You are on page 1of 9

Construction and Building Materials 28 (2012) 57–65

Contents lists available at SciVerse ScienceDirect

Construction and Building Materials


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

Investigation on the flexural behavior of UHPCC considering the effect of fiber


orientation distribution
Su-Tae Kang a,⇑, Jin-Keun Kim b
a
Department of Civil Engineering, Daegu University, 201, Daegudae-Ro, Gyeongsan, Gyeongbuk 712-714, South Korea
b
Department of Civil and Environmental Engineering, Korea Advanced Institute of Science and Technology (KAIST), 335, Gwahangno, Yuseong-gu, Daejeon 305-701, South Korea

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

Article history: This study aims to investigate the effect of fiber orientation distribution on the flexural behavior of ultra
Received 7 February 2011 high performance cementitious composites (UHPCC) and to propose an analytical approach which
Received in revised form 26 May 2011 enables to predict the flexural behavior considering probabilistic fiber orientation distribution. A
Accepted 6 July 2011
three-point bending test with the notched beams was carried out and the fiber orientation distribution
Available online 1 October 2011
was quantitatively estimated by the help of image analysis process. The measured fiber orientation dis-
tributions for two different flexural performances confirmed that the fiber orientation distribution has a
Keywords:
strong impact on the deflection hardening behavior in bending. Finite element analyses were performed
Ultra high performance
Cementitious composites
to predict flexural behavior of UHPCC considering the difference in fiber bridging behavior depending on
Steel fiber the fiber orientation distribution. The analytical results were in good agreement with the experimental
Fiber orientation results.
Flexural behavior Ó 2011 Elsevier Ltd. All rights reserved.

1. Introduction [18,19]. Wuest et al. [18] developed a meso-mechanical model to


predict the tensile response as a function of the volume, aspect ra-
Up to date, there have been extensive efforts to improve the tio, distribution and orientation of the fibers and the mechanical
strength of concrete; however, as an inevitable result accompany- properties of the matrix. Kang and Kim [19] also suggested an ana-
ing increased strength, brittle failure has limited the application of lytical model to predict the tensile behavior based on the fiber
high strength concrete. The incorporation of fibers into the higher bridging considering the fiber orientation effect after cracking.
strength of matrix can be employed to overcome the problem of Meanwhile, although the direct tension tests to investigate the uni-
brittleness by providing high energy absorption capacity and duc- axial tensile behavior of high performance cementitious compos-
tile behavior via fiber bridging. Even for ultra high performance ites such as UHPCC in this study are recently increasing, the
cementitious composites (UHPCC) matrix with a compressive worldwide standards for testing are still specifying the bending
strength of 200 MPa, incorporating steel fibers can guarantee suffi- test with beam specimens as a standard test method to evaluate
cient ductility of the composites. of the various tensile properties of cementitious composites.
The tensile behavior of fiber reinforced concrete or composites Therefore, in order to estimate the experimental results with bend-
is greatly influenced by the fiber characteristics, such as shape, ing test thoroughly, it is also necessary to make clear how the
amount, and orientation [1–8]. In particular, even when the com- influencing factors have an effect on the flexural behavior. This
posites include steel fibers having the same shape and in the same study aims to investigate the effect of fiber orientation distribution
amounts, variation in the fiber orientation distribution has a strong on the flexural behavior of UHPCC and to propose an analytical ap-
impact on their tensile behavior [9–14]. The influence is applied to proach which enables to predict the flexural behavior according to
UHPCC in the same way. In consequence, the structural perfor- the fiber orientation distribution.
mance of an UHPCC member is expected to be also dependent on
the fiber orientation distribution. It is therefore necessary to clarify
2. Modeling approach for uniaxial tensile behavior of UHPCC
the effect of fiber orientation on the tensile behavior of UHPCC.
There have been several studies on the effect of fiber orientation
It is commonly recognized for quasi-brittle materials such as
distribution on the uniaxial tensile behavior for UHPCC [15–17]
concrete that the tensile behavior is dominated by localization
and proposals to predict the behavior considering its effect
after cracking. Such behavior can be found in UHPCC, of which
the behavior quite differs from that of strain hardening cementi-
⇑ Corresponding author. Tel.: +82 53 850 6528; fax: +82 53 850 6529. tious composites (SHCC) presenting the increase in stress with
E-mail address: stkang@daegu.ac.kr (S.-T. Kang). numerous multiple cracking in uniaxial tension.

