

Specimen grip system

25 mm

Fiber grip system

Specimen Test machine fixed support Fig. 2. Single fiber pullout specimen and test setup.

at room temperature for the first 48 h prior to demolding. After demolding, steam curing at temperature of 90 ± 2 °C was performed for 3 days, and then the specimens were stored again in the laboratory with room temperature until testing. 3.2. Test setup and procedure 3.2.1. Uniaxial compression test A total of fifteen cylindrical specimens of 100 mm diameter and 200 mm long were produced to measure the compressive strength (ASTM C 39 [24]) by using a universal testing machine (UTM) with a maximum load capacity of 3000 kN (Fig. 1). In order to obtain the elastic modulus and strain capacity, a compressometer for the average compressive strain measurement from three linear voltage differential transformers (LVDTs) was installed. Two strain gages were attached in the circumferential direction to measure the lateral strain, and Poisson’s ratio was obtained by comparing the lateral strain with axial strain. 3.2.2. Single fiber pullout test The half-dogbone-shaped pullout specimens were fabricated to estimate the influence of SRA on the bond behavior of smooth steel fiber embedded in ultrahigh-strength matrix, identical to the UHPFRC mixture without fiber. The section of the pullout specimens used was 25  25 mm and the single fiber was embedded in the center. The embedment length of the fiber investigated was 6.5 mm, which is half the length of the fiber. The free end of the fiber was gripped tightly and a tensile load was applied as shown in Fig. 2. Based on a previous research [25], by assuming that the movement of the grip system is equal to the fiber slip and the elastic deformations of the fiber and specimen are negligible, the vertical displacement of the grip system was measured as the fiber slip using an LVDT. The pullout load was applied using UTM in displacement control at a rate of 0.018 mm/s during testing. 3.2.3. Biaxial flexure test A total of twelve round specimens with dimensions of 48 mm thickness and 420 mm diameter were used. The radius of support ring of 200 mm and the radius of loading ring of 50 mm were adopted in accordance with a previous study [17]. The specimens were fabricated by pouring concrete at the center of the form and concrete then flowed uniformly in a radial direction with high fluidity. Thus, the fibers are likely to tend to align perpendicular to the radius of the specimen and

4. Discussion on shrinkage behavior of UHPFRC SRA is mostly used to mitigate the early age shrinkage of concrete. Therefore, the effect of SRA on the shrinkage behavior of UHPFRC was first analyzed. Since UHPFRC exhibits very high autogenous shrinkage but low drying shrinkage, owing to its low w/cm and high fineness admixtures, only autogenous shrinkage was investigated according to the SRA dosage in this study. The detailed test setup and test procedures can be found in elsewhere [27]. As shown in Fig. 5, autogenous shrinkage of UHPFRC was reduced with the increase of SRA dosage; 30-day autogenous shrinkage of UHPFRC decreased by 15.2% and 28.4% by including SRA of 1% and 2%, respectively. In addition, it is interesting to notice that a significant reduction of autogenous shrinkage with the addition of SRA was observed at very early age. This is caused by the fact that since SRA degrades the surface tension of water in capillary pores, the magnitude of capillary stress is reduced. 5. Experimental results and discussion 5.1. Evaluation parameters Each compression test is described by a compressive stress– strain relationship in this study, and the value of elastic modulus is obtained from the stress–strain curve according to ASTM C 469 as follows [28].

Ec ¼

0:4  f ck  f 1 e2  0:00005

ð1Þ

where fck is the ultimate compressive strength, f1 is the stress corresponding to a longitudinal strain of 50 le, and e2 is the longitudinal strain produced by stress at 40% of fck. Poisson’s ratio in compression is associated with the axial and lateral strains, and it is obtained as

Load cell Loading ring ( 100 mm) Support ring ( 400 mm)

Specimen ( 420 48 mm) Rubber pads Steel frame

Strain gages Gypsum (with 2 mm thickness)

Fig. 3. Capping and strain gauge locations.

LVDT

Fig. 4. Test setup for BFT method.