0950-0618/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.conbuildmat.2011.07.003
58 S.-T. Kang, J.-K. Kim / Construction and Building Materials 28 (2012) 57–65

Before any cracking has taken place in the matrix, during the
early stages of loading, the tensile behavior can be described in
terms of stress–strain relation. The elastic shear transfer at the
interface between the fiber and the matrix is the major mechanism
to be considered for predicting the pre-cracking tensile behavior of
fiber reinforced composites [20]. The tensile stress of the compos-
ites is calculated based on the rule of mixture, expressed by Eq. (1),
as a function of the strain ec in the composites.

rc ðec Þ ¼ rm ðec ÞV m þ gl gh rf ðec ÞV f ð1Þ


In Eq. (1), rf(ec) represents the tensile stress in the fiber at a
strain of ec and rm(ec) is for the matrix. Vm and Vf indicate the vol-
ume fraction of the matrix and the fiber, respectively. gl represents
the length efficiency factor in the pre-cracking case for frictional
stress transfer mechanism. gh means the fiber orientation coeffi-
cient, which depends upon the length of fiber, the effect of friction
on the interface between the matrix and the fiber, etc. However, for
simplicity, if the geometrical arrangement of fiber is only consid-
ered; the fiber orientation coefficient can be defined as given in
Eq. (2) [21].
Z hmax
gh ¼ pðhÞ cos2 h dh ð2Þ
hmin

where p(h) is a probability density distribution for the fiber orienta-


tion and h represents the angle between each fiber axis and the ten-
sile stress direction. It is found that gh has the value of 0.5 for a
random 2-dimensional arrangement and 1/3 for a random 3-dimen-
sional arrangement.
After cracking, the localized behavior is normally expressed by Fig. 1. Probability density distribution in 3-D randomly oriented fiber distribution
the load versus crack opening relation. With regard to fiber rein- [22].
forced composites, the localized behavior can be obtained by com-
bining the fiber bridging behavior and the matrix tension softening bilinear softening curve as a resisting portion by the UHPCC matrix
behavior after cracking. when the localized behavior of UHPCC is intended to be predicted
The fiber bridging behavior is determined by summing up the [19]. Their proposal was introduced for the tension softening curve
pullout resistance of the individual fiber at the crack plane, when in this study. For the matrix in UHPCC, the stress versus crack
if assuming that all fibers are pulled out without breakage. When opening displacement in the cracked plane can be taken as
the pullout resistant force of a single fiber at the cracked plane is  
w
given as a function of the inclined angle of the fiber (h), the embed- rct ¼ ft 1  0:7 for 0:3f t 6 rct 6 ft ð4Þ
wc
ded length of the fiber (le) and the crack opening displacement (d),
which is denoted as P(h, le, d), the bridging stress of the composites
0:3f t
may be obtained from Eq. (3) [22]. rct ¼ ðwc  wÞ for 0 6 rct 6 0:3f t ð5Þ
wc  w1
Z p Z lf
4V f 2 2
w1 ¼ 0:2 mm; and wc ¼ 0:5 mm ð6Þ
rb ðdÞ ¼ 2
Pðh; le ; dÞpðle ÞpðhÞ cos h dle dh ð3Þ
p df 0 0
where ft is the tensile strength, w is the cracking opening (mm), and
where p(h) and p(le) are probability density functions for h and le rct is the tensile stress at a given crack opening. w1 is the cracking
respectively. For three-dimensional random distribution of fibers, opening (mm) at rct = 0.3ft, wc is the cracking opening (mm) at
p(h) is equal to sin h (see Fig. 1) and p(le) is given as a function of rct = 0.
fiber embedded length and inclination angle. lf means the length
of fiber. 3. Experimental program
Lee et al. [23] proposed an analytical model to predict the pull-
out behavior of inclined steel fiber in an ultra high strength cemen- The mix design of UHPCC differs significantly from that of nor-
titious matrix. Both the matrix and steel fiber used in their study mal concrete. UHPCC mix proportion applied in this study is given
have exactly the same properties as those in this study. Their mod- in Table 1. The water–binder ratio is determined as w/b = 0.2. Sand
el was applied for the pullout resistance of a single fiber, P(h, le, d). with grain size below 0.5 mm is used as fine aggregate, and coarse
At this moment, it should be noted that the bridging force obtained aggregates are not included. Superplasticizer has been introduced
from a pullout test may be a little different from that in a compos- to secure the required workability for a low water to binder ratio
ites, which is mainly caused by the difference in the bond condition composition. In addition, the filler with mean grain size of about
[19,24]. Therefore, additionally in order to consider the difference, 4 lm was used to enhance the density of UHPCC. 2 vol.% of straight
the pullout resistance is multiplied by a correction factor. steel fibers with 13 mm length and 0.2 mm diameter were incor-
As mentioned earlier, besides the fiber bridging, the contribu- porated. The property in a fresh state was evaluated by a flow test
tion of the matrix should be introduced in order to establish the (ASTM C1437) and the flowability was measured to be about
post-cracking localized behavior. However, unfortunately little 230 mm.
information is available to describe the tension softening behavior Experiments on the effect of fiber orientation on the flexural
of the UHPCC matrix after cracking. The authors proposed a behavior of the composites were carried out, and those were
S.-T. Kang, J.-K. Kim / Construction and Building Materials 28 (2012) 57–65 59