70

D.-Y. Yoo et al. / Construction and Building Materials 89 (2015) 67–75

where df is the diameter of the fiber. The average bond strength between the fiber and the matrix at maximum pullout load is calculated by dividing the maximum pullout load by the embedment area of the fiber, as follows

sav ¼

Pmax

ð6Þ

p  df  LE

The equi-biaxial stress in BFT is given by Eq. (7) which is obtained from the linear elasticity theory [21,29]. Note that Eq. (7) is valid only in elasticity regime.

rf Fig. 5. Autogenous shrinkage of UHPFRC with different SRA ratios [27].

et2  et1 e2  0:00005

ð2Þ

where et1 is the transverse strain at the stress corresponding to the longitudinal strain of 50 le, and et2 is the transverse strain at 40% of fck. In the case of FRC, the ratio of the maximum tensile stress induced in the fiber and the fiber tensile strength is an important factor for determining the adequate use of the fiber. The maximum tensile stress induced in the fiber is related to the maximum pullout load and area of the fiber and can be obtained by Eq. (3).

rf ;max ¼

Pmax Af

ð3Þ

where Pmax is the maximum pullout load and Af is the area of the fiber. The energy absorption capacity of FRC is closely related to the fiber pullout energy. The pullout energy defined by the area of the hysteresis loop in the load-slip curve up to complete pullout is affected by the pullout load and slip capacities between the fiber and the matrix and can be obtained as

Z

The average uniaxial compressive stress–strain responses of UHPFRC with various amounts of SRA are illustrated in Fig. 6. The average stress–strain curve was obtained from the test results of five specimens. In comparison with normal- and high-performance concretes, UHPFRC shows a very linear compressive stress–strain relationship up to failure regardless of the percentage of SRA. Linearity was determined using the ratio between the elastic modulus Ec calculated by Eq. (1) and the secant modulus for the strain at peak load E0 [30]. As the linearity of concrete increases, a lower ratio Ec/E0 is obtained. The ratio Ec/E0 is approximately 3.5 for concrete with 7 MPa compressive strength and 1.25 for concrete with 70 MPa compressive strength [30,31]. A ratio Ec/E0 of UHPFRC of approximately 1.1 was obtained in this study, regardless of the SRA amount. This implies that a higher linearity

s¼LE

PðsÞds

ð4Þ

s¼0

where s is the fiber slip during pullout, LE is the initial embedment length of the fiber, and P(s) is the pullout load at any slip. By assuming that the bond strength is equivalent over the entire embedment length of fiber, the equivalent bond strength can be computed from the pullout energy [18]. Even if the maximum pullout load is equal to each specimen, the equivalent bond strength can differ in terms of the pullout energy and is expressed by Eq. (5).

seq ¼

5.2. Compressive behavior

2  WP

(a)

ð5Þ

p  df  L2E

60

250 Compressive strength Elastic modulus

225

55

200

50

175

45

Elastic modulus (GPa)

WP ¼

ð7Þ

where P is the applied load, h is the thickness of the specimen, m is the Poisson’s ratio, R is the radius of the specimen, a and b are the radius of the support ring and radius of the loading ring, respectively.

Compressive strength (MPa)

l¼

  8 9 2 2 3P <ð1  mÞ a  b a= ¼ þ ð1 þ mÞ ln 2 b; 2R2 2ph :

40

150 0%

1%

2%

Amount of SRA (%) 250

150 100

0.2

0.15

50 0

0.3

0.25 Poisson's ratio

200

Stress (MPa)

(b)

UH-S0 UH-S1 UH-S2

0

0.001

0.002

0.003

0.004

0.005

Strain (mm/mm)

0.1 0%

1%

2%

Amount of SRA (%) Fig. 6. Average compressive stress–strain curve of UHPFRC with three different SRA to cement weight ratios.

Fig. 7. Effect of SRA on compressive strength, elastic modulus and Poisson’s ratio.

71

D.-Y. Yoo et al. / Construction and Building Materials 89 (2015) 67–75

96

3000

72

2250

48

1500 Ave. of UH-S0

24

0

750

0

2

4

6 Slip (mm)

8

10

Pullout stress (MPa)

Pullout load (N)

Fig. 8. Failure patterns in compression (UH-S0).

0

Fig. 9. Typical pullout load-slip (stress-slip) relationship of UH-S0.

is obtained for UHPFRC than for concretes with compressive strengths of 7 and 70 MPa. Based on a previous study [11], the compressive strength of UHPFRC was slightly reduced by adding SRA. As shown in Fig. 7(a), the highest compressive strength was obtained by 200.14 MPa for UH-S0. This value is 7.2% and 2.7% higher than those of UH-S1 and -S2, respectively. The average elastic modulus calculated by Eq. (1) for UH-S0 and -S2 showed similar values of approximately 51.50 GPa, which is higher than that of UH-S1. The strain at the peak load of UH-S0 was obtained by 0.0043 mm/mm, and it was slightly reduced by adding SRA. On the contrary, a higher Poisson’s ratio of approximately 0.197 was obtained for UH-S1 and -S2 than 0.192 for UH-S0 (Fig. 7(b)). In general, UHPFRC failed by lateral tensile expansion under uniaxial compressive load, and the lateral expansion is confined by the steel fibers. Even when the test specimens were failed in a brittle manner with a rapid load decrease after the peak load, insignificant fragmentation was observed in contrast to the performance of ultra-high-strength concrete without fiber as shown in Fig. 8.

results of five specimens. The peak pullout load of each test was within 16% of the average value. The average values of pullout parameters are synthetically summarized in Table 4. In addition, the average pullout load (or stress) versus slip relationship in terms of the weight ratio of SRA is shown in Fig. 10. Since identical smooth steel fiber was used, the shape of pullout load-slip curve was similar for each specimen. The typical pullout behavior was characterized by a steep increase followed by a gradual decrease of fiber pullout stress with an increased slip. The following three factors can explain why the pullout load was slowly reduced until complete pullout [25]; (1) the additional mechanical bond due to the cutting process, (2) the wedge effect and abrasion of fine adherent components of the matrix, and (3) the damaging and scratching of the fiber coating. Test results also indicated that the pullout resistance was significantly reduced with the addition of SRA and with the increase in the percentage of SRA. This is because the radial confinement pressure that results in the bond between the fiber and the matrix is reduced due to the decreased shrinkage by using SRA (see Fig. 5). Fig. 11 provides the maximum fiber tensile stress according to the SRA content. The highest maximum fiber stress rf,max of 981.12 MPa was obtained for UH-S0. For UH-S1 and -S2, the maximum fiber stresses of 815.37 and 643.67 MPa were obtained, respectively. The ratios of the maximum fiber tensile stress obtained from fiber pullout and the tensile strength of fiber (ft = 2500 MPa) varied from 0.26 to 0.39. As shown in Fig. 12(a), the maximum pullout load Pmax (bond strength sav) of 30.82 N (sav = 7.55 MPa) for UH-S0 was obtained. Due to the addition of SRA, the pullout load was decreased by 16.9% and 34.4% for UH-S1 and -S2, respectively. The highest equivalent bond strength seq of 9.26 MPa was also observed for UH-S0 (Fig. 12(b)), and the equivalent bond strength was reduced by 27.3% and 40.9% for UH-S1 and -S2, respectively. From the analysis of the pullout test results, it is obvious that the bond

96

3000

72

2250

Table 4 Summary of pullout parameters.

UH-S0 UH-S1 UH-S2

rf,max (MPa)

sav (MPa)

Wp (N mm)

seq (MPa)

981.12 (134.263) 815.37 (160.385) 643.67 (99.737)

7.55 (1.033) 6.27 (1.234) 4.95 (0.767)

122.86 (25.858) 89.30 (26.512) 72.56 (20.860)

9.26 (2.567) 6.73 (2.380) 5.47 (1.117)

Where, rf,max = maximum pullout stress, sav = average bond strength, Wp = pullout energy, seq = equivalent bond strength, (x.xxx) = standard deviation.

48

1500

UH-S0 UH-S1

24

UH-S2

750

0

Pullout stress (MPa)

Fig. 9 shows the typical pullout load-slip (pullout stress-slip) behaviors of UH-S0. Average result was obtained based on the test

Pullout load (N)

5.3. Single fiber pullout behavior

0 0

2

4 6 Slip (mm)

8

10

Fig. 10. Effect of SRA on pullout behavior of smooth steel fiber embedded in UHPFRC mortar.

72

D.-Y. Yoo et al. / Construction and Building Materials 89 (2015) 67–75

120

0.6

UH-S0

96

0.4

Load (kN)

1000

σf, max/ft

Max. fiber tensile stress (MPa)

1500

500

UH-S1

72 UH-S2

48

0.2 24

0

0%

1%

2%

0

0

0

3

Amount of SRA (%)

6 Deflection (mm)

9

12

Fig. 14. Average biaxial flexural load–deflection curve of UHPFRC according to the amount of SRA.