Table 1 distribution were induced by adopting different placing proce-


Mix design of UHPCC. dures (see Fig. 2). Two cases were considered for the placement
Relative weight ratios to cement Steel fiber directions: placing parallel to the longitudinal direction of the
(Vf a, %) beam specimen and placing transversely to the longitudinal
Cement Water Silica Fine Filler Superplasticizer
fume aggregates direction.
1.00 0.25 0.25 1.10 0.30 0.018 2
Flexural tests were implemented using beam specimens with
dimensions of 100  100  400 mm. Before performing the tests,
a
Fiber volume expressed as a volumetric ratio to the whole volume. a 10 mm notch was made on the lower side at the midspan of
the specimen, as shown in Fig. 3.
A three-point bending test with the notched specimen was con-
ducted to examine the relationship between the orientation distri-
bution of the fibers and the characteristics of tensile behavior of
UHPCC. The tests were carried out through displacement control
using an actuator with a capacity of 250 kN. The crack opening dis-
placement was measured by clip gauge located at the notch as
shown in Fig. 3. The deflection at the center of the flexural speci-
mens was also measured by means of a LVDT installed at the center
of the specimens. The flexural tensile test specimen and apparatus
are illustrated in Fig. 3.

3.2. Experimental results

Table 2 gives the measured loads, CMODs, and deflections at the


moment of first cracking as well as at the point of maximum resis-
tance for the two placing direction cases, respectively. In Table 2,
‘PL’ denotes UHPCC placed parallel to the longitudinal direction
of the beam specimen, which corresponds to the direction tensile
stress acts on in bending, and ‘TL’ indicates UHPCC placed trans-
versely to the longitudinal direction of the specimens. The flexural
Fig. 2. Specimen preparations by (a) placing concrete parallel to the tensile
test results for two different placing methods of UHPCC are pre-
direction of the specimen; and (b) placing concrete transversely to the tensile sented in Fig. 4. The behavior until the initiation of cracking is
direction of the specimen. shown in Fig. 4a and the whole flexural behavior described in the
load versus CMOD relation is given in Fig. 4b. In addition, the test
compared with the analytically obtained results. The comparison results are also presented in the load versus deflection relation,
provides verification of the analytical approach for the flexural seen in Fig. 5.
behavior of the composites. The average of the first cracking load is 24.77 kN for PL speci-
mens and 24.66 kN for TL specimens, respectively. The average
3.1. Test method maximum load is 76.78 kN for the PL specimens and 60.39 kN for
the TL specimens. The experimental results show that the post-
In order to investigate the effect of the fiber orientation distri- cracking flexural behavior is considerably sensitive to how to
bution on the flexural behavior, two types of fiber orientation place; the PL specimens show better performance than the TL