48

12

5.4. Biaxial flexural behavior of UHPFRC

36

9

24

6

12

3

5.4.1. Load versus mid-span deflection response The load–deflection curves of all test specimens under biaxial flexure are illustrated in Figs. 13 and 14. Four test specimens were used to obtain the average load–deflection curve. The details of the test results are summarized in Table 5, which provides the evaluated parameters regarding the biaxial flexural response of UHPFRC at various amounts of SRA. The peak load of each test was within 10% of average value for all test series in Fig. 13. Thus, it was concluded that the use of four specimens per series was adequate for the evaluation of biaxial bending performance. All test series exhibited deflection-hardening behavior and similar load–deflection responses. The UH-S0 specimen exhibited the highest load carrying capacity, but showed the lowest post-peak ductility in the softening region. The load carrying capacity decreased and the post-peak ductility increased as a higher amount of SRA was used. Similar observation was obtained for the case of direct tensile behaviors of UHPFRC [27]; lower tensile strength and strain capacity were obtained for the specimens with higher SRA contents. An accurate evaluation of the stress and deflection at first crack is of paramount importance. Therefore, to determine the first cracking point in this study, two strain gages were attached to the center of the bottom face as shown in Fig. 3, and the load– strain relationship was compared with the load–deflection curve. The first cracking point was defined as the point where nonlinearity in the load–deflection curve was obtained. This point is named the limit of proportionality (LOP) from the previous ASTM C 1018 [32]. As shown in Fig. 15(a), in the case of biaxial stress state, LOP was clearly described and this point was defined as the first cracking point. In comparison with the load–strain curve as shown in Fig. 15(b), the LOP obtained from the load–deflection curve corresponded to the point where the strain was sharply increased

0

Bond strength (MPa)

(a) Max. pullout load (N)

Fig. 11. Effect of SRA on the maximum fiber tensile stress.

0 1% Amount of SRA (%)

2%

15

160

12

120

9

80

6

40

3

0

0 0%

1%

2%

Amount of SRA (%) Fig. 12. Average and equivalent bond strengths according to SRA content; (a) average bond strength, sav, (b) equivalent bond strength, seq.

properties of the smooth steel fiber embedded in ultra-highstrength matrix are significantly affected by the amount of SRA, and especially, the fiber pullout resistance was deteriorated with the increase in the amount of SRA.

(c) 120

96

96

96

72

Load (kN)

(b) 120

Load (kN)

(a) 120

Ave. of UH-S0

48 24

Load (kN)

Pullout work (Nmm)

(b) 200

Equ. bond strength (MPa)

0%

72 Ave. of UH-S1

48 24

0 3 6 9 Deflection (mm)

12

Ave. of UH-S2

48 24

0

0

72

0 0

3 6 9 Deflection (mm)

12

0

3 6 9 Deflection (mm)

Fig. 13. Effect of SRA on biaxial flexural behavior of UHPFRC; (a) UH-S0, (b) UH-S1, (c) UH-S2.

12

73

D.-Y. Yoo et al. / Construction and Building Materials 89 (2015) 67–75 Table 5 Summary of parameters derived from biaxial flexure test results.

PLOP fLOP dLOP ToughLOP

Unit

UH-S0

UH-S1

UH-S2

kN MPa mm kN mm

51.43 (10.472) 21.27 (4.332) 0.25 (0.107) 8.27 (5.643)

41.57 (6.490) 17.24 (2.691) 0.20 (0.071) 4.59 (2.449)

38.52 (6.040) 15.97 (2.504) 0.16 (0.036) 3.18 (1.187)

90

Load (kN)

LOP

(a) 120

kN 97.10 (4.726) 85.89 (7.015) 79.41 (4.015) MOR PMOR dMOR mm 3.79 (0.295) 3.70 (0.985) 4.10 (0.697) ToughMOR kN mm 314.62 (38.339) 269.16 (66.822) 287.46 (65.706) kN 94.30 (5.894) 83.40 (7.298) Pd2.5 mm 2.50 (0.000) 2.50 (0.000) dd2.5 Toughd2.5 kN mm 192.77 (13.737) 172.52 (9.336)

d5

Pd5 dd5 Toughd5

kN 94.89 (4.072) 83.85 (7.237) 78.12 (4.781) mm 5.00 (0.000) 5.00 (0.000) 5.00 (0.000) kN mm 431.01 (24.706) 383.55 (27.215) 356.91 (17.871)

Pd10 dd10 Toughd10

kN 73.40 (3.617) 66.855 (7.949) 65.42 (7.663) mm 10.00 (0.000) 10.00 (0.000) 10.00 (0.000) kN mm 860.22 (38.978) 764.35 (59.265) 721.43 (50.202)

d10

0

77.19 (4.057) 2.50 (0.000) 162.80 (7.938)

Where, P = applied load, f = biaxial flexural stress, d = mid-span deflection, Tough = toughness, (x.xxx) = standard deviation.