Fig. 3. Flexural tensile test specimen and apparatus.

Table 2
Flexural test results.

Specimen no. At first cracking At maximum load


Load (kN) CMOD (mm) Deflection (mm) Load (kN) CMOD (mm) Deflection (mm)
PL1 24.41 0.015 0.123 79.78 0.780 0.865
PL2 26.39 0.019 0.308 72.82 1.167 1.113
PL3 23.52 0.016 0.383 77.75 1.183 1.342
Mean (St. dev.) 24.77 (1.469) 0.017 (0.002) 0.271 (0.134) 76.78 (3.579) 1.043 (0.228) 1.107 (0.239)
TL1 25.67 0.018 0.302 59.34 0.730 0.974
TL2 27.04 0.018 0.228 60.51 0.765 0.815
TL3 21.27 0.014 0.347 61.34 0.582 0.894
Mean (St. dev.) 24.66 (3.017) 0.017 (0.002) 0.292 (0.060) 60.39 (1.005) 0.692 (0.097) 0.894 (0.079)
60 S.-T. Kang, J.-K. Kim / Construction and Building Materials 28 (2012) 57–65

40 PL Specimens 40 TL Specimens

30 30

Load (kN)

Load (kN)
20 20

10 10

0 0
0 0.02 0.04 0.06 0 0.02 0.04 0.06
CMOD (mm) CMOD (mm)

(a) in the initial area including the first cracking

80 80
PL Specimens TL Specimens

60 Load (kN) 60
Load (kN)

40 40

20 20

0 0
0 2 4 6 0 2 4 6
CMOD (mm) CMOD (mm)

(b) overall load-CMOD curves


Fig. 4. Flexural tensile test results according to the placing direction (in the load–CMOD curve).

specimens. For the first cracking load, there is no noticeable differ- 4. Analysis and discussion
ence between the PL and TL specimens. On the other hand, for the
maximum load, the former is greater than the latter by about 30%, 4.1. Evaluation of fiber orientation distribution
and the corresponding CMOD and deflection are also higher in the
PL specimens. In order to estimate the tensile behavior of UHPCC depending
on fiber orientation distribution, the distribution characteristics

80 80
PL Specimens TL Specimens

60 60
Load (kN)
Load (kN)

40 40

20 20

0 0
0 2 4 6 8 0 2 4 6 8
Deflection (mm) Deflection (mm)

Fig. 5. Flexural tensile test results according to the placing direction (in the load–deflection curve).
S.-T. Kang, J.-K. Kim / Construction and Building Materials 28 (2012) 57–65 61

(a) PL1 (d) TL1

(b) PL2 (e) TL2

(c) PL3 (f) TL3


Fig. 6. Digital images of fiber distribution for PL and TL specimens.