(a) 120

60

30

PLOP

(b) 1000 Toughness (kNmm)

d2.5

UH-S0 UH-S1 UH-S2

800

Pd2.5

PMOR

Pd5

Pd10

Toughd2.5

ToughMOR

Toughd5

Toughd10

UH-S0 UH-S1 UH-S2

600 400 200 0 ToughLOP

Load (kN)

96 Fig. 16. Effect of SRA on load and toughness of UHPFRC panels; (a) applied load, (b) toughness.

72 48 24

Limit of Proportionality (LOP)

0 0

0.1

0.2

0.3

0.4

0.5

Deflection (mm)

(b) 120

Strain gage 1 Strain gage 2

Load (kN)

96 72 48

Fig. 17. Effect of SRA on the ratio between PMOR and PLOP.

24

LOP

0 0

200

400 Strain (με)

600

800

Fig. 15. Typical initial biaxial flexural behavior of UH-S0; (a) load–deflection curve, (b) load–strain curve.

without any change of load, indicating the initiation of flexural crack.

5.4.2. Load carrying capacity and energy absorption capacity (toughness) Flexural response of FRC is classified as either deflection-softening or deflection-hardening [33]. UHPFRC exhibits deflection-hardening behavior under biaxial flexure, thus both of the points at LOP and modulus of rupture (MOR) are basically required to evaluate the load carrying capacity and energy absorption capacity, based on toughness value. Besides the points at LOP and MOR, three other deflection points are also defined as follows; L/160 (d2.5), L/80 (d5), and L/40 (d10), similar to those suggested by ASTM C 1550 [22]. The reason for their inclusion, especially for d5 and d10, is

that UHPFRC exhibits superior load carrying capacity and toughness at large deflections. The influence of SRA on the load carrying capacity is shown in Fig. 16(a) and summarized in Table 5. At all points, the highest loads were obtained for UH-S0, and the magnitude of load was reduced with increasing the SRA content. For example, the highest first crack and peak loads (PLOP and PMOR) of UH-S0 were obtained by 51.43 and 97.10 kN, and these were reduced by 19.2% and 11.5% for UH-S1 and 25.1% and 18.2% for UH-S2, respectively. Fig. 16(b) exhibits the effect of SRA on the energy absorption capacity of UHPFRC panels using toughness value, which is defined by the area of the hysteresis loop in the biaxial flexural load–deflection curve up to a certain point. Since the energy absorption capacity is an important parameter for the structures subjected to dynamic and extreme loads, such as seismic, impact and blast [34], comparing energy absorption capacity can provide useful information for such applications. The toughnesses at all points were reduced with increasing the amount of SRA, because the load carrying capacity was decreased (see Fig. 16(a)). For example, the highest toughness at d10 was obtained by 860.22 kN mm for UH-S0, and this was decreased by 11.1% and 16.1% for UH-S1 and UH-S2, respectively. Based on these test results, it was noted that

74

D.-Y. Yoo et al. / Construction and Building Materials 89 (2015) 67–75

Fig. 18. Failure patterns of UHPFRC panels; (a) UH-S0, (b) UH-S1, (c) UH-S2.

the inclusion of SRA provides negative effect on the biaxial flexural performance of UHPFRC panels including load carrying capacity and toughness. 5.4.3. Deflection-hardening ratio (PMOR > PLOP) and cracking behavior Fig. 17 shows the effect of SRA on the ratio between PMOR and PLOP. The deflection-hardening response can be obtained when the ratio is larger than 1.0. The average ratio was barely affected by the amount of SRA, ranged from 1.89 to 2.07. This means that UHPFRC including 2 vol.% of smooth steel fibers provide deflection-hardening behavior under biaxial stress irrespective of the amount of SRA. The multiple cracking behavior and failure pattern of test specimens under biaxial flexure are shown in Fig. 18. All test specimens produced three or four major cracks with multiple micro-cracks. In addition, the crack pattern and failure mode were not influenced by the SRA amount. The first crack was initiated inside the loading ring, where uniform maximum stress occurs, and then propagated to the near support ring with increasing crack width. 6. Conclusions The effect of SRA on the compressive, fiber pullout and biaxial flexural behaviors of UHPFRC were evaluated in this study. Based on the above results, the following conclusions can be drawn: (1) From the results of the uniaxial compression test, compressive strength, elastic modulus and strain capacity were slightly reduced by adding SRA, whereas Poisson’s ratio was increased. All test specimens showed very linear compressive stress–strain response (Ec/E0 = 1.1) and were failed in a brittle manner with an insignificant fragmentation due to the confinement of fibers. (2) The fiber pullout resistance was significantly reduced by adding SRA, since the radial confinement pressure generating the bond between the fiber and the matrix was reduced by decreasing shrinkage. The highest average bond strength was obtained by 7.55 MPa for UH-S0 and it was reduced by 16.9% and 34.4% for UH-S1 and -S2, respectively. The equivalent bond strength of UH-S0 was obtained by 9.26 MPa, which was 27.3% and 40.9% higher than those of UH-S1 and -S2, respectively. (3) All the test specimens showed deflection-hardening behavior under biaxial flexural stress, regardless of the SRA content, and produced three or four major cracks with multiple micro-cracks. The definition of first cracking point of UHPFRC under biaxial flexure was proposed as the point where nonlinearity in the load–deflection curve was obtained. The highest load carrying capacity and toughness

at all points (i.e., LOP, d2.5, MOR, d5, d10) were obtained for UH-S0, and these values were all decreased with increasing the SRA content. (4) From these test results, it was concluded that even though SRA provides positive effect on the reduction of shrinkage, SRA should be carefully used because it also gives negative effect on the mechanical properties of UHPFRC.

Acknowledgments This research was supported by a Grant from a Construction Technology Research Project 13SCIPS02 (Development of impact/ blast resistant HPFRCC and evaluation technique thereof) funded by the Ministry on Land, Infrastructure, and Transport.

References [1] Federal Highway Administration (FHWA). Material property characterization of ultra high performance concrete. Federation Highway Administration Report, US Department of Transportation, VA; 2006. [2] di Prisco M, Lamperti M, Lapolla S, Khurana RS. HPFRCC thin plates for precast roofing. In: Proceedings of the second international symposium on ultra high performance concrete. Kassel, Germany; March, 2008. p. 675–82. [3] Saleem MA, Mirmiran A, Xia J, Mackie K. Ultra-high-performance concrete bridge deck reinforced with high-strength steel. ACI Struct J 2011;108(5): 601–9. [4] Kobler M, Sobek W. The introduction of high forces into tine-walled UHPC elements by the use of implants. In: Proceedings of the second international symposium on ultra high performance concrete. Kassel, Germany; March, 2008. p. 683–90. [5] Wille K, Kim DJ, Naaman AE. Strain-hardening UHP-FRC with low fiber contents. Mater Struct 2011;44(3):583–98. [6] Yoo DY, Park JJ, Kim SW, Yoon YS. Early age setting, shrinkage and tensile characteristics of ultra high performance fiber reinforced concrete. Const Build Mater 2013;41. 427–238. [7] Kang ST, Kim JK. The relation between fiber orientation and tensile behavior in an ultra high performance fiber reinforced cementitious composites (UHPFRCC). Cem Concr Res 2011;41(10):1001–14. [8] Koh KT, Ryu GS, Kang ST, Park JJ, Kim SW. Shrinkage properties of ultra-high performance concrete (UHPC). Adv Sci Lett 2011;4(3):948–52. [9] Park JJ, Yoo DY, Kim SW, Yoon YS. Drying shrinkage cracking characteristics of ultra-high-performance fibre reinforced concrete with expansive and shrinkage reducing agents. Mag Concr Res 2013;65(4):248–56. [10] Cheung AKF, Leung CKY. Shrinkage reduction of high strength fiber reinforced cementitious composites (HSFRCC) with various water-to-binder ratios. Cem Concr Compos 2011;33(6):661–7. [11] Soliman AM, Nehdi ML. Effect of drying conditions on autogenous shrinkage in ultra-high performance concrete at early-age. Mater Struct 2011;44(5): 879–99. [12] Yoo DY, Park JJ, Kim SW, Yoon YS. Combined effect of expansive and shrinkagereducing admixtures on the properties of ultra high performance fiberreinforced concrete. J Compos Mater 2014;48(16):1981–91. [13] Yoo DY, Park JJ, Kim SW, Yoon YS. Influence of ring size on the restrained shrinkage behavior of ultra high performance fiber reinforced concrete. Mater Struct 2014;47(7):1161–74.