of fibers in a composites should be quantitatively evaluated in ad- The probability density distribution of the fiber orientation for
vance. For this purpose, an image processing technique proposed each specimen is presented in Fig. 7. It can be easily found that
by Lee [25] was adopted. The images were obtained by using the fiber orientation distribution of the PL specimens is a left-
high-resolution camera. skewed distribution whereas the TL specimens show a right-
For the purpose of evaluating the orientation of the fibers skewed distribution. This means that the fibers in the PL specimens
according to the placement directions of UHPCC, the specimens have a tendency to be aligned such that they are more perpendic-
were cut at a cross section, that is, normal to the longitudinal direc- ular to the cut plane (i.e. parallel to the longitudinal direction)
tion of the beam specimen, as close as possible to the cracked plane whereas the TL specimens are more parallel to the cut plane (i.e.
after the flexural test to obtain the images of the cross section. perpendicular to the longitudinal direction). For the purpose of
High resolution photographs were taken at the cross sections using comparison, the probability density distributions of 2-D random
a digital camera. The images were analyzed using the image pro- distribution and 3-D random distribution are presented together.
cessing technique. A comparison of the probability density distribution of fiber orien-
The obtained images of the fiber distribution are presented in tation is presented in Fig. 8 for the aforementioned four cases; the
Fig. 6, where it can be seen that in the case of concrete placed par- probability distributions of PL2, TL2, 2-D random, and 3-D random.
allel to the longitudinal direction of the specimen (PL specimens), Actual fiber orientation distributions (PL2 and TL2) in the speci-
the fibers are more uniformly dispersed compared to the TL spec- mens are quite different from idealized orientation distributions
imens. Furthermore, most of the fibers in the PL specimens are (2-D and 3-D random).
aligned more parallel to the normal direction of the cut plane, com-
pared to the TL specimens. From the image analysis results, the fi- 4.2. Estimation of fiber bridging behavior
ber distribution characteristics, the number of fibers per unit area
(Fn, number/mm2) could be also calculated. Fn of the PL specimens From the probability density functions of fiber orientation, the
is equivalent to 0.335 and roughly 50% higher than that of the TL fiber bridging behaviors can be obtained by means of Eqs. (3)
specimens (0.217), which is closely related to the fiber orientation and (6). The obtained fiber bridging behaviors for the PL and TL
distribution. specimens are shown in Fig. 9, and it reveals that there is little
62 S.-T. Kang, J.-K. Kim / Construction and Building Materials 28 (2012) 57–65

2 Measured 2.5 Measured


2-D random 2-D random
3-D random 3-D random
1.6 2

Probability, p (θ)

Probability, p (θ)
1.2 1.5

0.8 1

0.4 0.5

0 0
0 30 60 90 0 30 60 90
θ (degree) θ (degree)
(a) PL1 (d) TL1

2.5 2 Measured
Measured
2-D random 2-D random
3-D random 3-D random

2 1.6

Probability, p (θ)
Probability, p (θ)

1.5 1.2

1 0.8

0.5 0.4

0 0
0 30 60 90 0 30 60 90
θ (degree) θ (degree)
(b) PL2 (e) TL2

2.5 2 Measured
Measured
2-D random 2-D random
3-D random 3-D random
2 1.6
Probability, p (θ)

Probability, p (θ)

1.5 1.2

1 0.8

0.5 0.4

0 0
0 30 60 90 0 30 60 90
θ (degree) θ (degree)
(c) PL3 (f) TL3
Fig. 7. Probability density distributions for PL and TL specimens.

variation in each case. Therefore, hereafter, PL2 and TL2 are chosen though its performance does not come up to that of the PL
as representatives of the PL and TL specimens in analyzing the ef- specimen.
fect of fiber orientation distribution or comparing with experimen-
tal results. 4.3. FEM analysis and comparison with experimental results
The fiber bridging behaviors for four different cases are com-
pared in Fig. 10. The PL specimen shows much better tensile per- In order to verify the effect of the fiber orientation distribution
formance than the others. The TL specimen is also better than on the flexural behavior, finite element analyses were performed,
the cases with the 2-D and 3-D random fiber distributions, even and the results were compared with the experimental results.
S.-T. Kang, J.-K. Kim / Construction and Building Materials 28 (2012) 57–65 63

2.5 Measured (PL2) 16 by 3-D randomly distributed fibers


Measured (TL2) by 2-D randomly distributed fibers
2-D random for PL Specimen
3-D random for TL Specimen
2
12

tensile stress (MPa)


Probability, p (θ )

1.5

4
0.5

0
0
0 30 60 90
0 2 4 6 8
θ (degree)
Crack width (mm)
Fig. 8. Comparison of probability density distributions for different fiber orienta-
tion distributions. Fig. 10. Comparison of fiber bridging behaviors calculated with Eq. (3) for different
fiber orientation distributions.
16
PL1
PL2
PL3 16
Tensile bridging stress (MPa)

for PL Specimen
12
for TL Specimen
Tensile bridging stress (MPa)

12 for 2-D distribution

8 for 3-D distribution

0
0 2 4 6 8
0
Crack width (mm)
0 2 4 6 8
(a) PL specimens
Crack width (mm)
16 Fig. 11. Comparison of the simplified curves by Eqs. (7)–(9) with the fiber bridging
TL1
TL2 behaviors obtained from the fiber orientation distribution.
TL3
Tensile bridging stress (MPa)

12
In the finite element modeling, crack propagation at the notch
was simulated by adopting the fictitious crack model, for which
the interface elements were introduced. The pre-cracking tensile
behavior of the interface elements is linear elastic, and the post-
8
cracking behavior follows the combined curve of the matrix soften-
ing and the fiber bridging curve which is determined considering
the fiber orientation distribution.
The matrix softening curve is expressed by Eqs. (4)–(6). For the
4
FEM analysis, it is necessary to introduce a simplified equation for
the fiber bridging behavior. In order to express the behavior, a
quartic equation was adopted for the ascending branch of the
0
bridging curve, and the descending branch was assumed to be lin-
ear. The quartic equation can be expressed as follows:
0 2 4 6 8
"  2  3  4 #
Crack width (mm) 4 lf w w w 1 w
rB ¼ V f sf 2 3 þ2 
(b) TL specimens p df ws ws ws 2 ws
for w 6 ws ð7Þ
Fig. 9. Calculated fiber bridging behaviors of PL and TL specimens.
64 S.-T. Kang, J.-K. Kim / Construction and Building Materials 28 (2012) 57–65

80 by 3-D randomly distributed fibers 80


by 2-D randomly distributed fibers
for PL Specimen Experiment
for TL Specimen
Analysis

60 60

Load (kN)
Load (kN)

40 40

20 20

0 0
0 2 4 6 8 10 0 2 4 6 8
CMOD (mm) CMOD (mm)

(a) load-CMOD curves (a) PL specimen

80 by 3-D randomly distributed fibers 80


by 2-D randomly distributed fibers Experiment
for PL Specimen
for TL Specimen Analysis

60 60
Load (kN)

Load (kN)

40 40

20 20

0 0
0 2 4 6 8 0 2 4 6 8
Deflection (mm) CMOD (mm)

(b) load-deflection curves (b) TL specimen


Fig. 12. Comparison of flexural behavior according to the fiber orientation Fig. 13. Comparison of analytical and experimental flexural behavior (in load–
distribution: (a) load–CMOD curves; and (b) load–deflection curves. CMOD relation).

The bridging behaviors for four different cases are compared with
the simplified curves obtained from Eqs. (7)–(9) in Fig. 11, demon-
In Eq. (7), ws is the crack width when the fiber starts to slide at strating that the equations express the bridging behaviors very well.
the interface, which corresponds with the point when the bridging After determination of the bridging curves, FEM analyses were
stress reaches the maximum, and sf is the frictional shear strength. performed. The analytical results obtained from the FEM analysis
lf is the length of the fiber, df is the diameter of the fiber, and Vf is for the notched beams are presented in Fig. 12, which provides
the volume fraction of the fiber. Eq. (7) is based on the equation load–CMOD curves as well as load–deflection curves. The analyti-
suggested by Grzybowski and Shah [26,27]. They proposed a rela- cally obtained flexural behaviors are compared with experimental
tionship between the pullout force of the fiber that normally acts results for the PL and TL specimens in Figs. 13 and 14. The analyt-
on the unit area of the crack surface and the crack width, based ical curves are in good agreement with the experimental results in
on pullout test results. In this study, considering the characteristics both load–CMOD curve and load–deflection curve. Therefore, it can
of the pullout behavior of UHPCC [23], the modified quartic equa- be said that the prediction of flexural behavior based on the actual
tion was introduced. fiber orientation distribution in UHPCC is pretty reliable.
For the descending branch of the tensile bridging behavior, the
linear softening curve was assumed based on Fig. 10, expressed by: 5. Conclusions
 