D.-Y. Yoo et al. / Construction and Building Materials 89 (2015) 67–75 [14] Passuello A, Moriconi G, Shah SP. Cracking behavior of concrete with shrinkage reducing admixtures and PVA fibers. Cem Concr Compos 2009;31(10): 699–704. [15] Hwang SD, Khayat KH. Effect of mixture composition on restrained shrinkage cracking of self-consolidating concrete used in repair. ACI Mater J 2008;105(5):499–509. [16] Wang JY, Banthia N, Zhang MH. Effect of shrinkage reducing admixture on flexure behaviors of fiber reinforced cementitious composites. Cem Concr Compos 2012;34(4):443–50. [17] Kim J, Kim DJ, Zi G. Improvement of the biaxial flexure test method for concrete. Cem Concr Compos 2013;37:154–60. [18] Kim DJ, El-Tawil S, Naaman AE. Correlation between single fiber pullout behavior and tensile response of FRC composites with high strength steel fiber. In: Cachan SARL, editor. Proceedings of the fifth international workshop of high performance fiber reinforced cementitious composites, RILEM proceedings, Pro. 53. France; July 2007. p. 67–76. [19] Kanakubo T. Tensile characteristics evaluation method for ductile fiber-reinforced cementitious composites. J Adv Concr Technol 2006;4(1): 3–17. [20] Yoo DY, Lee JH, Yoon YS. Effect of fiber content on mechanical and fracture properties of ultra high performance fiber reinforced cementitious composites. Compos Struct 2013;106:742–53. [21] Zi G, Oh H, Park SK. A novel indirect tensile test method to measure the biaxial tensile strength of concretes and other quasibrittle materials. Cem Concr Res 2008;38(6):751–6. [22] ASTM C 1550-12a. Standard test method for flexural toughness of fiberreinforced concrete (using centrally loaded round panel). American Society of Testing and Materials; 2012. [23] Park JJ, Kang ST, Koh KT, Kim SW. Influence of the ingredients on the compressive strength of UHPC as a fundamental study to optimize the mixing

[24] [25] [26]

[27]

[28]

[29] [30] [31] [32]

[33] [34]

75

proportion. In: Second international symposium on ultra high performance concrete. Kassel, Germany; 2008. p. 105–12. ASTM C 39. Standard test method for compressive strength of cylindrical concrete specimens. American Society of Testing and Materials; 2012. Wille K, Naaman AE. Pullout behavior of high-strength steel fibers embedded in ultra-high-performance concrete. ACI Mater J 2012;109(4):479–87. Barnett SJ, Lataste JF, Parry T, Millard SG, Soutsos MN. Assessment of fibre orientation in ultra high performance fibre reinforced concrete and its effect on flexural strength. Mater Struct 2010;43(7):1009–23. Yoo DY. Performance enhancement of ultra-high-performance fiber-reinforced concrete and model development for practical utilization [Ph.D. thesis]. Seoul, Korea: Korea University; 2014. 586 p. ASTM C 469M-10. Standard test method for static modulus of elasticity and Poisson’s ratio of concrete in compression. American Society of Testing and Materials; 2010. Timoshenko SP, Woinowsky-Krieger S. Theory of plates and shells. 2nd ed. New York: McGraw-Hill; 1959. Graybeal BA. Compressive behavior of ultra-high-performance fiberreinforced concrete. ACI Mater J 2007;102(2):146–52. Popovics S. Strength and related properties of concrete. New York: John Wiley & Sons, INC.; 1998. 522 p. ASTM C 1018-97. Standard test method for flexural toughness and first-crack strength of fiber reinforced concrete (using beam with third-point loading). American Society of Testing and Materials; 1998. Naaman AE, Reinhardt HW. Proposed classification of HPFRC composites based on their tensile response. Mater Struct 2006;39(5):547–55. Yoo DY, Yoon YS. Influence of steel fibers and fiber reinforced polymers on the impact resistance of one-way concrete slabs. J Compos Mater 2014;48(6): 695–706.