2 lf w  ws In this study, the effect of the fiber orientation distribution on
rB ¼ V f sf 1  for w 6 0:85Lf ð8Þ
p df 0:85lf  ws the flexural behavior of UHPCC was investigated. For the purpose,
a three-point bending test with the notched specimens having dif-
rB ¼ 0 for w > 0:85Lf ð9Þ ferent fiber orientation distributions was carried out and the fiber
S.-T. Kang, J.-K. Kim / Construction and Building Materials 28 (2012) 57–65 65

80 obtained in a structure or flow-induced fiber rotational movement


Experiment in fresh composites can be predicted in advance, the flexural per-
formance can be easily expected with the help of the analytical ap-
Analysis
proach suggested in this study.
60
Acknowledgments
Load (kN)

This study has been a part of a research project supported by


40 Korea Institute of Construction Technology (KICT). The authors
wish to express their gratitude for the financial support that made
this study possible.

20 References

[1] Mangat PS. Tensile strength of steel fiber reinforced concrete. Cem Concr Res
1976;6(2):245–52.
[2] Fanella DA, Naaman AE. Stress-strain properties of fiber reinforced mortar in
0 compression. ACI J 1985;82(4):475–83.
[3] Ezeldin AS, Balaguru PN. Normal- and high-strength fiber-reinforced concrete
0 2 4 6 8
under compression. J Mater Civil Eng 1992;4(4):415–29.
Deflection (mm) [4] Jianming G, Wei S, Keiji M. Mechanical properties of steel fiber-reinforced,
high-strength, lightweight concrete. Cem Concr Compos 1997;19(4):307–13.
(a) PL specimen [5] Nataraja MC, Dhang N, Gupta AP. Stress–strain curve for steel-fiber reinforced
concrete under compression. Cem Concr Compos 1999;21(5):383–90.
[6] Song PS, Hwang S. Mechanical properties of high-strength steel fiber-
80 reinforced concrete. Constr Build Mater 2004;18(9):669–73.
Experiment [7] Yazici S, Inan G, Tabak V. Effect of aspect ratio and volume fraction of steel fiber
on the mechanical properties of SFRC. Constr Build Mater 2007;21(6):1250–3.
Analysis [8] Kang ST, Lee Y, Park YD, Kim JK. Tensile fracture properties of an ultra high
performance fiber reinforced concrete (UHPFRC) with steel fiber. Compos
60 Struct 2010;92(1):61–71.
[9] Morton J, Groves GW. The cracking of composites consisting of discontinuous
reinforced concrete. J Mater Sci 1974;9(9):1436–45.
Load (kN)

[10] Brandt AM. On the optimal direction of short metal fibres in brittle matrix
components. J Mater Sci 1985;20:3831–41.
40 [11] Chiang CR. A statistical theory of the tensile strength of short-fiber-reinforced
composites. Compos Sci Technol 1994;50(4):479–82.
[12] Ozyurt N, Mason TO, Shah SP. Correlation of fiber dispersion, rheology and
mechanical performance of FRCs. Cem Concr Compos 2007;29(2):70–9.
[13] Kim JK, Kim JS, Ha GJ, Kim YY. Tensile and fiber dispersion performance of ECC
20 (engineered cementitious composites) produced with ground granulated blast
furnace slag. Cem Concr Res 2007;37(7):1096–105.
[14] Kang ST, Lee BY, Kim JK, Kim YY. The effect of fibre distribution characteristics
on the flexural strength of steel fibre-reinforced ultra high strength concrete.
Constr Build Mater 2011;25(5):2450–7.
0 [15] Kang ST, Park JJ, Ryu GS, Kim SW. Investigation of fiber alignment of UHSFRC in
flexural members. In: Proceedings of 8th International symposium on
0 2 4 6 8
utilization of high-strength and high-performance concrete, Tokyo, Japan;
Deflection (mm) 2008. p. 709–14.
[16] Wuest J, Denarié E, Brühwiler E, Tamarit L, Kocher M, Galluci E. Determination
of fibre distribution and orientation in fibre reinforced composites. Exp Tech
(b) TL specimen 2009;33(5):50–5.
[17] Wuest J, Denarié E, Brühwiler E. Measurement and modelling of fibre
Fig. 14. Comparison of analytical and experimental flexural behavior (in load– distribution an d orientation in UHPFRC. In: Proceedings of the Fifth
deflection relation). International RILEM Workshop on High Performance Fiber Reinforced
Cement Composites (HPFRCC5), Mainz, Germany; 2007. p. 259–66.
[18] Wuest J, Denarié E, Brühwiler E. Model for predicting the UHPFRC tensile
orientation distribution was quantitatively estimated by the help hardening response. In: Proceedings of the Second International Symposium
of image analysis process. The measured fiber orientation distribu- on Ultra High Performance Concrete, Kassel, Germany; 2008. p. 153–60.
tion from two kinds of beam specimens, which were fabricated [19] Kang ST, Kim JK. The relation between fiber orientation and tensile behavior in
an ultra high performance fiber reinforced cementitious composites
with different placing methods to induce different flexural perfor- (UHPFRCC). Cem Concr Res 2011;41(10):1001–14.
mance, confirmed that the fiber orientation distribution has a [20] Cox HL. The elasticity and strength of paper and other fibrous materials. Brit J
strong impact on the flexural behavior, especially on the deflection Appl Phys 1952;3(3):72–9.
[21] Piggott MR. Short fibre polymer composites: a fracture-based theory of fibre
hardening behavior. reinforcement. J Compos Mater 1994;28(7):588–606.
In order to verify the effect of the fiber orientation distribution [22] Li VC, Wang Y, Backer S. A micromechanical model of tension-softening and
on the flexural behavior, finite element analyses were performed. bridging toughness of short random fiber reinforced brittle matrix composites.
J Mech Phys Solids 1991;39(5):607–25.
Crack propagation was simulated by adopting the fictitious crack [23] Lee Y, Kang ST, Kim JK. Pullout behavior of inclined steel fiber in an ultra-high
model, for which the interface elements were introduced. The strength cementitious matrix. Constr Build Mater 2010;24(10):2030–41.
pre-cracking tensile behavior of the interface elements is linear [24] Markovic I. High-performance hybrid-fiber concrete-development and
utilisation. Delft University of Technology, Doctorial Thesis, Netherland;
elastic, and the post-cracking behavior follows the combined curve
2006. p. 50–7.
of the matrix softening and the fiber bridging curve which is deter- [25] Lee BY. Fiber distribution evaluation using digital image processing and its
mined considering probabilistic fiber orientation distribution. The effect on tensile behavior of fiber reinforced cement composites. Korea
FEM analysis revealed that the analytical curves correspond well Advanced Institute of Science and Technology, Doctorial Thesis, Korea; 2009.
p. 5–43.
with the experimental results. Therefore such analytical approach [26] Grzybowski M, Shah SP. Model to predict cracking in fibre reinforced concrete
makes it possible to predict the flexural behavior considering the due to restrained shrinkage. Mag Concr Res 1989;41:125–35.
effect of fiber orientation distribution. Furthermore, if the mea- [27] Grzybowski M, Shah SP. Shrinkage cracking of fiber reinforced concrete. ACI
Mater J 1990;87(2):138–48.
surement of the actual fiber orientation distribution can be

You might also like